Early Thiamine Deficiency

Contemporary Clinical Neuroscience
Series Editors
Ralph Lydic, Ph.D.
Department of Anesthesiology
The University of Michigan
Ann Arbor, Michigan
United States
Helen A Baghdoyan, Ph.D.

Department of Anesthesiology
The University of Michigan
Ann Arbor, Michigan
United States
For other titles published in this series, go to

www.springer.com/series/7678
David W. McCandless
Thiamine Deficiency
and Associated Clinical
Disorders
David W. McCandless
Department of Cell Biology & Anatomy
Rosalind Franklin University
3333 Green Bay Road
North Chicago, IL 60064
USA
david.mccandless@rosalindfranklin.edu
ISBN 978-1-60761-310-7 e-ISBN 978-1-60761-311-4

DOI 10.1007/978-1-60761-311-4
Library of Congress Control Number: 2009930939
© Humana Press, a part of Springer Science+Business Media, LLC 2010

All rights reserved. This work may not be translated or copied in whole or in part without the written
permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring
Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly
analysis. Use in connection with any form of information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed is
forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are
not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject
to proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of going
to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any
errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect
to the material contained herein.
Printed on acid-free paper
springer.com
This volume is dedicated to Sue, April,
Elizabeth, and Andrew
Preface
The past 20 years have seen remarkable advances in neuroscience, neurology, imag-
ing techniques, and diagnostic strategies. These advances have been successfully
applied to many different diseases, including thiamine deficiency and associated
clinical disorders. Syndromes such as beriberi, Wernicke’s disease, Leigh’s dis-
ease, African Seasonal Ataxia, and various inherited ataxias have all benefited
from improved scientific approaches. It is timely therefore to examine new knowl-
edge related to the effects of thiamine deficiency on organ systems and to specific
thiamine-related clinical disorders.
The elucidation of the biochemistry of thiamine took many scientists many years
to complete and is inextricably linked to the study of thiamine deficiency. Thiamine
(vitamin B1) is a water-soluble compound which consists of a pyrimidine nucleus
and a thiazole ring. A key derivative of thiamine is thiamine pyrophosphate (TPP),
called cocarboxylase. This form is a coenzyme which participates in the decarboxy-
lation of several essential intermediates involved in carbohydrate metabolism. ATP
is involved in the transfer of phosphate to thiamine. The two key decarboxylation
reactions are the decarboxylation of pyruvic acid and of alpha-ketoglutaric acid. The
enzyme transketolase is another enzyme requiring TPP as coenzyme. This enzyme
is important in the hexose monophosphate shunt. The dietary requirement for thi-
amine is based largely on the caloric intake, and primarily that of carbohydrates.
Thiamine is widely distributed in plants and animal tissues and is found in high
concentrations in yeast, beans, wheat germ, oats, ham, soy beans, etc.
Thiamine is a naturally occurring vitamin which has many functions in mam-
mals. An early and frequently studied role of thiamine is its participation as
coenzyme in enzymatic functions in energy metabolism. Clinical descriptions of
thiamine-related disorders were published many years before the association of their
effects to thiamine deficiency. For example, beriberi was first described in medical
literature in 1642 by the Dutch physician Jacobus Bontius (some of his writings
related to beriberi have been translated and appear in Appendix A). It was many
years later before the association between thiamine and beriberi was discovered.
The enormous interest in thiamine deficiency and the associated clinical disorders
stems from the fact that these represent classical metabolic encephalopathies, and
as such are reversible (treatable) when recognized early.
vii
viii Preface
As stated above, thiamine plays an important role as coenzyme in reactions in

the tricarboxylic acid cycle and in the hexose monophosphate shunt. Thiamine defi-
ciency produces symptoms which are primarily localized to cardiac and nervous
tissues. These symptoms can be rapidly reversed by thiamine administration, which
led to the important concept of a biochemical lesion. The term “biochemical lesion”
implies a lesion which occurs before morphological lesions, the presence of which
may render the disorder less amenable to treatment and reversal.
Thiamine may also play a role in neurotransmission, as well as several other key
biochemical pathways. Perturbation of these important metabolic events may also
contribute to key neurochemical alterations, and their reversal in thiamine deficiency
and associated clinical syndromes. The fact that frequently the symptoms associated
with both pure thiamine deficiency and those in human disorders related to thiamine
deficiency can be reversed (successfully treated) in early stages is promising. If
the biochemical lesion can be recognized and corrected before the onset of less
reversible structural changes, recovery and return to normal function is possible.
Clinical disorders shown to be directly related to thiamine deficiency include
beriberi, Wernicke’s disease, Leigh’s disease, African Seasonal Ataxia (ASA),
inherited ataxias, central pontine myelinolysis, and others.
Pure beriberi heart disease is rare in the United States; however, it is endemic
in other parts of the world. The majority of cases of beriberi (85%) are subacute
and mild. The remainder of cases may be more severe, and if not treated, death
may ensue. Since beriberi is a complicated nutritional disorder, it is not necessarily
a disorder of thiamine deficiency alone; however, treatment with thiamine usually
reverses symptoms, sometimes in only hours. There is a long, rich history relating
to beriberi, including the pioneering work of Christian Eijkman, which resulted in
the awarding of a Nobel Prize (Dr. Eijkman’s Nobel Prize speech is reproduced in
Appendix C).
Wernicke’s disease, first described in 1881 by Carl Wernicke, occurs frequently
in the United States and is usually associated with chronic alcoholism. Like beriberi,
Wernicke’s disease is a complicated nutritional disorder, but in the early stages,
many symptoms can be reversed by thiamine administration. Wernicke’s disease
may have a rapid onset, and neurological symptoms consist largely of confusion,
ataxias, nystagmis, and ocular palsies. Other brain stem symptoms may include stu-
por and coma. Many Wernicke’s patients may lapse into the more chronic disorder,
Korsakoff’s psychosis. In later stages, symptoms may be associated with structural
alterations in specific brain sites. Classic studies by Dr. Maurice Victor and col-
leagues on Wernicke’s disease were carried out over many years, and are the basis
for many publications in this area.
Leigh’s disease (subacute necrotizing encephalomyelopathy) is an uncommon
disease with around 200 cases reported in the literature. This disorder seems to be
genetically transmitted. In patients diagnosed with Leigh’s disease, there was found
in urine an inhibitory substance which blocked the formation of thiamine triphos-
phate (TTP). Low levels of TTP were found in tissues of Leigh’s disease patients.
This was postulated to be responsible for thiamine-deficiency-like symptoms seen
in these patients. Leigh’s disease usually begins to be symptomatic around postnatal
Preface ix
month 6–12 following normal development. There appears lethargy, nystagmus,

ataxia, failure to thrive, stupor, coma, and death. The lesions in the brain are usually
more widespread than those in Wernicke’s disease, but similar. Thiamine levels in
Leigh’s disease patients’ blood and tissue are normal, and thiamine administration
does not ameliorate the condition. A test for the inhibitor did not provide a con-
clusive indication of the presence of Leigh’s disease, there being significant false
positives and false negatives. Recent studies have shifted away from the inhibitor
and examined supposed Leigh’s disease patients for genetic defects including those
with cytochrome c oxidase deficiency. Leigh’s disease may represent one variation
of the inherited ataxias. These studies and their ramifications will be described in
depth. There are many unanswered questions relating to this fascinating disorder.
African Seasonal Ataxia (ASA) is another interesting recently described clini-
cal entity which presumably has a thiamine-related foundation. This syndrome has
been recently described in people who live in Western Nigeria and is characterized
by ataxia, tremors, and decreased levels of consciousness. These symptoms occur
during the rainy season of July through October. ASA usually follows a large carbo-
hydrate meal. At its peak incidence, ASA can account for well over 70% of hospital
and clinic admissions.
Various hypotheses have been advanced to explain ASA; however, strong evi-
dence supports a mechanism related to thiamine deficiency. There is a clinical triad
of cerebellar ataxia, ocular disturbances, and encephalopathy usually seen in acute
thiamine deficiency. Upon examination of the dietary intake of the low socioe-
conomic strata of patients, it was found that almost all had consumed significant
amounts of roasted silkworm larvae Anaphe venata. The availability of these larvae
in the marketplace corresponds to the rainy season. The larvae represent a valuable
protein source for rainforest people.
The practice of entomophagy in low socioeconomic cultures is accepted. Protein
sources are relatively scarce for these people, who subsist largely on carbohydrate-
rich diets. Subsequently, it has been shown that there is a thiaminase present in the
Anaphe venata larvae. During the rainy season, these larvae fall from specific trees,
and are gathered and sold in markets. Subsequent consumption of larvae containing
a thiaminase by people ordinarily eating carbohydrate-rich diets can explain the
rapid onset of symptoms resembling those of thiamine deficiency. As a corollary
to this recent description are earlier descriptions of similar outbreaks of thiamine
deficiency. There was, for example, the outbreak of the so-called Chastek paralysis,
a disorder of silver foxes on a fox ranch in Minnesota. These foxes were fed raw
fish and within a few weeks developed ataxia, changes in consciousness, seizures,
and death. Pathologically, brain lesions resembled those seen in thiamine deficiency.
Subsequent work showed the presence of a thiaminase in the viscera of the raw fish,
which had precipitated the disorder.
Inherited ataxias are a group of relatively rare neurological disorders genetically
transmitted, which have as a common denominator ataxia and the possibility of
successful thiamine treatment. These diverse, yet related ataxias include Refsum’s
disease and Friedreich’s ataxia. This group of disorders has a defect in the enzyme
pyruvate decarboxylase. Pathologically, these disorders show mitochondrial damage
x Preface
in selective brain regions. Treatment regimes include thiamine and ketonic diet ther-
apies.
The disorder called central pontine myelinolysis (CPM) represents one variant
almost always associated with some other serious chronic disease. Over one half
of the reported cases were associated with alcoholism and Wernicke’s disease. As
the name implies, the key feature is a pathological lesion consisting of a symmetric
focus of demyelination localized at the center of the pons. Microscopically, there is
demyelination of medulated fibers. A variation of CPM are cases which, in addition
to the lesion in the pons, demonstrate demyelinating lesions in other brain regions
including the thalamus, cerebellum, cerebral cortex amygdala, putamen, etc. In spite
of the predominance of alcoholism and Wernicke’s disease as a feature of CPM
patients, there are many instances of CPM in serious disorders not associated with
thiamine involvement. The relation between thiamine and CPM therefore remains
somewhat unclear.
Marchiafava-Bignami disease is yet another relatively rare disorder, which has
been associated with thiamine deficiency. This clinical entity has as its distinguish-
ing characteristic, lesions in the corpus callosum. These lesions lead to specific signs
and symptoms, which together with MRI can lead to early diagnosis. Before imag-
ing advances, the diagnosis was almost always made at autopsy. As a result of rapid
diagnosis, treatment regimes involving thiamine administration have evolved, and
cases of Marchiafava–Bignami disease have been successfully treated.
In summary, thiamine deficiency and associated clinical disorders represent an
intriguing area of both basic and clinical investigation. Modern imaging and diag-
nostic strategies have facilitated the rapid treatment, and potential reversal of these
clinical disorders. The fusion of laboratory and clinical knowledge serve as an exam-
ple of how research can translate to successful treatment. This book is designed
to bring together cogent results from both basic and clinical investigation. These
data will be of interest to neurologists, internists, nutritionists, biochemists, neu-
rochemists, neuroscientists, and many others with interest in thiamine deficiency.
Many questions regarding these clinical disorders as well as thiamine deficiency
remain unanswered. We hope this book may serve to stimulate further investiga-
tions in these areas.
Acknowledgments
The author wishes to thank Dr. Joseph X. DiMario for support and consultation
regarding several aspects of this book. The author also thanks Dr. Roland Auer and
Dr. Jeffrey Joseph, University of Calgary, Department of Pathology and Labora-
tory Medicine, Calgary, Alberta, Canada, and Dr. Michael D. Norenberg, Direc-
tor of Neuropathology, Department of Pathology, University of Miami School of
Medicine, Miami, FL, for supplying many fine figures from brain slices, and pho-
tomicrographs. These illustrations add much to the visualization of features such
as the highly localized nature of lesions in areas such as the mammillary bodies.
Dr. Bola Adamolekun, Department of Neurology, University of Tennessee College
of Medicine, graciously supplied me with several pounds of Anaphe venata larvae,
shown in Figs. 1 and 2 of the chapter “African Seasonal Ataxia.”
This volume would not have been possible were it not for the dedicated partic-
ipation of Ms. Cristina Gonzalez. She was tireless in searching out old and new
references, preparing tables, and organizing figures and figure permissions. Ms. Vil-
mary Friederichs also participated in organizing the manuscript and gave advice
in the preparation of the volume. My son Dr. Jeffrey McCandless, NASA Ames
Research Center, Palo Alto, CA, helped in the design of many excellent figures and
in improving others. Finally, Ms. Ann Avouris, editor at Springer Science and Busi-
ness Media, was outstanding in answering questions, helping obtain permissions
for use of previously published figures, and in general shepherding this manuscript
through the publishing process.
xi
Contents
Early Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Early Thiamine Deficiency . . . . . . . . . . . . . . . . . . . . . . . . . 9
Thiamine Deficiency in Mammals . . . . . . . . . . . . . . . . . . . . . 17
Beriberi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Wernicke’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Central Pontine Myelinolysis, Alcoholic Cerebellar
Degeneration, and Marchiafava–Bignami Disease . . . . . . . . . . . . 65
Central Pontine Myelinolysis . . . . . . . . . . . . . . . . . . . . . . . . . 65
Alcoholic Cerebellar Degeneration . . . . . . . . . . . . . . . . . . . . . . 72
Marchiafava–Bignami Disease . . . . . . . . . . . . . . . . . . . . . . . . 75
Leigh’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
African Seasonal Ataxia . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Inherited Ataxias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Thiamine Deficiency in Serious Illness . . . . . . . . . . . . . . . . . . . 131
Beriberi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Intensive Care Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Hematologic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Gastrectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
5-Fluorouracil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Wernicke’s Disease and Gastrectomy . . . . . . . . . . . . . . . . . . . . . 139
HIV/AIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Malaria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Amyotrophic Lateral Sclerosis . . . . . . . . . . . . . . . . . . . . . . . . 142
Alzheimer’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
World Health Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Epilogue: Future Prospects . . . . . . . . . . . . . . . . . . . . . . . . . 151
xiii
xiv Contents
Animal Models of Thiamine Deficiency . . . . . . . . . . . . . . . . . . . 151

Beriberi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Wernicke’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
CPM/ACD/MBD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Leigh’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Inherited Ataxias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Serious Illnesses and Thiamine Deficiency . . . . . . . . . . . . . . . . . . 159
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Appendix A: Bontius Translation . . . . . . . . . . . . . . . . . . . . . . 161
Appendix B: Thiamine Synonyms . . . . . . . . . . . . . . . . . . . . . 163
Appendix C: The Nobel Prize Acceptance Speech
of Dr. Christiaan Eijkman, 1929 . . . . . . . . . . . . . . . . . . . . . . 165
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Early Chemistry
The early chemistry of thiamine deficiency goes back many years. The purpose of
this introductory chapter is to review some of the remarkable discoveries relating to
thiamine, thiamine-dependent enzymes, the relation to a deficiency of thiamine, and
results, which led to the awarding of a Nobel prize to Christiaan Eijkman in 1929 in
physiology or medicine for his work involving thiamine.
Remember that the first medical description of beriberi was published in 1642 by
Jacobus Bontius (see Appendix A for a translation). Christiaan Eijkman, a physi-
cian, was able to demonstrate 250 years later that birds that were fed a diet of pol-
ished rice gradually progressed to a point where they became moribund and died.
Eijkman had earlier tried unsuccessfully to isolate infectious agents that might have
been responsible for beriberi. Eijkman next demonstrated that if the birds were fed
unpolished rice, their symptoms reversed. The reversal feature of what was thiamine
deficiency (but not known by Eijkman) was demonstrated very early in the investi-
gations of the neurological (polyneuritis) outcomes of feeding polished rice. Also,
Eijkman observed that the symptoms seen in pigeons closely resembled those seen
in human beriberi.
The following is a description from Eijkman quoted verbatim concerning the
symptoms in birds:
“The beginning of the disease is characterized by an unsteady gait which first of all mani-
fests itself in walking about on the perch, as if the animal cannot squeeze its toes around it
firmly enough, and must exert itself in order not to fall off. The disturbance in mobility soon
increases in intensity and speed. The fowl no longer has the strength to climb up; because of
weakness it holds its limbs spread apart and bent at the knee and ankle joints; when running
it frequently collapses or falls over. Finally, it remains lying on its side and in its fruitless
efforts the developing paralysis of the wing muscles also becomes noticeable. The paralysis
of the body musculature rapidly progresses from below upward.”
“The involvement of the peripheral nerves is the most important feature that post-mortem
investigations reveal to date. It involves both the sensory and motor portions, which occur
focally in the nerve trunks, and produces the picture of non-inflammatory atrophic degen-
eration such as is observed after transection of a nerve in the distal portion of the divided
fragment. However, definite changes in the spinal cord and spinal cord roots are also not
lacking. These show, likewise, the appearance of degeneration and atrophy.”
D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 1

Contemporary Clinical Neuroscience, DOI 10.1007/978-1-60761-311-4_1,

C Humana Press, a part of Springer Science+Business Media, LLC 2010
2 Thiamine Deficiency and Associated Clinical Disorders
Some years later, Eijkman and Vorderman (Eijkman, C. and Vorderman, A.,
1897) were able to show that it was the “polishing” of the rice that was causing
the symptoms of polyneuritis that were prevalent in birds and in man suffering the
disease beriberi. Later it was discovered that a very small amount of a substance
called “vitamine” by Funk (1911) was the compound which was essential for nor-
mal health, and the substance whose absence was responsible for symptoms in birds
and man. The final “e” of vitamine was dropped when it was learned that the sub-
stances were not amines.
Early on Eijkman knew that polyneuritis was linked to the type of rice eaten.
He believed that the symptoms were due to an infectious agent or toxin on the rice.
Later it was shown that there was no toxin or infectious agent associated with the
rice or its preparation.
Many investigators tried unsuccessfully to purify and isolate the mysterious vita-
min until Jansen and Donath (1926) succeeded. It was a decade later before the
vitamin was synthesized in the laboratory of R. R. Williams. Jensen and Donath
sent a small amount of the compound to Eijkman who was able to show that as
little as 2 µg was enough to protect pigeons from the deleterious effects of a diet
of polished rice. Dr. Christiaan Eijkman received the Nobel Prize in Physiology or
Medicine in 1929. His acceptance speech appears in Appendix C.1
Later in 1936, the structure (Williams, R., 1936), then the synthesis of the vitamin
(Williams, R. and Cline, J., 1936) appeared as a communication to the editor of the
Am. J. Chem. Soc. “Many laboratories were working on this problem, and later
were able to confirm Williams” work.
The structure for thiamine put forth by Williams is as follows (Fig. 1):
Fig. 1
Using liquid ammonia, Williams was able to produce a free base of thiamine,
C6 H10 N4 . A couple of months later, the synthesis of thiamine was briefly outlined
by Williams. The synthesis included ammonia, POCl2 , and HBr. The structure and
synthesis of thiamine was the culmination of many years of hard work.
1 Itis important to note that Frederick Gowland Hopkins shared the Nobel Prize in Physiology or
Medicine with Eijkman in 1929. It was awarded to Hopkins for his work on vitamins and for his
finding that muscle contraction leads to lactate accumulation.
Early Chemistry 3
The active form of thiamine is thiamine pyrophosphate or cocarboxylase. This

active form has a structure as follows (Fig. 2):
Fig. 2
One of the first to synthesize thiamine pyrophosphate was Weil-Malherbe (1940),

who synthesized the compound by treating thiamine bromide with silver pyrophos-
phate. It was also shown about the same time that yeast, bacteria, and animal tissues
could also generate thiamine pyrophosphate enzymatically (Fig. 3).
Fig. 3
Finally it was shown (Lohmann and Schuster, 1937) that thiamine pyrophosphate
was the coenzyme “active” form of thiamine.
Three key enzymes in carbohydrate metabolism require thiamine in its phos-
phorylated form (thiamine pyrophosphate) in order to function properly. The two
enzymes pyruvate and alpha-ketoglutarate dehydrogenases are of particular interest
because of their pivotal role in metabolism. Pyruvate dehydrogenase is important
in lipid synthesis, gluconeogenesis, and oxidation via the Krebs (TCA) cycle. The
Krebs cycle produces much needed energy in the form of ATP. The third thiamine-
requiring enzyme is transketolase, which is the rate limiting step in the hexose
monophosphate shunt, which produces reducing power (NAD/NADH) and ribose.
The pyruvate dehydrogenase complex is regulated largely by product inhibition.
ATP in micromolar concentrations is able to inhibit kidney pyruvate dehydrogenase
(Reed, L., 1976). Similar compounds such as ADP, CTP, and GTP were not shown
to be regulatory. Magnesium is also an important component of the enzyme com-
plex, and when pyruvate dehydrogenase is phosphorylated, but inactivated, Mg can
activate the enzyme.
The alpha-ketoglutarate dehydrogenase complex is large (2.7 million M.W.). It

can be separated from other dehydrogenase complexes by fractionation (Koike, M.
et al., 1976). Highly purified enzymes were analyzed for substrate specificity.
Results showed that the pyruvate dehydrogenase enzyme complex has a rather wide
range of substrate specificity in that it can decarboxylate other substrates besides
pyruvate, including alpha-ketobutyrate, alpha-ketovalarate, and alpha-ketoceproate,
among others. The enzyme alpha-ketoglutarate dehydrogenase was found to be spe-
cific for alpha-ketoglutarate. Both enzymes catalyze the first step of the two enzymes
which decarboxylate pyruvate or alpha-ketoglutarate.
The possibility of important regulation of the pyruvate dehydrogenase enzyme
is based on its critically important location in glycolysis, in lipid metabolism, and
in the Krebs cycle. Studies have shown that the enzyme is inhibited by its products
such as acetyl-CoA and NADH, and the enzyme reverts to normal activity in the
presence of NAD and CoA. Other keto acids mentioned above were inhibitory for
the pyruvate dehydrogenase enzyme complex. The binding of thiamine pyrophos-
phate to pyruvate dehydrogenase is loose, making the activity highly dependent on
available thiamine pyrophosphate.
Examination of the effect of the three major methods of producing thiamine defi-
ciency in mammals (diet only, diet plus oxythiamine, and diet plus pyrithiamine)
shed some light on the enzymes that thiamine pyrophosphate serves as coenzyme
(Gubler, C. and Johnson, L., 1967). Animals were Sprague-Dawley rats, and con-
trols received 10 µg/100 g/day of thiamine supplementation. After sacrifice, tissue
was separated, and pyruvate decarboxylase and alpha-ketoglutarate decarboxylase
was measured in heart, liver, kidney, and brain. Results showed that pyrithiamine
plus deficient diet decreased pyruvate decarboxylase in all tissues examined, while
alpha ketoglutarate decarboxylase was decreased only in brain and kidney. Oxithi-
amine plus deficient diet caused a decrease in pyruvate decarboxylase in liver, heart,
and kidney, while alpha-ketoglutarate was only reduced in the heart. Replacement
of the carbohydrate diet (thiamine deficient) with a high-fat diet (which could have
“fed” the system) had only a trivial effect.
Further studies showed that pyrithiamine was an inhibitor of the enzyme thi-
amine pyrophosphokinase, thus blocking the formation of thiamine pyrophosphate.
Oxythiamine formed oxythiamine pyrophosphate, which was in turn an inhibitor
of both pyruvate decarboxylase and alpha-ketoglutarate decarboxylase, as well as
transketolase.
The structure of the actual enzymes pyruvate dehydrogenase and alpha-
ketoglutarate dehydrogenase have been studied in part with electron microscopy
(Reed, L., 1967).
The thiamine-requiring enzyme alpha-ketoglutarate dehydrogenase is actually
composed of three enzymatic subunits: alpha-ketoglutarate decarboxylase, dihy-
drolipoyl transsuccinylase, and dihydrolipoyl dehydrogenase. When the three sub-
units are combined, they automatically form a large molecule, which resembles the
structure of the alpha-ketoglutarate dehydrogenase.
The thiamine-requiring enzyme pyruvate dehydrogenase is also composed
of three subunits: pyruvate decarboxylase, dihydrolipoyl transacetylase, and
Early Chemistry 5
dihydrolipoyl dehydrogenase. This composition is very similar to that of alpha-

ketoglutarate dehydrogenase. These three components of pyruvate dehydrogenase
form a three-dimensional configuration such that during enzyme functioning, the
components are brought close enough to each other to function properly.
Electron microscopic studies of pyruvate dehydrogenase enzyme show a poly-
hedral form with a diameter of about 400 Å. Electron micrographs show that the
transketolase moiety is located in the center of the polyhedron, and the other two
complexes are located around the transacetylase. Electron micrographs also show
the transacetylase moiety has eight subunits at the eight points of a cube. The
transacetylase component is key to the overall three-dimensional form of the pyru-
vate dehydrogenase complex.
In terms of the biosynthesis of pyruvate dehydrogenase, studies indicate that
the three-dimensional configuration of the complex is under direct genetic con-
trol. Little wonder that such a complex structural component, under genetic con-
trol, might have a variety of point mutations, any one of which could affect the
functional capacity of either of these two thiamine-requiring enzyme complexes,
pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase.
Many of the structural/three-dimensional studies have initially been performed
in bacteria. Later studies in mammalian tissues such as heart and kidney show quite
similar structural results. Electron microscopic examination show striking similari-
ties to the enzyme in bacteria-based studies. Even down to basic detail, such as the
position of the subunits on the transacetylase, components are similar to the bacteria
counterpart.
A specific three-dimensional structure may enhance the ability of the complex
to perform its duty. In addition, there may be a significant increase in the efficiency
of the complex to dehydrogenate substrate. The close association of the three com-
ponents should facilitate activity. This might be a sort of evolution in terms of effi-
ciency of enzyme capacity to function.
Experimental evidence supports the concept that there is a protein which binds
thiamine in bacteria (Escherichia coli) (Nose, Y. et al., 1976). Data examining the
inhibition of thiamine binding to protein showed that classical thiamine analogs
such as oxythiamine and pyrithiamine had little effect on binding, but that thiamine
monophosphate, thiamine diphosphate, and thiamine triphosphate had over 50%
inhibition. Chemical treatments such as iodine and bromosuccinimide had an over
50% inactivation of the binding of thiamine to protein.
The regulation of pyruvate dehydrogenase has been examined (Shen, L. et al.,
1968). In this study, pyruvate dehydrogenase was isolated from E. coli. The enzyme
was purified by protamine sulfate precipitation and elution (Koike, M. et al., 1976).
Results showed that pyruvate dehydrogenase is inhibited by S-acetyl-CoA, and this
inhibition is directed to the first reaction enzyme, a thiamine-requiring enzyme. Fur-
ther regulating this first enzyme are nucleotide concentrations. Thus AMP, ADP,
and ATP act to increase the rate of activity of the enzymes whether S-acetyl-CoA is
present or not. While ATP in this preparation was stimulatory, in vivo it is probably
inhibitory. It is important to consider only real physiological variables in vivo in
attempting to assess regulation of pyruvate dehydrogenase.
Since the brain is not a gluconeogenic organ, the regulation of the thiamine-
dependent enzyme pyruvate carboxylase by rat brain mitochondria has been inves-
tigated (Patel, M. and Tilghman, S., 1973). Rat brain mitochondria were isolated
after decapitation and viability assured by measuring the P:O ratio and respira-
tory quotient. When all cofactors were present, the mitochondria fixed H14 Co3 at
a rate of 269 nmoles/30 min/mg protein. If pyruvate, MgCl2 , or ATP were left
out of the mixture, fixation was nearly zero. Omission of Pi reduced decarboxy-
lation by 65%. The combination of adenonucleotide and Pi data tend to show that
the decarboxylation is regulated by intramitochondrial adenine nucleotide concen-
tration. Except for citrate and isocitrate, the other dicarboxylate intermediates of
the TCA cycle tended to decrease the decarboxylation of pyruvate as much as
70%.
The concentration of pyruvate also played a role in the regulation of mitochon-
drial decarboxylation. The role in regulation for the cofactor acetyl-CoA was not
examined, but speculation is that it plays a role. The acetyl moiety of acetyl-CoA
is vital for lipogenesis as well as the formation of acetylcholine. These are cytoso-
lic processes, and the acetyl-CoA does not cross mitochondrial membranes. Data
tend to show that citrate may act as a transporter to carry the acetyl moiety to the
cytoplasm.
In another study (Tomita, I. et al., 1974), analogues of thiamine pyrophosphate
were examined for their effects on pyruvate decarboxylase and transketolase, both
thiamine-requiring enzymes. Results showed that both pyruvate decarboxylase and
transketolase reconstitution was inhibited by thiazole pyrophosphate as well as by
methyl and benzyl thiazole pyrophosphate.
The phosphorylated esters of 2-methyl and 2-benzyl thiazoles were inactive in
combination with apopyruvate decarboxylase, or with apotransketolase. It appears
to be the pyrimidine moiety of thiamine pyrophosphate which is important in
binding studies, causing hydrogen bonding to take place. It is speculated that the
pyrimidine moiety could act to tighten the binding of coenzyme by negating ionic
repulsion.
Another study (Taylor, S. et al., 1975) describes the effects of the ATP/ADP
ratios and other variables on the activity of pyruvate dehydrogenase. A rapid freez-
ing method was developed to minimize any possible artifact change between active
and inactive pyruvate dehydrogenase.
Results showed that the rapid freezing method was successful in extracting pyru-
vate dehydrogenase with little interconversion between the active and inactive forms
of the enzyme. The investigators found a good correlation between the active form
of pyruvate dehydrogenase and the intramitochondrial ATP/ADP ratio. Also noted
was that pyruvate increased the active form of pyruvate dehydrogenase without any
alteration in the ATP/ADP ratio. Octanoate acted to lower the active form of the
enzyme, also without any significant change in the ATP/ADP ratio.
These data, taken together, lend strong support for the concept that the ATP/ADP
ratio is an essential component for pyruvate dehydrogenase activity. This observa-
tion might be best explained by the fact that ADP is a competitive inhibitor of active
pyruvate dehydrogenase kinase. The authors speculate that the ATP/ADP ratio in
Early Chemistry 7
mitochondria may also be a key regulator of the active form of the enzyme in intact
cells.
The developmental pattern of pyruvate carboxylase has been studied in newborn
rats because of its importance in gluconeogenesis (Chang, L., 1977). In these stud-
ies, animals from –5 days (premature) to 36 days after birth were utilized. Results
showed that about 70% of pyruvate carboxylase activity was located in the mito-
chondrial fractions of liver. Twenty percent was found in the nuclear fraction. Only
about 5% of adult levels of pyruvate carboxylase were found in fetal liver. The lev-
els of enzyme remained low until birth, when activity levels rapidly rose, reaching
a peak at about 5 days after birth. The actual rise after birth occurred after the first
8 hours, suggesting that the birth process in and of itself was not significant.
The developmental pattern of pyruvate carboxylase in vitro agrees with that seen
previously in vivo (Snell, K., 1974). The author speculates that the rise in activity
of the enzyme could be related to liver maturation. More likely is that the rise after
birth is due to a sudden change from a high-carbohydrate diet (transplacental) to a
diet of milk high in fat (nursing). This would spell a clear adaptation to changing
dietary balance.
Early Thiamine Deficiency
Early thiamine deficiency studies were performed on pigeons because these birds
were uniquely sensitive to polished rice. Symptoms frequently were evident by 21
days, which was as early as or earlier than any other animal model. This sensitivity
also took the form of pronounced symptoms and certain death when not reversed
with thiamine. The reversal feature was key in determining the significance of bio-
chemical changes. If symptoms reversed with thiamine treatment along with key
substrates such as decreased lactate, then this implied a role for lactate in the gener-
ation of the symptoms (Fig. 1).
Fig. 1 Pigeon showing opisthotonic posturing due to thiamine deficiency. Reproduced from
Yoshinori (1995) with permission from Springer
Several studies from about 1910 to 1916 were pivotal in developing concepts
relating to vitamin physiology and biochemistry. For example, it was discovered
that certain chemicals were necessary for continued health, and these were termed
“vital amines,” from which the word vitamin was coined. It was shown that adding
carbohydrate to a vitamin-free diet brought about the development of symptoms
more rapidly. This suggested that the carbohydrates require vitamins, and the addi-
tion of carbohydrates increases vitamin demand and consumption.


Casimir Funk was instrumental in some of these early studies. In one report
(Funk, C., 1911), he placed pigeons on several diets and the results were com-
pared. The data showed that all pigeons placed on diets with high concentrations
of carbohydrates developed hyperglycemia. Pigeons maintained on a vitamin-free
diet also developed hyperglycemia. When hyperglycemic animals were treated with
vitamins, the hyperglycemia diminished and liver glycogen levels increased. These
studies show early correlations between carbohydrate metabolism and vitamins.
By the mid-1920 s several studies had been published, sometimes with conflict-
ing data. Most investigators, however, were in agreement that animals deprived
of vitamin B had significantly decreased oxygen consumption in various tissues.
This was largely attributed to a reduction in the amount of activity of oxidizing
enzymes. However, even in related measurements such as the respiratory quotient,
there was disagreement. Hess and Messerle (1922) suggested that there was a sim-
ilarity between the polished rice-fed beriberi pigeons and the result of sublethal
doses of cyanide. He also speculated that the biochemical results in animals eating
polished rice were manifest in a decrease in the amount or efficiency of oxidizing
enzymes.
Given the somewhat controversial situation regarding various hypotheses that
had been put forth to explain the relation of these findings to vitamin B metabolism,
a rigorous study was published (Drummond, J. and Marrian, G., 1927). In one set
of experiments, the authors compared the rates of reduction of methylene blue as
a hydrogen acceptor to assess the oxidation/reduction status. Normal controls and
beriberi tissues (pigeon breast muscle) were compared. Results did not confirm stud-
ies by Hess (Hess and Messerle, 1922). The Hess results showed that it took longer
for the vitamin-deficient (beriberi) tissue to produce a reduction of methylene blue
than in the case of normal non-deficient tissue. In the study by Drummond and Mar-
rian, there was a small increase in time for the reduction of methylene blue, but it
was not deemed significant.
Similarly when these investigators examined oxygen consumption in normal and
beriberi tissues, results showed no real difference in breast muscle and only a small
decrease in liver. These workers also documented the weight loss seen in the final
days on vitamin B deficient diet. The authors conclude that on the basis of their
findings, there is no evidence to support the view that vitamin B is essential for the
functioning of the oxidative mechanism in tissues. The authors do point out that they
cannot rule out the possibility that some small stimulation in oxidation might occur
in intracellular locations. This could explain the small rise in temperature seen when
animals deficient in vitamin B are administered extracts containing the vitamin.
In another study (Kinnersley, H. and Peters, R., 1930) published 2 years later,
the relation between lactic acid in the brains of avitaminosis pigeons and normal
control birds was described. Pigeons were stunned and the head was cut off and
quickly frozen in liquid air. Frozen brain samples were extracted in 10% TCA, and
following suitable preparation, lactic acid in the brain and in the blood was mea-
sured in normal pigeons and in opisthotonic pigeons. The vitamin-deficient pigeons
were made opisthotonic by feeding them with polished rice. Results showed a much
higher concentration of lactic acid in the brains of opisthotonic pigeons than in that
of normal control birds.
Early Thiamine Deficiency 11
Worried that some artifact of lactic acid might have occurred due to the few sec-
onds delay before the brain was frozen, the authors constructed a special guillotine.
This devise was described as “T” shaped with two blades at different levels. The
lower blade decapitated the pigeon, while the second, higher blade split the skull.
This effectively severs the blood supply to the brain, and the splitting of the skull
facilitates more rapid freezing.
Results using the improved guillotine showed elevated lactic acid, as was seen
earlier. The increased lactic acid was noted during the opisthotonic stage, not earlier,
and the elevated lactic acid rapidly returned to normal when the animals were treated
(reversed). Results also showed that stunning produced higher levels of lactic acid
than did anesthesia with ether. The authors conclude that lactic acid levels seen in
brain are more than those that could result from blood levels, thereby suggesting
storage of carbohydrate in brain. The rapid reversal with vitamin B1 indicates no
other food material or ions could have been deficient enough to produce the severe
symptoms seen in these pigeons.
It is amazing to see that in 1925 investigators (Kinnersley and Peters) knew about
the lability of brain metabolites. They designed and built a two-bladed guillotine to
first cut off the head and then split the skull, rendering easier access to the brain
for freezing. This worked to prevent or minimize artifacts after death. Interestingly,
in the late 1960 s into the early 1970 s there was intense concern about the very
real risk of changes occurring in brains of experimental animals following killing.
In fact, devices were designed to minimize changes. These included freeze blowers,
microwave ovens, and brain choppers which “fixed” or froze tissue in milliseconds,
rendering the measurement of “true” levels of labile metabolites. One wonders if
those investigators knew of Peters’ attempts along the same lines, which were done
50 years earlier.
In another study by Peters (Gavrilescu, N. and Peters, R., 1931), oxygen uptake
was determined in pigeon brain regions during vitamin B deficiency induced symp-
toms. In this study, oxygen uptake was performed rapidly after killing, the elapsed
time being 12–17 minutes. The guillotine developed and described above was used.
Brain regions were isolated and measured separately. Regions sampled included
cerebellum, cerebrum, optic lobes, and “lower parts.” The speed of tissue prepa-
ration was attributed in part to a new air-damped Sartorius balance. The Homer
pigeons were fed in the standard way for producing severe symptoms.
Results showed the usual symptomology of opisthotonus, and death if not treated.
Oxygen uptake was determined in these birds with head retraction or just before
these symptoms were apparent. Oxygen uptake in symptomatic birds’ cerebellum
was comparable to normal birds, but values in the cerebrum, optic lobes, and lower
parts were decreased in symptomatic as compared to normal birds. The most signif-
icant decreased areas were the cerebrum and optic lobes. It was further shown that
the oxygen uptake reverted to normal when the animals were “cured.”
The authors state that these oxygen uptake data, together with earlier published
lactate data, show conclusively that the disorder in vitamin-deficient pigeons repre-
sents a biochemical defect that is central in origin, not peripheral. The rapid reversal
of opisthotonus and decreased oxygen uptake with vitamin B treatment prove that
these features are due to vitamin B1 deficiency, not starvation. The fact that the cere-
bellum was unaltered as regards oxygen uptake serves as an internal control, arguing
against changes in blood circulation in the affected pigeon brain. The authors state
that they do not think some circulating toxin is involved.
It is interesting to note that even today (2009) many neurochemical studies are
performed on whole brain, neglecting the brain’s heterogeneity, whereas 80 years
ago regional brain studies were being performed, and regional differences in neuro-
chemical responses to perturbations were being described.
In a later study (Gavrilescu, N. and Peters, R., 1931) the issue of the possible
local lack of vitamin B1 was addressed. Arguing in favor of a local change are data
showing regional alterations in biochemical paradigms such as oxygen uptake, and
the rapid (30 minutes) reversal of symptoms by injection of thiamine into the cra-
nium. The counterargument was that some hormone or some other substance might
be stimulated and/or released into the circulation by vitamin B1, thereby causing
the reversal of biochemical changes and symptoms. The authors chose to examine
oxygen uptake in vitro when vitamin was added to a brain preparation from symp-
tomatic pigeons.
Animals used were pigeons in the final stages of vitamin B1 deficiency and show-
ing the symptom of opisthotonus. Brain areas used were optic lobes, cerebrum, and
lower brain parts. Tissue was collected as before, minced, and oxygen uptake mea-
sured as previously described. After equilibration, vitamin B1 was added to the mix-
ture. Results showed that when B1 was added to optic lobe and to the lower parts,
there was a clear increase in oxygen uptake by the homogenate. Various methods of
inactivation of vitamin B1 were tried, and only when the vitamin was completely
inactivated did the addition fail to increase oxygen uptake.
Earlier results indicated the cerebrum acted differently, and in these experiments
the effect of vitamin addition had little effect on oxygen consumption. The authors
conclude that whatever decrease in oxygen consumption is present in the cerebrum,
it is not due to vitamin B1 decrease, as it is not reliably reversed by the addition of
vitamin B1.
Adding vitamin B1 to a regional brain preparation seemed to rule out any stimu-
lation in vivo from some other agent or hormone in vitamin-deficient pigeons. Sim-
ilarly, these data correlate well with the lactate data gathered earlier. Collectively,
the data on lactate and oxygen consumption indicate a local (regional) biochemical
lesion, which correlates with an abnormal condition in the CNS. The authors con-
clude that vitamin B1 is directly involved with oxidative metabolism in lower brain
parts in symptomatic pigeons.
In a later paper (Rydin, H., 1935), pigeons that were not in opisthotonus but
rather were ataxic due to vitamin B1 deficiency were studied. The goal was to see if
there was a change in oxygen uptake in regional brain parts when other symptoms
were present.
As before, pigeons were made symptomatic by feeding them polished rice.
Birds were on this diet for at least 30 days and were selected for the study based
on the absence of opisthotonus and presence of leg weakness and ataxia. Ani-
mals were killed by decapitation and brains quickly removed and minced with a
spatula. Oxygen consumption was then measured in the cerebrum and in the lower
parts of the symptomatic pigeons’ brains.
Results showed that pigeons exhibiting leg weakness had a greater increase in
oxygen uptake due to added vitamin B1 in the cerebrum. There was a negligible
effect in the optic lobes and the rest of the brain. In previous work the optic lobes
and the rest of the brain were affected during the opisthotonus phase of vitamin B1
deficiency.
In this study, as seen earlier, the decrease in oxygen uptake and symptoms cor-
related with vitamin B1 deficiency, not inanition. This study shows this even in the
absence of opisthotonus, but with leg weakness, a decreased oxygen uptake was
reversed by vitamin addition in vitro, again arguing against the stimulation of some
other hormone or chemical release. Finally this study demonstrates that a different
brain region – the cerebellum – appears to respond to leg weakness. This shows
that the heterogeneity of the brain accounts for different responses, which correlate
with different neurological symptoms. This study points out again that the leg weak-
ness was central in origin and not manifest due to altered muscle biochemistry or
peripheral nerve involvement.
In a subsequent paper the oxygen/pyruvate ratio and the pyruvate RQ in nor-
mal and vitamin-deficient pigeon brain (McGowan, 1937) were determined. The
methodology was as previously described from the Peters’ laboratory.
Results showed that pyruvate was incompletely oxidized, and it was likely that
some intermediate was accumulating. Also it seemed that some pyruvate was com-
pletely oxidized, while some was metabolized along some other route. Data from
another worker showed that some pyruvate was metabolized to lactic acid and CO2 .
Assuming that two-thirds of pyruvate is completely oxidized, and one-third goes
into the Krebs cycle, then the theoretical oxygen/pyruvate ratio is around 450 and
the RQ is 1.29, and these are about the actual values found.
The experiments described above were carried out in pigeons suffering from
vitamin B1 deficiency due to eating polished rice. Results showed that the oxy-
gen/pyruvate ratio was significantly lower than that of normal pigeons. The author
suggests that there may be two different routes for pyruvate, and absence of vitamin
B1 may influence which route is utilized in vitamin B1 deficient birds.
In a neuropathological study, pigeons were made thiamine deficient by dietary
means, then brains and peripheral nerves were examined in animals exhibiting
opisthotonus, leg weakness, and in controls without symptoms (Swank, R., 1940).
Experimental groups consisted of the following: (1) starved and thiamine deficient,
(2) acutely thiamine deficient, (3) chronically thiamine deficient, and (4) starved on
normal diet. Acute thiamine deficiency was produced by tube feeding a deficient
diet until vomiting occurred. Then thiamine was administered along with the diet
for 3–6 weeks. After that, symptoms (opisthotonus) occurred in 7–12 days if the thi-
amine supplementation was stopped. In the chronic group, after vomiting occurred,
a small dose (7–15 µg) of thiamine was administered. In these birds, ataxia and leg
weakness occurred between the 16th and the 30th day.
Results in the acutely thiamine-deficient group showed occasional degeneration
in the sciatic nerve only. Lesions were also seen in the ventral funiculus of the
cervical spinal cord. The lesions were those of degenerating myelin sheaths. It was
thought that there were more affected nerves, but that the chlorate-osmic acid stain-
ing method would not show them.
In the chronically thiamine-deficient group, pigeons developed leg weakness
and ataxia, not opisthotonus. In addition most of these pigeons showed evidence
of heart failure. Results showed that the greater and longer lasting symptoms cor-
related with more pronounced evidence of demyelination in peripheral nerves. In
mildly symptomatic birds, the pathology was distal, and as the intensity of symp-
toms grew worse, the lesion moved centrally. The process of degeneration was Wal-
lerian in appearance. Centrally, lesions start in the lateral funiculus and follow the
spinocerebellar tract and can be traced into the cerebellum. This “ascent” was simi-
lar to that seen peripherally where lesions first appeared distally and moved centrally
in increasingly severely symptomatic birds.
In birds “reversed” with thiamine treatment, the total number of myelinated nerve
fibers in the sciatic and brachial nerves increased as the birds recovered. Complete
recovery took 6–9 weeks in birds in which leg weakness had lasted for 7–10 days
before instituting treatment. This is in contrast to the rapid recovery (hours) in
rats reversed with thiamine therapy. Microscopically, dorsal ganglia continued to
show chromatolysis until symptoms cleared. It seemed that complete regeneration
of nerve fibers had to occur before symptoms disappeared.
The author suggests that the lesion associated with opisthotonus is biochemical
in nature, and that explains the rapid reversal of symptoms when the symptom is
opisthotonus. By contrast, the leg weakness appears to be a lower motor neuron
type of paralysis, and nerve fibers in the sciatic nerve degenerate. Adding thiamine
to the diet reverses the lesion and symptoms, but the symptom reversal is dependent
upon reversal of the nerve lesion. Hence, this reversal time is lengthy. The large
myelinated peripheral nerve fibers degenerated first in thiamine-deficient birds. In
the CNS, the long ascending spinocerebellar fibers were affected early in the pro-
cess. When an axis cylinder has been damaged for longer than 4 days, changes in
the cell bodies consisting of chromatolysis occur. The chromatolysis remains after
initiation of the thiamine reversal period and finally disappears after the myelin
cylinder shows complete repair. In the CNS, damaged neurons did not regenerate.
It is suggested that in early thiamine deficiency, CNS neurons are at first damaged
only chemically. If thiamine is administered during this early period cell recovery is
about 100%. Later, structural changes occur in the cell body, which disappear only
if the axis cylinder makes a full recovery.
Opisthotonus is a manifestation of decerebration. Leg weakness, a more chronic
disorder, is due to degeneration of nerves such as the sciatic nerve. Reversal is pos-
sible given thiamine treatment and time.
The blood lactate/pyruvate relation in thiamine-deficient pigeons was investi-
gated (Stoltz, E. and Bessey, O., 1942). Pigeons were fed the usual polished rice diet
and blood was collected from a wing vein. Birds were on a thiamine-deficient diet
for 9 days. At this point, blood sampling revealed increases in pyruvate, at a time
when the pigeons were still asymptomatic. The birds were then allowed 5 more
days of thiamine-deficient diet, at which time some pigeons were exhibiting mild
opisthotonus. This was the time at which the pyruvate values were greatest. Treat-
ment of symptomatic pigeons with thiamine reversed these elevated pyruvate levels.
This study demonstrated the usefulness of blood pyruvate/lactate measurements as
a way to evaluate thiamine deficiency even before the onset of symptoms.
In an early study in rats (Harper, H., 1942), the rate of absorption of glucose
from the intestine was examined. Rats were placed on a thiamine-deficient diet and
maintained in a way that allowed a “subacute” thiamine-deficient state to exist. At
the time of killing, glucose was administered, and after 1–2 hours, the entire diges-
tive tract was removed and analyzed for glucose. Results showed that there was a
significant decrease (17%) in glucose absorption. This decrease in glucose absorp-
tion was not surprising in light of the decreased function in the GI tract. In another
series of experiments, the author found decreased glycogen in the liver of thiamine-
deficient rats 3–6 hours after feeding glucose. This finding could be attributed to
lower absorption of glucose.
The effect of oxygen and thiamine pyrophosphate (cocarboxylase) on the forma-
tion of citrate and alpha-ketoglutarate has been studied (Coxon, R. and Peters, R.,
1950). In this study, pigeons were used and treated by administering polished rice as
the diet until opisthotonus developed. Animals were killed and brain homogenates
were prepared. Both citrate and alpha-ketoglutarate were measured. Results showed
that the addition of cocarboxylase and oxygen to pigeon brain extracts promoted
the generation of citrate and alpha-ketoglutarate. In pigeon brain preparation from
thiamine-deficient animals, pyruvate accumulates. Previously it had been suggested
that some seven-carbon compound was the first compound formed in the decar-
boxylation of pyruvate. This seems not to be the case given that it is pyruvate which
accumulates when citrate and alpha-ketoglutarate do not form in thiamine-deficient
pigeon brain. Thus it is shown that citrate and alpha-ketoglutarate are formed in the
presence of oxygen and cocarboxylase. Further, it is concluded that a two-carbon
compound is formed when pyruvate is decarboxylated.
Based on earlier work showing that thiamine deficiency interferes with the degra-
dation of pyruvate, the author of this paper (Liang, C., 1962) examined pyruvic acid
changes in thiamine deficiency.
Albino rats were made thiamine deficient by feeding a diet of polished rice pow-
der. Pyruvic acid and total keto acids were measured. Results showed that rats on a
thiamine-deficient diet died after 10 weeks, and they demonstrated weight loss and
ataxia.
Pyruvate and keto acids in brain, heart, liver, and muscle showed a steady rise
after administration of the deficient diet. Glyoxylic acid was detected in all tissues
tested, including brain.
A slight decrease was seen in the rise in pyruvic acid and keto acids between days
15 and 22, and the reason is unclear. The reason for the elevation in brain of gly-
oxylic acid also remains unclear, but may be related to the deamination of glycine
and the amination of alpha-ketoglutarate to glutamate and glutamine. Glyoxylic acid
is toxic. These data are in keeping with previous reports showing impairment in
enzyme activity of pyruvate decarboxylase and of alpha-ketoglutarate decarboxy-
lase in the brains of thiamine-deficient rats.
In a paper by Lofland et al. (1963), two different strains of pigeons were stud-
ied as to the effect of thiamine deficiency on thiamine-dependent enzymes. Results
showed that the “snow racer” pigeon was more susceptible to thiamine deficiency
than the “white carneau” breed. This was viewed as an indication of a possible
genetic control of thiamine-dependent enzymes. The enzyme alpha-ketoglutarate
decarboxylase seemed to show the most decreased activity among the three enzymes
studied. Results also showed that transketolase decreased in blood, but remained
about normal in brain. The activity of pyruvate oxidase was decreased, which
seemed to correlate with the elevation in pyruvic acid. Overall these data are in
keeping with previous data showing an alteration in the thiamine-requiring enzymes
in carbohydrate metabolism (for further review, see Itokawa, 1995).
Thiamine Deficiency in Mammals
The use of animal models of thiamine deficiency has proved quite beneficial in the
elucidation of underlying mechanisms of pathogenesis. Brain lesions, time course,
symptoms, and reversibility of symptoms all correlate very well with features seen
in humans. The issue of mode of action of thiamine antagonists has been investi-
gated in thiamine-deficient rats (Gubler, 1961). Although several antagonists have
been discovered such as oxythiamine, pyrithiamine, amprolium, and deaminoth-
iamine most studies have used either oxythimine or pyrithiamine. The Gubler study
examined the effects of a pure dietary thiamine-deficient diet, a thiamine-deficient
diet plus oxythiamine, and a thiamine-deficient diet plus pyrithiamine on the oxida-
tion of rat brain pyruvate and alpha-ketoglutarate.
Thiamine-deficient rats on this regime showed the usual growth curve (see Fig. 1,
for example). After sacrifice, brain homogenates were prepared and the oxida-
tion of pyruvate and alpha-ketoglutarate measured in brain homogenates. Results
from these biochemical analyses showed that the oxidation of neither pyruvate nor
Fig. 1


alpha-ketoglutarate was decreased by oxythiamine, but both were decreased with

thiamine deprivation and pyrithiamine treatment. These decreases could be reversed
by the addition of thiamine diphosphate in vitro.
Heart enzymes were also measured, and pyruvate oxidation was decreased in all
three types of thiamine deficiency, but only oxythiamine plus thiamine-deficient diet
had a lowering effect on alpha-ketoglutarate oxidation. These data indicated that at
least two enzymes are involved in the oxidation of pyruvate and alpha-ketoglutarate.
In a separate study (Brin, 1962), the effects of thiamine deficiency plus oxythi-
amine was examined on rat brain transketolase. This study showed that in all tissues
examined – kidney, intestine, spleen, heart, liver, etc. – transketolase levels were
significantly lower. The one exception was brain. The decreases were evident in
tissue before the leveling off of the growth curve. The level of decrease in transke-
tolase with oxythiamine administration was noteworthy in that, at that time period
(13 days), neurological symptoms were absent.
In another study (Dreyfus, 1965), transketolase was measured in brain regions
of thiamine-deficient rats maintained on diet for 30+ days. Brain regions assayed
included myelinated areas such as spinal cord and pons and less myelinated areas
such as cerebral cortex and thalamus. High transketolase levels were found in spinal
cord and lowest levels in the cerebral cortex. The greatest effect of thiamine defi-
ciency was noted in the lateral pontine tegmentum, the site of the most striking
neuropathological changes. These data suggested that transketolase activity might
be involved in oligodendroglial metabolism.
A problem of older studies using Warburg techniques to measure pyruvate dehy-
drogenase activity is a lack of sensitivity. To overcome this, a study was performed
comparing four different measures of pyruvate dehydrogenase activity: (1) War-
burg manometry, (2) the Clark electrode method, (3)14CO2 evolution, and (4) fer-
ricyanide reduction (Bennet et al., 1966). Tissue analyzed was liver and brain from
pyrithiamine-induced thiamine-deficient rats. Animals were maintained on this diet
until symptomatic and thenkilled. Results showed that all four procedures for esti-
mating pyruvate dehydrogenase activity gave similar results. The authors found no
effect of thiamine deficiency on liver enzyme activity. Brain pyruvate dehydroge-
nase activity was reduced about 70% compared to control rats. Cerebral 2-oxo-
glutarate oxidation was reduced by 30%, whereas succinate and malate oxidation
was not changed (Fig. 2).
In a study comparing the oxidation of pyruvate and the oxidation of ribose
in intact rats (Brin, 1967), 14C-pyruvic acid and 14C-ribose were injected into
thiamine-deficient rats, and 14CO2 evolution measured. Results showed that
14C-pyruvate was oxidized to 14CO2 in deficient rats at the same rate as controls.
14C-ribose, however, showed a reduced oxidation at only 1 week of deficiency and
continued to decrease throughout the time course. The defect showed improvement
within 24 hours following thiamine treatment. Transketolase activity was measured
in six different tissues in thiamine-deficient rats and found decreased in kidney,
muscle, and liver, but not brain.
The use of the thiamine antagonists oxythiamine and pyrithiamine became
increasingly widespread, and the mode of action of these antagonists was examined
Thiamine Deficiency in Mammals 19
Fig. 2 Schematic representation of glycolysis and the TCA cycle. Modified from Desjardins, P.
and Butterworth, R. (2005), and McCandless, D. and Schenker, S. (1968)
(Steyn-Parve, 1967). When antimetabolites are administered, many symptoms noted

in pure dietary thiamine deficiency are present. These effects can be reversed by
increasing thiamine doses. The effects on mice and rats of the two antagonists
are different. Oxythiamine seems to alter heart function, whereas pyrithiamine
inhibits thiamine phosphate synthesis and has its major effect on the nervous sys-
tem. Oxythiamine acts as a competitor of thiamine pyrophosphate. Oxythiamine
produces bradycardia, elevates blood pyruvate significantly, and cannot enter brain
tissue. Pyrithiamine is capable of entering brain tissue, thereby producing CNS
symptoms. These studies tend to dissociate the heart and CNS symptoms in thiamine
deficiency by showing that the two clinical characteristics can exist independent of
each other.
Additional studies (for review, see Cooper and Pincus, 1967; and Karup-
pagounder, S. and Gibson, G., 2009) were directed toward determining if thi-
amine (or thiamine triphosphate) might play a role in nerve conduction. Initial
studies had shown that pyrithiamine applied to nerve fibers in vitro blocked the
action potential, whereas oxythiamine did not. Another key observation was that
in vitro frog spinal cord released thiamine when stimulated. Other studies showed
that vagus nerve preparations to which pyrithiamine was administered showed an
increase in amplitude of the compound action potential and reversal of post-tetanic

hyperpolarization. These changes could be reversed by thiamine administration, and
oxythiamine did not have an effect on vagus nerve electrical activity. Results on
experiments to analyze the effects of pyrithiamine on vagus nerve preparations were
equivocal.
To determine if thiamine might be involved in nerve conduction, a further look at
the mode of action of pyrithiamine was undertaken. In one set of experiments, vagus
nerves were incubated for 10 minutes both with and without pyrithiamine. Nerves
incubated with pyrithiamine had about a 40% decrease in thiamine as compared to
those not incubated with pyrithiamine.
In a second series of experiments, rats were made deficient by dietary means.
When symptomatic, labeled thiamine was injected over 3 days. Upon killing, spinal
cords were removed, divided longitudinally, and placed on electrodes in a Locke’s
bath. The spinal cords were stimulated in the bath both with and without pyrithi-
amine. Results showed that both pyrithiamine and electrical stimulation released
labeled thiamine from the spinal cord preparations. This correlates with earlier
pyrithiamine studies showing an increased urinary output of thiamine during pyrithi-
amine administration. These data, overall, suggested a possible nerve conduction
role for thiamine, independent of its role as co-enzyme in key enzymatic reactions.
Direct evidence of participation of thiamine or thiamine triphosphate in nerve ion
transport is lacking.
The possibility that there was a relationship between thiamine deficiency and
magnesium deficiency was suspected on clinical grounds. Therefore, studies were
undertaken to examine possible correlations (Zieve et al., 1968). Results showed that
in magnesium-deficient rats, blood and liver thiamine levels were slightly reduced,
and in thiamine-deficient rats, there was a transient decrease in magnesium in liver.
The activity of thiamine-requiring enzymes was significantly decreased in the liver
of magnesium-deficient rats. The recovery of enzyme activity was incomplete upon
thiamine administration. Interestingly, magnesium in vitro did not reverse decreased
enzyme activity.
The explanation that seemed most logical to the authors was that magnesium
deficiency interfered with activation of magnesium-dependent enzymes. The effect
on transketolase, rate limiting reaction in the hexose monophosphate shunt, seemed
most sensitive. The authors suggest a defective incorporation of magnesium into the
enzyme complex. These data are somewhat in keeping with earlier studies (Brin,
1967) suggesting the transketolase alteration in thiamine deficiency is more critical
than that of the decarboxylation of pyruvate or oxyglutarate.
The authors speculate that a state of magnesium deficiency, appearing clinically
similar to thiamine deficiency, could explain why thiamine alone does not amelio-
rate symptoms. The authors, however, found it strange that Mg replacement ther-
apy did not result in a reversal of deficiency symptoms. This was interpreted as
further evidence of an effect on transketolase and that this effect carries over to
the recovery phase. Both animals and humans could be compromised by magne-
sium deficiency because it would result in a reduced ability to properly utilize
thiamine.
The effect of thiamine deficiency on rat brain acetylcholine has been examined
(Cheney et al., 1969). Thiamine-deficient rats treated with the antagonist pyrithi-
amine developed neurological symptoms and seizures. There was a lack of corre-
lation between these symptoms and either total acetylcholine activity or between
symptoms and the rate of acetylcholine formation in the deficient rat brains. On the
other hand, prevention of acetylcholine destruction by injection of eserine prior to
convulsions seemed to prolong death. This suggests that perhaps acetylcholine does
play some (minor) role in the evolution of neurological signs.
Most of the above studies were carried out in whole brain, but neuropathological
studies in thiamine deficiency indicate a highly focal lesion in the brainstem and
cerebellum, as is seen in Wernicke’s disease. When analyzing whole brain, inclusion
in the sample of non-affected tissue could easily dilute the sample, thereby masking
adjacent affected brain areas.
In a study aimed at looking at biochemical parameters in brain regions (McCan-
dless and Schenker, 1968), thiamine deficiency was produced by dietary methods
only. This regime produces neurological symptoms in 4.5–5 weeks. Since loss of
appetite and decreased food ingestion is a feature of dietary thiamine deficiency,
pair-fed controls were utilized in order to have weight loss controls. Animals were
killed in a dry ice acetone mixture (–120ºC) in order to preserve labile brain metabo-
lites.
By 4.5 weeks, thiamine-deficient rats began to show neurological symptoms.
These consisted of ataxia, impaired righting response, opisthotonic posturing, and
drowsiness. Seizures were not a feature of this dietary regime. When the deficient
rats were injected with 10 µg of thiamine, symptom reversal occurred within 16–
36 hours. The ability of the rats to be “reversed” allows for the brain biochemical
changes to be re-examined following symptom reversal.
Cerebral thiamine levels dropped to about 15% of control values when symp-
toms were most severe. Thiamine dropped to 20% of control before any symptoms
occurred, indicating a significant brain reserve of the vitamin. There was no differ-
ence between ad lib fed controls, and pair-fed controls.
The thiamine dependent enzyme pyruvate decarboxylase was decreased by 36%
and 29%, respectively, in brainstem and cerebellum of symptomatic rats. Reversal
of symptoms with B1also reduced the enzyme depletion. The thiamine-requiring
enzyme transketolase was also measured, and the most significant decrease was
found in thiamine-deficient cerebral cortex, an area not affected in thiamine defi-
ciency. The enzyme decrease was least in the brainstem of deficient rats.
Lactate levels were elevated in the cerebellum and brainstem of thiamine-
deficient rats. They were unchanged in the cerebral cortex. Brainstem pyruvate was
also elevated in symptomatic animals. Reversal of symptoms with thiamine adminis-
tration reduced the decreases in these metabolites. Analysis of ATP levels in cortex,
cerebellum, and brainstem showed no alteration in ATP in moderately or severely
thiamine-deficient rats.
This study defined the critical range of thiamine depletion necessary to result
in neurological symptoms, and the amount of increase related to the reversal of
symptoms with thiamine treatment. Transketolase was decreased to a greater extent
than pyruvate decarboxylase, duplicating previous studies. Regional results showed

a greater decrease in brainstem/cerebellum than in cerebral cortex. These findings
are consistent with the neuropathologic observations of lesions largely localized to
the cerebellum and brainstem.
The rapid reversal (16–36 hours) of symptoms with thiamine administration per-
mits an opportunity to re-measure enzyme levels to determine if one or the other also
reverts toward normal. In spite of the greater effect of thiamine deficiency on trans-
ketolase levels, it was the pyruvate decarboxylase enzyme that had a more dramatic
increase toward normal. This would suggest that reversal of the pyruvate decarboxy-
lase enzyme activity was more closely tied to the reversal of symptoms.
Since ATP production is linked to pyruvate decarboxylase and TCA cycle func-
tion, the finding of normal ATP levels suggests that some other mechanism linked to
the enzyme level should be examined. This does not mean that some smaller region
might show changes in high-energy phosphates or that there could be a decrease in
production of ATP coupled with a decrease in utilization leading to unchanged net
levels of high-energy phosphates. Remember, the lesions of thiamine deficiency are
highly localized.
Utilization of the Lowry “closed system” (Lowry et al., 1964) assesses turnover
of adenine nucleotides and has been used to examine flux in thiamine-deficient rat
brain (McCandless and Cassidy, 1976). In this study, after decapitation, the brains
are maintained at 37ºC for 30 seconds and then frozen in liquid N2 . This effectively
stops high-energy phosphate synthesis, and only utilization occurs (Lowry et al.,
1964). Results from this study showed that metabolism of the adenine nucleotides
ATP, ADP, and AMP were not altered in severely thiamine-deficient rats. Pyruvate
decarboxylase activity was also measured and found to be 63% decreased in the
brainstem of thiamine-deficient rats as compared to pair-fed controls. This indicates
that there is a significant reserve capacity for pyruvate decarboxylase.
The above studies, although performed regionally, still included many cell
types, which may not be involved in severe thiamine deficiency. To circumvent
this problem, cerebral metabolism in thiamine-deficient rats has been examined in
small (100–500 ng) samples from the lateral vestibular nucleus (McCandless and
Schwartzenburg, 1981).
The results from this study showed that phosphocreatine was elevated in the lat-
eral vestibular nucleus in severely thiamine-deficient rats as compared to pair-fed
controls. Reversal of symptoms resulted in a return of phosphocreatine to con-
trol levels. An increase in phosphocreatine could be viewed as an indication of a
decrease of energy utilization rate in very discrete samples. ATP was unaffected in
this study. Glucose and glycogen levels increased in the recovery stage, a finding
noted in the reversal of other models of metabolic encephalopathy (McCandless
et al., 1979).
It has been shown previously that transketolase is depleted in thiamine-deficient
brain to a greater extent than pyruvate decarboxylase. This observation has led many
to speculate that transketolase and the hexose monophosphate shunt play a vital role
in thiamine deficiency. This hypothesis has been advanced despite the fact that upon
reversal of symptoms with thiamine administration, transketolase levels returned
toward normal only slightly, whereas pyruvate decarboxylase returns much more
dramatically. Interestingly, thiamine triphosphate (with a supposed neurotransmis-
sion role) returns not at all (Pincus and Grove, 1970). Since enzymes are not a
true indicator of flux through a metabolic pathway, the effect of thiamine defi-
ciency directly on the hexose monophosphate shunt (pentose phosphate pathway)
was examined.
In these studies (McCandless et al., 1976a), rats were rendered thiamine defi-
cient to the point of moderate to severe neurological signs. Flux through the hex-
ose monophosphate shunt was estimated based on the differentially labeled glucose
method. The C1/C6 ratio derived from the specific yields of 14CO2 from 1 to 14C
and from 6 to 14C glucose shows that there was no decrease in flux through the
pathway. In fact there was a 20% increase in flux through the shunt. This probably
represents a compensatory mechanism that is not understood. The effect of thiamine
deficiency on transketolase in this study was a 65% decrease in activity. There are
some limitations to this method (Katz–), and whole brainstem was used, so a more
anatomically discrete study might yield different results. This study does not, how-
ever, argue in favor of a significant role for the hexose monophosphate shunt in the
pathogenesis of the symptoms of thiamine deficiency.
Pyrithiamine-induced thiamine deficiency can be produced in mice in 10–12
days, and this model was the basis for an electron microscopy study (Watanabe
et al., 1981). Groups of mice included the following: (1) those made thiamine defi-
cient using a thiamine-deficient diet plus pyrithiamine; (2) mice treated as in group
1, then reversed by injection of thiamine; and (3) mice made thiamine deficient
as above, then reversed, and then maintained on a thiamine-containing diet for 2.5
weeks. Lesions in the brains of thiamine-deficient symptomatic mice were local-
ized in the thalamus, pontine tegmentum, and mammillary bodies. These lesions
were characterized by edematous necrosis localized to neuron dendrites, myelin
sheaths, and astrocytes. Also present was hemmorhagic necrosis. Following rever-
sal of symptoms (48–96 hours), edema was still present, and fat-laden macrophages
were noted. After 2.5 weeks of “reversal,” some edema was still present, as well
as Wallerian type of degeneration. It was noted that the lesions resembled those of
non-hemmorhagic Wernicke’s disease. Vascular permeability to horse radish perox-
idase was seen in severely affected mice, but it reversed when the symptoms were
reversed with thiamine treatment. The findings suggest that both nerve and vascu-
lar lesions are present, but that the nervous lesions are more significant than the
vascular changes.
Another study examined neuropathological findings in a series of dogs made thi-
amine deficient by dietary means (Read and Harrington, 1986). Lesions occurred
in two varieties: (1) in this group, lesions were confined to the inferior colliculus,
and (2) a group consisting of much more widespread lesions in the cerebral cor-
tex, inferior colliculus, cerebellar nodulus, and medial vestibular nucleus. In both
groups, gray matter was the site of lesions, and the lesions were bilaterally symmet-
rical. Microscopically, the lesions were varied, but included spongiosis, vacuolation
of neuropil and myelin sheaths, neuronyl necrosis, hypertrophy and hyperplasia of
endothelial cells, glial necrosis, and accumulation of lipid containing phagocytes.
The presence of two different neuropathological patterns was seen as a different

chronic pattern and different symptoms in the affected dogs. The range of time for
the development of overt thiamine deficiency was 32–134 days.
Another study (Aikawa et al., 1984) examined the neuropathological and bio-
chemical effects of an acute onset of pyrithiamine-induced thiamine deficiency. Wis-
tar rats were used in this study, and results showed the usual weight loss, and by days
11–12, a rapid onset of symptoms including ataxia, opisthotonus, tonic seizures,
etc. Grossly, the brains appeared normal except for occasional petechial hemor-
rhages. Results from light microscopic examination showed spongy changes in
the thalamus, mammillary bodies, and lateral vestibular nucleus in the pons. These
changes were bilaterally symmetric. Except for severe cases, lesions were limited to
astrocytes and oligodendrocytes. In severe cases neurons and myelin sheaths were
involved. Petechial hemorrhages were infrequently seen. Electron microscopic find-
ings were consistent with those seen in light microscopy.
Biochemical analysis included measuring thiamine and high-energy phosphates.
Results showed that cerebral thiamine dropped to less than 20% of control lev-
els, a finding previously described (McCandless and Schenker, 1968). High-energy
phosphates were decreased in symptomatic animals; ATP was down 10% and
phosphocreatine decreased by 30%. These data are somewhat at odds with other
investigators’ results, but the rats in this study were experiencing tonic seizures.
Seizures are known to have a deleterious effect on cerebral energy metabolism
(McCandless et al., 1979). Most previous studies on thiamine-deficient animals
stress that seizures were not a clinical feature of murine thiamine deficiency. There-
fore, care must be taken in the interpretation of the high-energy phosphate data in
the paper by Aikawa et al. (1984).
Another study looking at development examined the effects of pyrithiamine-
induced thiamine deficiency on rat brain myelination (McCandless et al., 1976b).
Newborn rats were injected with a low or high dose of pyrithiamine, then killed
at various times and brains analyzed for transketolase and pyruvate decarboxylase,
and brains were stained for myelination, thought to be dependent on the hexose
monophosphate shunt.
Results were that pyruvate decarboxylase, as indicated by pyruvate levels, were
depleted by about 30%. Similarly, transketolase levels were depleted by as much as
36% in high-dose rats. Staining of brains by the Weil technique showed no alter-
ation in amount or timing of myelin deposition. Myelin components such as phos-
pholipids, cholesterol, and cerebrosides were not depleted in experimental animals.
The newborn rats were symptomless and had normal weight gain. These data indi-
cate that in spite of an enzymatic defect that is lethal in adult rats, the newborn rats
develop normally, and their myelination is normal. This is interpreted as suggesting
alternative pathways supporting myelination were available to newborn rats or that
some heretofore undescribed protective device is in place permitting myelination.
Cerebral blood flow in thiamine-deficient rats has been measured regionally
(Hakim, 1986). 14C-iodontipyrine was used to image and measure regional cerebral
blood flow in rats rendered thiamine deficient using a thiamine-deficient diet plus
pyrithiamine injections. Animals became anoxic and started loosing weight after
day 10. By day 18, animals suddenly started showing neurological symptoms such
as ataxia, opisthotonus, and loss of righting response. PO2 levels were decreased,
and symptomatic rats were hyperglycemic. 14C-iodoantipyrine autoradiographs
showed evidence of early hyperfusion, especially in areas known to be affected
in thiamine deficiency, such as the thalamus and inferior colliculus. There was a
delayed hypoperfusion seen later in the pathologic process. The author attributes
these local blood flow changes to metabolic changes in local cerebral glucose uti-
lization described earlier (Hakim and Pappius, 1981).
Acetylcholine has been measured in the brains of thiamine-deficient animals with
divergent results. The rationale for looking at acetylcholine is that its production
might be decreased either by interference with acetyl-CoA production via pyruvate
decarboxylation activity depletion or by decreased production due to lowered ATP
levels.
In the study described (Vorhees et al., 1977), normal controls and pair-fed con-
trols were compared to thiamine-deficient rats in terms of acetylcholine levels and
also turnover of acetylcholine. Microwave irradiation was used for killing; this tech-
nique minimizes any chance for decay of labile metabolites before assay. Animals
were made thiamine deficient using dietary means, and pair-fed controls were fed
normal diet in the amount the thiamine-deficient rats consumed the previous day.
Acetylcholine was measured in several brain regions, including cerebral cor-
tex, diencephalon, midbrain, cerebellum, and medulla-pons. Results for net acetyl-
choline levels showed no significant differences in levels in any region in deficient
rats as compared to either pair-fed controls or rats fed a normal ad lib diet. Uti-
lization of acetylcholine following inhibition of synthesis with hemicholinium-3
showed a small but significant regional decrease in utilization. This decrease was
only statistically significant in the midbrain. The data, taken together, point to
impaired synthesis of acetylcholine. If one supposes decreased synthesis of acetyl-
choline, the mechanism may not be through acetyl-CoA since it has been shown to
not be altered in thiamine deficiency (Reynolds and Blass, 1975).
Since transketolase, a thiamine-requiring enzyme, is rate limiting for the hex-
ose monophosphate shunt, lipids and myelination have been investigated in young
developing rats (Geel and Dreyfus, 1975; McCandless and Cassidy, 1976). In the
first study, the lipid composition was studied in brains of 25-day-old rats made
thiamine deficient from gestational day 14. These experimental animals displayed
weight loss and neurological symptoms such as tremor, opisthotonus, and ataxia.
The undernutrition pair-fed control rats had a reduction in brain lipids. Thiamine-
deficient rats had no further change in regional brain lipids as compared to pair-fed
controls. The authors interpret these findings to cast doubt on the hypothesis that
depleted transketolase limits the hexose monophosphate shunt’s ability to gener-
ate reducing power (NADPH) for lipid synthesis. In general agreement with these
results is a study looking at myelination in thiamine-deficient newborn rats (McCan-
dless et al., 1976b). This study showed normal myelination throughout the critical
period for myelin deposition, in spite of thiamine deficiency.
In another interesting study (Mesulam et al., 1977), rhesus monkeys were grad-
ually weaned onto a thiamine-deficient diet. When the deficient group became
severely affected, they were reversed with thiamine and then made deficient again.
Clinical manifestations were carefully monitored, and blood transketolase monitor-
ing confirmed thiamine deficiency.
Clinical signs and symptoms included apathy, anorexia, nystagmus, ataxia, and
heart failure. Onset of these symptoms ranged from 40 to 100 days. These symptoms
were reversed with thiamine administration for several days and then the animals
were again made thiamine deficient. Several episodes of this regime were per-
formed, and clinical signs tended to worsen with each episode. This progression
may have been due to slowly developing brain damage.
This study is reminiscent of the history of chronic alcoholics, and the signs and
symptoms displayed by these thiamine-deficient rhesus monkeys were strikingly
similar to those seen in the Wernicke-Korsakoff syndrome. Transketolase levels
in blood confirmed chemically that the monkeys were indeed thiamine deficient.
The authors make the point that the treatment of chronic alcoholics in hospi-
tals with large thiamine doses may increase variations in thiamine levels. The
present study in monkeys shows that reversal of symptoms with large thiamine
doses may make subsequent deficiency status worse clinically. There is a possi-
bility that similar “up and down” thiamine levels in humans may have a deleterious
effect.
The effect of thiamine deficiency on the neurotransmitter serotonin has been
examined (Plaitakis et al., 1978). The rationale was that evidence existed for a
possible role of thiamine in nerve conduction, as described above. In this study,
both pyrithiamine and a low thiamine diet were used to produce thiamine defi-
ciency. Results showed that other neurotransmitters studied such as GABA, glu-
tamic acid, norepinepherine, and choline were not affected by thiamine deficiency.
Serotonin, however, was affected in that synaptosomes from the cerebellum of
thiamine-deficient rats showed the Vmax for serotonin uptake was only 50% that
of controls.
The decreased uptake of serotonin was limited to the cerebellum; other regions
were not affected. Thiamine added in vitro did not restore decreased serotonin
uptake; however, when thiamine was added in vivo, reversal of symptoms occurred,
as well as reversal of decreased serotonin uptake. Thiamine deficiency in rats pro-
duces seizures in some studies, and hypothermia in all studies. These two symp-
toms alone are associated with decreased serotonin uptake. Serotonon uptake is
energy dependent and there are some mild effects of thiamine deficiency on energy
metabolism. Taken together, it seems possible that the changes in serotonin in
thiamine-deficient rat brain are secondary to some other key changes.
A close relationship exists between the excitatory amino acids glutamate and
aspartate, and the thiamine dependent enzyme alpha-ketogluterate dehydrogenase.
This prompted investigators to examine the effect of pyrithiamine-induced thi-
amine deficiency on regional cerebral glutamate and aspartate metabolism (Butter-
worth and Heroux, 1989). Measurement of these two amino acids in symptomatic
pyrithiamine-induced and symptomatic thiamine-deficient rats showed decreases
in glutamate in the thalamus and pons and decreases in aspartate in the tha-
lamus, pons, cerebellum, and cerebral cortex. The enzyme alpha-ketoglutarate
dehydrogenase was decreased regionally, whereas activities of pyruvate dehydro-

genase were unchanged.
Upon reversal of symptoms with thiamine administration, concentrations of glu-
tamate and aspartate, and alpha-ketoglutarate returned toward normal. Alanine (also
measured) levels increased in the areas studied during the symptomatic stage and
returned to normal with thiamine treatment. Alanine is often associated with pyru-
vate decarboxylase activity, but the enzyme was unchanged in this study.
Pyrithiamine is used to produce thiamine deficiency, coupled with thiamine-
deficient diet. The advantage of using the antagonist is that symptoms are usually
produced in 10–12 days, as compared to 30+ days required with thiamine-deficient
diet alone. Another advantage is that the symptoms are compressed into a day or
so, whereas onset of symptoms generated by dietary means alone might take 3–4
days. This makes the pyrithiamine-induced thiamine deficiency model more effi-
cient. It is, however, a different model as shown clearly in the above study. For
example, pyruvate decarboxylase activity is unaltered in the above pyrithiamine
model, and it is always changed in the dietary method. When making correla-
tions to human syndromes such as the Wernicke-Korsakoff disease, one should
be careful to use a murine model (dietary) most closely resembling the human
counterpart.
In another study (Fournier and Butterworth, 1990), thiamine-dependent enzymes
were measured in thiamine-deficient pregnant rats and their offspring. Maternal rats
had decreased transketolase levels in the brain regions cerebral cortex, cerebellum,
and brain stem in thiamine deficiency lasting 22 days. Pyruvate dehydrogenase and
alpha-ketoglutarate dehydrogenase activities were unaltered in maternal rats in the
three regions examined. Neurological symptoms were not mentioned in this paper,
but symptoms do not occur in dietary-induced thiamine deficiency until around day
28. Red blood cell transketolase activation (TPP effect) was used to assess thiamine
deficiency, and it was decreased.
The offspring at 13 days of age showed decreases in pyruvate dehydrogenase
and alpha ketoglutarate dehydrogenase in the cerebral cortex, and transketolase was
decreased in all three regions: cerebral cortex, cerebellum, and brainstem. These
changes could certainly have deleterious effects on brain development during crit-
ical periods for metabolic activity. The peculiar findings as regards the distribu-
tion of effects on pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase
remain unexplained, but could be a non-descript early deficiency change. Transke-
tolase activity changes might affect lipid/myelination metabolism; however, earlier
studies show no effect on myelin formation, and flux through the hexose monophos-
phate shunt is not affected by a moderate drop in transketolase activity.
A method for producing murine pyruvate dehydrogenase deficiency has been
developed in pregnant mice. The mutation is x-linked, and results in no male
live births, and a 50% reduction in female live births (Pliss et al., 2004). Bio-
chemical results showed a reduction in brain and liver pyruvate dehydrogenase
activity by about 25%. Cells showed a decrease in immunostaining for pyruvate
dehydrogenase in the neocortex, cerebellum, and basal ganglia. The caudate and
putamen showed a decreased organization of fibers as compared to controls. The
cerebellum was overall smaller, with granular and Purkinje cells being reduced in
numbers. Midbrain structures were largely unaffected.
This model of pyruvate dehydrogenase deficiency allows the assessment of the
deficiency on the developing CNS. Results showed that a threshold exists for the
activity of pyruvate dehydrogenase such that a loss of more than about 25% is not
compatible with life. This is in keeping with earlier studies in adult mammals show-
ing that a drop of about 25% of activity in pyruvate dehydrogenase was associ-
ated with severe neurological symptoms and death (McCandless, D. and Schenker,
1968).
The question of selective vulnerability in thiamine deficiency has been one of the
key unanswered problems and has been recently examined. Selective vulnerability
means highly localized brain regions, and/or individual cell types are affected. This
question arises because of the striking localization in human cases of Wernicke’s
disease of lesions to sites such as the mammillary bodies. Since the highly localized
nature of the lesions is at the lower limits of direct biochemical analysis, alternative
techniques have been utilized.
In one such study (Pannunzio, P. et al., 2000), cerebellar granule cells
(CGCs) were utilized in culture to examine various aspects of thiamine defi-
ciency in a specific cell type. Briefly, CGCs were prepared from 7-day-old rat
cerebella, and cultured cells were given Dulbecco’s modified Engle’s medium,
deficient in thiamine. Pyrithiamine was also utilized to produce thiamine
deficiency.
Results showed that CGCs exposed to the thiamine-deficient media resulted
in a 97% decrease in thiamine, whereas pyrithiamine treatment yielded a less
dramatic decrease of 63%. Transketolase activity was decreased by exposure to
thiamine-deficient media and to media containing pyrithiamine. In addition, alpha-
ketoglutarate dehydrogenase activity was decreased by pyrithiamine treatment,
whereas thiamine-deficient media alone did not affect alpha-ketoglutarate dehydro-
genase activity. Pyruvate dehydrogenase activity was not changed by either treat-
ment. In addition, ATP levels were lowered by 35 and 68% with deficient media
and media plus pyrithiamine, respectively. Pyrithiamine treatment resulted in a 27%
increase in cell death.
This study represents the first to report effects of thiamine deficiency on cul-
tured cerebellar cells. The decrease in alpha-ketoglutarate dehydrogenase activity
was greater in pyrithiamine-treated cells than was the decrease in transketolase.
Thiamine-deficient media alone did not alter enzyme activity levels, although thi-
amine biphosphate levels were lower. That the 80% decrease in alpha-ketoglutarate
dehydrogenase activity in pyrithiamine-treated cells led to a drop in ATP is not sur-
prising. This drop in enzyme activity is greater than that seen in vivo. In addition, a
lowering of pyruvate dehydrogenase activity is seen in vivo, but was not observed
in these cell culture experiments.
In another study (Ke, Z. et al., 2003), the time course and length of time before
irreversible effects occurred were determined. The authors determined that by days
10–11 after initiation of thiamine deficiency via dietary means plus pyrithiamine
injections, lesions occurred that were essentially not reversible.
Results showed the most sensitive brain area was the thalamus, specifically the
submedial thalamic nucleus. Clinically, the mice were reversible up to days 10–11,
at a time when symptoms were severe. By day 11, the altered histological picture
included the entire thalamus. Symptoms could be reversed up to days 10–11. Neu-
ropathological changes included hemorrhagic necrosis and degeneration of neurons.
This started in the submedial thalamic nucleus and then spread to the other thalamic
nuclei. Hemoxygenase-1 microglia were increased by day 10 of treatment, while
NeuN-positive neurons were decreasing in numbers.
Results also showed that after 8 days on the thiamine diet with pyrithiamine, thi-
amine administration could not reverse impending cell death but prevented further
cell death. Microglial responses were similar to those of neurons. Some irreversible
changes had occurred by day 9, but treatment with thiamines slowed or stopped the
process.
The authors point out that studies of neuronal cell death need to be correlated
with the responses of other cells such as microglia, astrocytes, endothelial cells, and
mast cells. The authors suggest that in this model, somewhere in day 9, the tide turns,
as it were, and irrevocable changes cannot be avoided. Administration of thiamine
on day 11 produces an increase in the numbers of microglia and HO-1 positive
microglia, thereby increasing phagocytic activity. That there are clear differences
between thiamine deficiency in man and the rapid production of thiamine deficiency
in animals using pyrithiamine should be remembered.
In an attempt to localize and examine mechanisms of the distribution of lesions in
thiamine-deficient rats, histamine positive neurons and localization of mast cells in
the inferior colliculus and the thalamus were studied (Meng, J. and Okeda, R., 2003).
Immunohistochemistry was used to determine the distribution of mast cells, and vul-
nerable and non-vulnerable regions of thiamine-deficient rats. Results showed simi-
lar results between vulnerable and non-vulnerable regions in thiamine-deficient rats.
The authors concluded that a role for histamine in thiamine-deficient lesions could
not be supported from these studies. The authors speculate that spongy lesions seen
in the inferior colliculus and the thalamus are due to ischemia. These spongy lesions
are not seen in the cerebral cortex or other unaffected areas. The acute regional neu-
ronyl changes may be due to the ischemic condition of the spongy areas. Proof of
this awaits further experimental studies.
In another recent paper (Wang, X. et al., 2007), endoplasmic reticulum
stress (ER stress) was examined in thiamine-deficient rats and mice. Tissue cul-
ture from the cerebellum was utilized; thiamine deficiency was produced by
administering a thiamine-deficient diet supplemented by daily injections of
pyrithiamine.
The effect of thiamine deficiency was examined using various methods. For
example, there was an increase in GRP78 expression in thiamine-deficient mice,
reaching a 5.5-fold increase on day 7 of treatment. The expression of GADD153
and phosphorylated eLF2 alpha correlated with the expression of GRP78. Caspase-
12 activation was also deemed increased by 4.6 times on day 6 of thiamine defi-
ciency. The neurological status of thiamine-deficient animals was not stated, and
pair-fed controls were not used.
The thalamus was also subjected to analysis since that brain area is affected
by thiamine deficiency. Results were similar to those described in the cerebellum,
except that caspase-12 was unchanged.
A thiamine analog, amprolium, was also used to evaluate cerebral changes.
Results showed that amprolium-induced alterations in markers of ER stress simi-
lar to changes noted with a low-thiamine diet plus pyrithiamine. Caspase-12 was
also increased.
The authors speculate that these data support the idea that ER stress may be tied
to CNS lesions. Other CNS lesions such as those of Alzheimer’s disease and Hunt-
ington’s disease are associated with changes suggestive of ER stress. The authors
further suggest that since thiamine deficiency and other neurodegenerative disor-
ders share some common features, thiamine deficiency might serve as a model for
these other disorders. The authors state that the change in ER stress occurs on day
5 after inanition of thiamine deficiency, therefore preceding the lesions. Whether
these changes in ER stress actually are directly responsible for the neuropatholog-
ical changes in brains of thiamine-deficient animals remains unclear. The changes
in caspase-12 in lesions of thiamine-deficient animals can be blocked by Z-ATAD-
FMK, a synthetic peptide that inhibits caspase-12. This inhibition acts to protect
against thiamine deficiency induced cell death. The mechanism of this finding is
also unclear. Remember that this study was performed using the antithiamine com-
pound pyrithiamine, which produces thiamine deficiency very quickly (three times
faster than thiamine-deficient diet alone), and without the benefit of pair-fed con-
trols, which account for possible effects of starvation. Starvation alone produces
significant alterations in brain chemistry.
Beriberi
The condition known as beriberi has been recognized for hundreds of years. The
first written description was published in 1642 by Jacobus Bontius. Bontius was
a Dutch physician who traveled to the West Indies and observed several tropical
disorders including beriberi. These observations were published in a book enti-
tled de Medicina Indorv Lib 4. A translation of beriberi observations appears in
Appendix B. Interestingly, a couple of years after the Bontius book appeared,
another Dutch physician, Tulp, also published a description of beriberi. Tulp was
also well known as the subject of a famous painting by Rembrandt titled “Dr Tulp’s
Anatomy Lesion.”
It is important to note that beriberi is largely a disease of the past. In the chapter
on thiamine deficiency and world health, various outbreaks of thiamine deficiency
and beriberi are described, but these are isolated occurrences. Beriberi was a major
health problem 100 years ago in areas such as Malaya. In areas where wages were
low, workers consumed large quantities of polished rice. Chinese mine workers,
for example, working in jungle locations relied on food brought in. This usually
consisted of large amounts of polished rice. Frequently all daily meals consisted
mainly of rice. It was estimated that a worker might consume as much as 2 lbs (!) of
rice per day. The sad outcome was that in 1900, one estimate was that in a remote
mine, 800 of 2400 workers died from beriberi in a 2-year period.
As time went on, roads were built to remote mining areas, facilitating shipment
of food. Mining areas became towns, with increased demand for food variety. This
brought a decrease in the incidence of beriberi. Even reported beriberi incidences
are low due to the fact that many cases are of a mild chronic neuropathy, which
was unreported. Another interesting phenomenon was that with prosperity came
increased cases of beriberi. The reason was that people could purchase highly pol-
ished rice, whereas in bad times, they polished their own with less effective means.
There are many reported cases in which the incidence of beriberi rose following the
introduction of small rice mills into a community. WW2 certainly brought about
conditions in which people consumed what was available, increasing rice consump-
tion and the incidence of beriberi.
Beriberi in male adults usually falls into one of three categories: a chronic dry
atrophic type, a mild subacute form, or an acute fulminating type. The subacute mild


form of beriberi constitutes about 80–90% of all cases. The first type of beriberi is
usually seen in older patients. This form is also associated with alcohol consump-
tion. In these cases, wine is frequently warmed in lead-lined pots, and lead intox-
ication is part of the condition. There is also an association with opium addiction
(Platt, B. and Alley, R., 1942). The acute phase of beriberi can convert to the dry
chronic form in certain circumstances (Lim, E., 1938).
Clinical manifestations of the mild subacute cases of beriberi vary somewhat.
The signs and symptoms of cardiovascular involvement in this form of beriberi are
more significant than any others. They generally consist of breathlessness on any
activity, and of chest palpations. There is edema present in most cases, and it may
be associated with congestive heart failure. The heart is almost always enlarged, and
tachycardia is present. Cardiac dilation may be the first cardiovascular change and
usually precedes neurological changes (Aalsmeer, W. and Wenckebach, K., 1938)
(Fig. 1).
Beriberi Clinical Symptoms
Source Breathlessness Chest Palpations Edema Enlarged Heart Tachycardia Neuropathy
Aalsmer,
Et.al. + + + + + +
De Langen &
Lichtenstein + + + +
Denny-Brown + + + + +
Fig. 1
Frequently, the onset of symptoms is linked to a febrile episode. This is asso-

ciated with an increased need for thiamine, documented many times elsewhere.
Patients in the subacute group may also show a deficiency in other essential nutri-
tional substances such as ascorbic acid, niacin, and vitamin A. Most of these patients
also suffered from protein deficiency, evidenced by leg ulcers of a type associated
with protein deficiency (Platt, B., 1953).
In one situation, in an area where there were several factories that provided meals,
workers from one factory had a much lower incidence of beriberi than others. The
difference was that the beriberi-free factory served rice which was milled, washed,
and cooked in a way such that almost no thiamine was lost. Other groups used
commercial sources of rice prepared in a way that lost almost all of the thiamine.
In the factories using commercial rice, over time almost 75% of workers developed
edema and early signs of thiamine deficiency, whereas workers at the site carefully
preparing their rice daily had only a couple of cases of edema. Workers in factories
fed commercial rice were apathetic and listless, while workers fed daily prepared
rice that maintained thiamine content were “high spirited.”
The acute fulminating form of beriberi has an incidence of only about 5% of all
beriberi cases. This form of beriberi is the most serious form, and until thiamine
Beriberi 33
treatment was developed and utilized, nearly all patients died. The signs and symp-
toms of patients with the severe form have been described (de Langen, C. and Licht-
enstein, A., 1938). This clinical account has been cited as an excellent description by
others (Hawes, R. et al., 1937) and is now quoted verbatim from the above reference
(de Langen):
“The whole picture here is dominated by the insufficiency of the heart and vessels, and
usually runs a fatal course. The milder attacks may pass over to this type, but as a rule per-
nicious beriberi breaks out suddenly in a peracute form, before any other manifest symptom
has proclaimed the presence of beriberi in the patient. When the nerve lesions develop early,
we do not see Shoshin; the heart is saved from this extreme insufficiency the patient’s being
forced to a complete rest at an early stage of his attack. If the nerve symptoms come on late,
and the patient with his labile heart and vessels can remain for some time on his feet, and
must often even continue to work in order to earn his daily bread, the best possible condi-
tions are present for the sudden development of Shoshin. Patients are sometimes known to
die of this form of beriberi when next to no edema are as yet demonstrable.
The death from Shoshin beriberi is often very terrible. The patients are severely dysp-
noeic. Violent palpitations of the heart allow them no moment of rest. An intense precordial
agony is often one of the most distressing symptoms. The patient is very restless; he cannot
lie still for an instant, but tosses and turns about violently from side to side in bed the whole
time. The precordial oppression is often felt in the epigastrium. It is a feeling of heaviness,
constriction, and oppression, sometimes of actual pain. The patient moans and shrieks, and
his cries take on a rather special character because of the coincident hoarseness or some-
times aphonia. He is often intensely thirsty, but reacts to drinking by vomiting. He has an
anxious look on his face. His pupils are dilated; his respirations frequent and superficial.
In spite of his dyspnoea his accessory muscles of respiration are brought but little into
action, often not at all. He does not want to sit up in bed and help his respirations with
both hands as is the case with other dyspnoeic patients, but, so far as his restlessness will
permit it, he prefers to lie flat in bed. He does not cough or spit. In the heart region, the
epigastrium and the neck, widespread and powerfully undulating pulsations are to be seen.
There is an obvious cyanosis, more marked during inspiration. The pulse is fuller than
might be expected, is usually regular, even and weak, with a frequency of from 120 to
150 per minute. A very high frequency of the heart is seldom seen. Over the heart area
a wave-like motion may be felt. The liver is enlarged and tender. The epigastric region
is spontaneously painful. Percussion shows the heart to be enlarged in all directions, but
mostly to the right. The broadening of the area of aortic dullness, caused by the vascular
stem, is plainly demonstrable extending into the left. Changes in position affect the heart
outline markedly; it conforms to the position assumed by the patient. Over the body of the
organ systolic murmurs may sometimes be heard, but not always. The heart action may
only be described as tumultuous, although the sounds are not so powerful as in the last type
of the disease described. Murmurs and sounds are difficult to localize. Vessel sounds are
heard not only over the femoral artery, but also over the vessels of the elbow. These may
become so loud that, as Shimazono has reported, they may even be heard without the aid of
a stethoscope and in some instances at a distance from the bed. The systolic pressure, which
at first remains normal, finally falls slowly to below 100, often reaching 80 or even 70.
Strangely enough there is but little disturbance to be found in the lungs; at most there
is a mild typany on percussion. Just before death there tends to be a slight dullness at the
lower borders. The respiration is rough and whistling, but without rales. As the condition
becomes worse, rales may appear, at first dry and high-pitched, but soon becoming more
moist. At the same time the pulse becomes thinner, the veins dilate more and more and the
patient dies intensly dyspnoeic, but usually fully conscious. It is but seldom that a patient
survives an attack, but even if we happen to be fortunate enough to bring him into a less
immediately threatening condition, the risk of a sudden relapse is never absent.”
There are some differences in symptoms in the severe form, one of which is the
degree of edema shown by various patients. Some patients have little to no edema,
while others have a great deal. Since classification of the type of beriberi between
wet or dry depends in large part on the degree of edema, the absence or near absence
of edema can make classification difficult. It should be remembered that the relative
amount of edema in any given patient depends on conditions such as vomiting, heart
failure, renal failure, and possible alcohol consumption.
Pyruvic acid is elevated in blood of patients with the severe fulminating variety of
beriberi and is also elevated to a lesser extent in the mild form. Upon administration
of thiamine, pyruvic acid returned to normal values in only a few hours. At the same
time, there was an overall improvement in signs and symptoms, emphasizing the
“biochemical lesion” nature of the disorder. Other changes reverting toward normal
occurring with thiamine treatment, such as heart dilation, take several weeks. Also,
diuresis in patients with severe edema may take several days to begin and several
more days for edema to subside (Fig. 2).
Beriberi: Frequent blood findings

Increased pyruvic acid
Increased lactic acid
Increased blood volume
Decreased thiamine
Increased thiamine pyrophosphate effect
Decreased magnesium
Fig. 2
Most knowledge of the pathological changes seen in patients dying of beriberi

are derived from early studies (Follis, R., 1958a). Much of this is because, since the
development of thiamine knowledge, and treatment programs, which are widespread
and simple, death from beriberi is rare. At a time when autopsies were being done
on beriberi patients, descriptions and techniques were somewhat primitive.
Cardiopathology at autopsy is usually described as a dilation of the heart with
concomitant hypertrophy (Follis, R., 1958b). Microscopically, fragmentation of car-
diac fibers and fatty infiltration was seen. Also noted was cloudy swelling and areas
of frank necrosis. Edematous and hydropic changes in cardiac muscle have also been
noted. No areas of inflammation have been described. These pathological changes
may be only moderate because this is a biochemical lesion, and more structural
changes could be expected over a longer time course.
Changes in peripheral nerves consist of myelin loss and nerve fiber axoplasm dis-
ruption. Degeneration in the spinal cord was described (Durck, H., 1958), although
minimal changes in the brain were described. Sometimes changes in medullary neu-
rons were noted, but certainly none of the dramatic changes seen in patients with
Wernicke’s disease.
Beriberi has been described from a neurological point of view in prisoners of
war (Denny-Brown, D., 1947). In this description, the key feature in prisoners with
Beriberi 35
beriberi was the presence of foot and wrist drop. This finding was associated with
tenderness of muscles over the lateral portion of the legs, lateral aspect of the thighs,
and lateral forearms. Also seen were pain in the foot, paresthesias, and absence of
ankle jerks. Ataxia was not an early sign of beriberi.
It is noted that alcoholic polyneuropathy could be quite similar to that seen in
beriberi. A key difference, however, is that in the case of alcohol-related symptoms,
treatment of the patient with thiamine while continuing alcohol intake improves the
symptoms. Also, in alcoholism, there is usually more widespread sensory neuropa-
thy and burning paresthesias in the feet.
Cardiac changes were seen in essentially all prisoners suffering from beriberi.
Tachycardia was usually present, as well as palpitations. Also present is conges-
tion in the venous circulation. Warm climates, and/or fever can precipitate more
overt symptoms in chronic patients. The seasonal association of onset of symp-
toms in warm climates acts to increase congestion in venous circulation, precip-
itating cardiac pathology instead of neurological complications. The edema is of
course a concomitant of the peripheral circulatory condition. This might explain
why thiamine deficiency in the orient produces beriberi, whereas thiamine defi-
ciency in cooler climates (US) is more conducive to the production of a neuro-
logical disorder (Wernicke’s disease). In keeping with the above, elevated pyru-
vic acid levels in blood are usually present in beriberi in warm climates, but
elevated pyruvic acid in cooler climates and in Wernicke’s disease is not so
constant.
While beriberi is usually thought of as an adult disease, more descriptions recog-
nize a form of beriberi that affects infants. The infantile form of beriberi has been
described by Burgess (1958). He states that there were incidence rates which varied
between countries. For example, in Japan, only 3.5 deaths per 1000 births could
be attributed to beriberi. In the Phillippines, the incidence of beriberi in infants
under 1 year at autopsy was 56%. Also noted was that nearly all deaths occurred
between 1 and 3 months of age and that these infants were breast fed. Andrews
(Andrews, V., 1912) demonstrated the connection to breast milk by feeding new-
born puppies human breast milk from mothers with beriberi, and thus producing
beriberi in the newborn dogs. Around 1912, investigators (Chamberlain, W. et al.,
1912) were able to use an extract from rice polishings, which was able to treat mild
cases of beriberi and could be used to prevent the disorder. By the early twentieth
century (Albert, J., 1947), extensive use of the extract from rice polishings had cut
the mortality rate of infantile beriberi by 80%. By the end of WW2, the incidence
had further declined.
Other countries reported outbreaks of beriberi. There was an outbreak in the
Nauru Island (Bray, G., 1928); see also the World Health chapter in this volume. The
age of onset was from the 8th to 12th week of age, and the course was rapid. Many
infants died, but the mothers showed little beriberi symptoms. Another location of
infantile outbreaks was Singapore where hundreds of cases were seen. In 1945 over
1000 cases of infantile beriberi were described in Madras (Krishnan, B. et al., 1945).
In Hong Kong there was a large incidence of infantile beriberi, accounting for a mor-
tality rate of 300/1000. At least 18% of newborns coming to clinics were affected,
and the mortality rate was 90% (Fehily, L., 1944). There are many more reports
of similar outbreaks. Many other outbreaks from isolated areas are probably under-
reported or not reported at all.
One reason given for the decline in the incidence of infantile beriberi after WW2
is the increased status of women. As their status improved, they were able to eat all
of the foods prepared for meals instead of leftovers. In some countries, there was
a habit of restricting the diet of mothers just having given birth. The restriction to
rice and dried fish lasted around 6 weeks, putting the mother at risk for thiamine
deficiency. Once this practice was reduced, infant mortality from beriberi dropped.
It is noted that the mother of infants with beriberi frequently have signs and
symptoms of beriberi. This finding is associated with the acquisition of infantile
beriberi by the offspring. In these cases, breast milk seems to be the direct cause
of morbidity and mortality, whereas in most cultures breast milk promotes healthy
babies.
The onset of infantile beriberi between the first and fourth postnatal months prob-
ably reflects some storage of thiamine in the newborn due to placental transfer of
the vitamin, coupled with lower requirements for B1 immediately after birth. Onset
after 4 months of age usually signals a more chronic course.
Clinically, onset is rapid. The first signs may be a sudden attack of screaming,
with stressful respiration and cyanosis. This phase passes, but recurs with increased
frequency and intensity. Unless thiamine treatment is quickly initiated, death ensues.
Aphonia may accompany this acute attack in which the child appears to be cry-
ing/screaming, but no sound is heard. This is attributed to laryngeal nerve paralysis.
A variation of infantile beriberi is one in which some features of Wernicke’s dis-
ease are seen, including nystagmus, strabismus, seizures, and an elevation of body
temperatures.
Pathological studies of infantile beriberi are spotty, as they are in adult beriberi.
The heart in infantile beriberi is described as being greatly enlarged – as large as
that of a 10-year-old child. The pulmonary artery and aorta are equal in size, and
microscopically, muscle fibers appear swollen. Congestion and edema have been
described in liver, spleen, and brain. Biochemical tests of maternal breast milk
show a lower thiamine level in areas in which infant beriberi is common as com-
pared to breast milk in areas where the disorder is non-existent (Ramasatri, B. and
Indravathi, D., 1957). Levels in areas of infantile beriberi were one half those of
healthy areas. The volume of milk produced by mothers of infantile beriberi infants
seems to be about the same as healthy mothers.
Treatment of infantile beriberi with thiamine may have obvious signs of improve-
ment in as little as 6 hours and possibly earlier. This is clearly a biochemical lesion,
and certainly pure thiamine deficiency. Such rapid return of function has repeatedly
been seen in experimental animals with thiamine deficiency.
Reducing and/or eliminating the incidence of beriberi both in adults and in
infants involve increasing the intake of thiamine. One way to achieve this is to
improve the diet such that rice becomes a much less important part of the diet.
This goal is being achieved around the world. A second way is to decrease the
amount of milling of rice, which increases the thiamine content of the rice. This
Beriberi 37
too is decreasing as knowledge of the dangers of highly milled rice increases.

Washing is another procedure in the preparation of rice that decreases thiamine
content significantly. Another method of increasing thiamine content of rice are
enrichment programs. The actual enrichment process is relatively simple, but there
are many small mills throughout affected areas, and it is not practical to oversee the
preparation (and enrichment) of these widespread mills. All of these various pro-
grams have been responsible in part for the highly significant decrease in beriberi
world wide. Finally, when beriberi does occur, modern medical abilities for rapid
diagnosis and treatment have lowered the mortality rate.
As regards animal models of beriberi (see also the chapter on animals models of
B1 deficiency in this book), a large number of studies have been published. When
a thiamine-free diet is fed to mice and rats, the main symptom complex relates to
the nervous system. Two major thiamine antagonists have been used: pyrithiamine
and oxythiamine. Of these two, pyrithiamine primarily has an effect on the nervous
system, whereas oxythiamine not only decreases the excitability of cardiac cells
and increases the duration of the action potential but also reduces Purkinje fiber
conduction velocity and spontaneous beat.
Experiments using oxythiamine in conjunction with a thiamine-free diet have
been conducted in mice, rats, and pigeons. Polyneuritic symptoms in experimental
animals on oxythiamine are rarely seen; however, bradycardia is easily produced in
rats (Gurtner, H., 1961).
Results from oxythiamine administration include a decrease in conversion of
pyruvate to acetoin, and a lowered transketolase activity (Brin, M., 1962). Another
study showed that oxidation of pyruvate was decreased in rat heart in oxythiamine-
induced thiamine deficiency, whereas oxidation of alpha-ketoglutarate activity was
decreased only in heart tissue (Gubler, C., 1961). Blood levels of pyruvate are also
elevated in animals rendered thiamine deficient with oxythiamine.
The main oxythiamine-based antagonist is not oxythiamine itself, but also
the phosphate ester, oxythiamine pyrophosphate (OTPP) (Koedem, J. and
Steyn-Parve, E., 1960). OTPP is a strong inhibitor of transketolase and thiamine
pyrophosphate. OTPP has a greater affinity for transketolase than does thiamine
pyrophosphate, thereby displacing it from the haloenzyme. Therefore, the mode of
action of oxythiamine is as a competitive inhibitor for sites on the haloenzyme.
It successfully competes with thiamine pyrophosphate for these sites, then cannot
function as a coenzyme, thus lowering enzyme activity. The advantage of using
antimetabolites in experimental animals is that they speed up the development of
symptoms. The two antimetabolites pyrithiamine and oxithiamine allow the “split-
ting” of the symptoms into neurological (pyrithiamine) and cardiac (oxythiamine).
Thus, while their use has not uncovered any particular phenomenon not already
known, they have emphasized changes in brain and in heart by separating the two
symptom complexes.
In man, clinical symptoms in beriberi consist of those of cardiac and neuropatho-
logical origin. Beriberi heart disease consists of exertional dyspnia, tachycardia,
and heart failure. This progression can occur very quickly, but can be reversed by
thiamine administration. This condition may be accompanied by edema, a usual
hallmark of beriberi heart disease. Key symptoms of Wernicke’s disease such as nys-
tagmus, opthalmoplegia, and amnesis are almost never seen in beriberi heart disease.
In an early study of cardiac lesions in thiamine-deficient rats (Ashburn, L. and
Lowry, J., 1942), thiamine deficiency was produced by dietary means, and the hearts
were examined for pathology. Results showed several rats had a clear fluid in the
chest cavity. Hearts were enlarged especially in the right auricle. The left auricle was
also dilated, but to a lesser extent. The wall thickness appeared normal. Microscop-
ically, heart muscle fibers were hyalinized, and showed increased oxyphilia in the
cytoplasm, and the nuclei were pyknotic or even absent. Necrotic muscle fibers were
noted; some fibers were fragmented. Interstitial proliferation was seen, and muscle
fibers in these areas were usually absent. There was a significant proliferation of
fibroblasts. In its worst manifestation, the entire wall thickness of the chambers
was composed of fibroblasts. Thrombi in the auricular wall were occasionally seen.
From this sampling method, all parts of the auricles were affected by thiamine defi-
ciency. Examination of the ventricles showed much less frequency of lesions. Out of
57 hearts examined, only 9 showed any ventricular involvement. When present, ven-
tricular lesions were smaller than those in the auricles. It was also concluded based
on the time course of the development of symptoms and lesions that cardiac pathol-
ogy is a feature of the late stages of dietary thiamine deficiency in the rat (Fig. 3).
In a study of thiamine deficiency in dogs, it was shown that myocardial utiliza-
tion of pyruvate and lactate was abnormal (Hackel, D. et al., 1953). In fact it was
found that the extraction of pyruvate by myocardial tissue was decreased, while the
utilization was increased. The coefficient of extraction of lactate was also lower, as
was lactate utilization. The utilization of lactate was greater than the utilization of
pyruvate. These data suggest defects in myocardial metabolism of pyruvate and lac-
tate, and in these studies, starvation was not a factor since animals were eating well
and had only minor weight loss.
In another study of thiamine deficiency in rats, EKGs were performed on nor-
mal controls and deficient rats (Yoshitoshi, Y. et al., 1961). The authors state that
there are many variations in reported findings from EKGs in thiamine deficiency.
Findings in this study included arrhythmias in 17% of cases in later stages. The
heart weight increased from the second week through the fifth week. After day 16,
fatty infiltration appeared in the myocardium representing a disturbance in glycoly-
sis. Changes in ST and T in the EKG were due to myocardial lesions after the fifth
week on the deficient diet. These ST and T changes consisted of both elevation and
depression. Other cardiac changes included an increase in right axis deviation, A-V
block, premature beat, and interpolated nodal rhythm.
In another study of thiamine deficiency, effects on sarcosomes’ oxidative phos-
phorylation and electron transport were described (Arcos, J. et al., 1964). Results
showed an increase in the heart weight–body weight ratio. Oxidative rates of cardiac
tissue with pyruvate and alpha-ketoglutarate showed a faster depletion of thiamine
pyrophosphate with pyruvate as substrate than alpha-ketoglutarate. With malate
as substrate, respiratory rates were barely changed from controls. In contrast to
respiratory rates, P:O ratios were unchanged. ATP values (not directly measured)
were probably depleted as indicated by the inhibition of sarcosomal swelling by
Beriberi 39
Fig. 3 Chest X-rays: top left = before beriberi, top right = during acute beriberi, bottom = after
11 days of thiamine treatment. Reproduced from Attas et al. (1978) with permission from Wolters
Kluwer
pyruvate–fumarate substrates. Decreased inhibition of swelling shows a decreased

ability of sarcosomes to generate ATP from pyruvate metabolism.
Since the myocardial isometric tension is an indicator of working capacity of
muscle, a study was performed to determine the effects of thiamine deficiency in rats
on this physiological phenomenon (Aldinger, E., 1965). Thiamine deficiency was

produced by dietary means. The measured myocardial tension in thiamine-deficient
rats was only a small fraction (25%) of that in control rats. For the first 2 weeks,
however, contractility was no different than controls, only beginning to drop at 3
weeks. The author also noted arrhythmias and ventricular fibrillation in a significant
number of rats. Some rats on the diet for 4 weeks, which were administered thiamine
as a treatment, returned to normal physiologically in only a few hours, whereas rats
on diet for as long as 5 weeks (and terminally sick) did not recover with thiamine
treatment.
The results of this study are in keeping with previous studies in that by 5 weeks
on diet, when animals are severely affected, thiamine treatment is to no avail. This
finding suggests structural damage. The author implies his findings are consistent
with the hypothesis that the thiamine deficiency produces lowered levels of thi-
amine pyrophosphate, thereby decreasing the potential for oxidative metabolism and
energy production. This would in turn lead to the findings described in this paper on
myocardial contractibility.
In a study using electron microscopy to examine thiamine-deficient hearts, Wistar
rats were fed polished rice and compared to control rats fed the same diet plus
thiamine (Suzuki, T., 1967). With this feeding regime severe thiamine deficiency
took up to 8 weeks to develop.
Results showed that hearts harvested from thiamine-deficient rats on diet for 5–
8 weeks were slightly dilated and slightly hypertrophied. Examination with light
microscopy, some hearts showed fatty degeneration, disappearance of cross stria-
tions, some necrosis, and edema and cell infiltration. Examination using electron
microscopy showed swollen mitochondria at 5 weeks on the thiamine-deficient diet.
Some mitochondria showed destroyed internal structure, with cristae absent. By 8
weeks on diet, swelling of mitochondria was widespread. Most were irregular in
shape, with cristae quite deranged. The sarcoplasmic reticulum was smaller and
ruptured. The severity of these structural changes including the absence of cross
striations at this time could easily explain the sudden death in the eighth week.
When the thiamine-deficient rats were “reversed” by thiamine administration
these structural changes showed diminution and a return toward normal structure.
Thus, in a group that were on a deficient diet, then treated with thiamine for 2 weeks,
mitochondria were only slightly enlarged. The cristae had an arrangement approach-
ing normal, but were still scanty. The sarcoplasmic reticulum was nearly normal in
shape, and myofibril cross striations were distinguishable.
Following 4 weeks of treatment of 8-week deficient rats, mitochondria were only
very mildly enlarged, and the cristae were all well preserved and densely arranged.
The sarcoplasmic reticulum was not significantly different from controls. Myofib-
ril cross striations could easily be seen. The author states that the finding of a
few degenerated mitochondria in the hearts from the 8-week diet/4-week treatment
group suggests that severely affected mitochondria were still not functional but that
more mildly affected mitochondria had recovered function.
An interesting significance of these observations is the ability of thiamine
to reverse not only the signs and symptoms, but also the structural changes.
Beriberi 41
Common thinking was (and is) that thiamine deficiency represents a true biochem-
ical lesion, and if it lasts long enough, then irreversible structural changes occur.
These reversible structural occurred at the microscopic level. A separate example is
the Wernicke-Korsakoff disease in which changes seen in signs and symptoms in
early Wernicke’s disease can be dramatically reversed with thiamine. Conversely,
by the time a patient reaches the Korsakoff psychosis stage, reversal with thiamine
treatment is problematic. Perhaps the structural changes need to reach a greater evo-
lution than early microscopic alterations. So here in the heart is clear evidence of
structural changes being completely reversed with thiamine treatment. This prob-
ably represents the “highest” level at which structural changes can occur without
permanent deficit.
In another study (McCandless, D. et al., 1970), cardiac metabolism was directly
assessed in order to determine if thiamine deficiency was a factor in the cardiac
signs and symptoms of beriberi heart disease. In this study, thiamine deficiency was
produced in rats using dietary means. Controls consisted of normally fed control diet
rats and rats pair-fed to the deficient rats. Deficient rats were fed thiamine-deficient
diet only.
Results showed rats at 5 weeks with neurological signs had decreased transke-
tolase activity in blood, and a higher TPP effect as compared to pair-fed controls.
In these symptomatic rats, whole heart transketolase and pyruvate dehydrogenase
activities were depressed by about 65% and 80%, respectively. The hearts of symp-
tomatic rats showed an 18% hypertrophy. In addition, cardiac ATP was decreased
by about 17% at 4 weeks in rats on the deficient diet. At 5 weeks the ATP decrease
was 34%. The depletion of cardiac ATP correlated with both the fall in PDH activ-
ity, and with the rise of cardiac pyruvate. The drop in ATP returned to normal
values within 24 hours after “reversal” with thiamine. The fall in ATP did not cor-
relate with changes in transketolase. Both of these enzymes are thiamine (thiamine
pyrophosphate) dependent. A decrease in ATP can result from decreased synthesis
or increased utilization. The data suggest the ATP drop was due to a decrease in
synthesis. The observed fall in cardiac ATP may produce an impairment in cardiac
contractibility. Overt heart failure was not seen in spite of a 34% drop in ATP, pos-
sibly because many animals died from neurological complications. The ATP drop
was about equal in atria and ventricles.
In a follow-up study (McCandless, D. et al., 1976a), the effect of thiamine defi-
ciency on rat heart pentose phosphate pathway activity was assessed. Only ani-
mals displaying severe signs and symptoms of thiamine deficiency were used. The
enzymes glucose-6-phosphate dehydrogenase (G-6-PDH) and 6-phosphoglucose
dehydrogenase (6PGDH) were measured in pair-fed controls and thiamine-deficient
rats. Results showed both G-6-PDH and 6-PGDH were not changed in deficient as
compared to pair-fed control rats. Flux through the pentose phosphate pathway was
estimated using the methodology of Katz (Katz, J. et al., 1966). While transketo-
lase is the rate-limiting step in the shunt, G-6-PDH and 6-PGDH mediate flux. In
keeping with the finding that G-6-PDH and 6-PGDH activities were not changed,
in vivo flux through the pentose phosphate pathway was not changed in deficient as
compared to control rats.
These studies were performed at a time in the course when symptoms were
severe, and transketolase levels in previous studies showed activities only a frac-
tion of control values. Based then on the levels of G-6-PDH and 6-PGDH, and
flux through the pathway, these in vivo data do not support the hypothesis that flow
through the pentose phosphate pathway is an important feature of the cardiac pathol-
ogy seen in severe thiamine deficiency. Similar findings have been noted by us in
regarding flux through the pentose phosphate pathway in liver and brain.
In another paper (Hauschildt, S. et al., 1973), the effect on isolated perfused rat
heart of thiamine deficiency was analyzed. Rats were made thiamine-deficient via
dietary means. After 21–28 days on the diet, at a time when neurological symp-
toms were severe, animals were decapitated, and the heart isolated and perfused
with Krebs-Henseleit buffer. In this preparation, contractile force was measured in
thiamine-deficient rats and controls. Lactate, pyruvate, ATP, ADP, AMP, and phos-
phocreatine were measured enzymatically. Results showed contractility strength in
thiamine-deficient rats lasted about 15 minutes in this paradigm, compared to 2–3
hours in control rats.
As regards high-energy phosphates in this in vitro model, phosphocreatine and
ATP were depleted, and ADP and AMP were elevated in thiamine-deficient rats
as compared to controls. This was despite a high oxygen tension and adequate
perfusion. When thiamine-deficient rats were reversed with thiamine treatment,
high-energy phosphates and cardiac contractility returned to normal values. This
again emphasizes the “biochemical lesion” nature of early thiamine deficiency. The
authors state that two views are held regarding the pathogenesis of cardiac malfunc-
tion. The first states that thiamine depletion results in lowered activity of thiamine-
requiring enzymes, with a resultant decrease in high-energy phosphates for cardiac
work. The second hypothesis states that blockage of thiamine-dependent enzymes
allows accumulation of a wide variety of metabolites associated with the affected
pathway and that these accumulating metabolites serve as metabolic inhibitors
and/or toxins for the affected organ. This in vitro study strongly suggests that the
effect is direct, since there is good correlation between enzyme activities described
in earlier studies and high-energy phosphates, and the correlation continues in the
recovery phase. Certainly, in the present study the good correlation of lactate, pyru-
vate, citrate, alpha-ketoglutarate, isocitrate, ATP, ADP, AMP, and phosphocreatine
lends strong credence to the direct effect hypothesis.
In a study (Takahashi, K. and Nakamura, H., 1976), nine cases of typical beriberi
were studied as regards both cardiac involvement and peripheral neuropathy. The
study emphasizes the fact that in classical beriberi cases, there is a significant neu-
ropathological component.
These patients had eaten a polished rice diet with very little vitamin B1 supple-
mentation. Patients may have presented acutely or subacutely. All showed edema,
apical heart murmurs, accentuated pulmonary sounds, and cardiac enlargement.
Blood lactate levels were elevated, while pyruvate levels were normal. Blood thi-
amine levels were below that of controls. Following treatment with thiamine, blood
values, cardiac pathology, edema, and neuropathy returned toward normal in a few
weeks.
Beriberi 43
During the pretreatment phase, sural nerve biopsy was performed in some
patients, and in some other patients, biopsies were performed after treatment was
begun.
Results showed a loss of myelinated fibers and an increase in collagen fibers
were noted in the sural nerve. Fragmented myelin forms were seen in methylene
blue stained sections. These changes were correlated with leg sensory loss. In elec-
tron microscopy sections, flattened vesicular forms were seen in axoplasm. Axonyl
degeneration was noted in all cases examined. Also seen were degenerated mito-
chondria, vesicles, small myelin figures, and deranged Schwann cell cytoplasm. The
“jelly role” appearance of myelin was frequently preserved suggesting the primary
defect was axonyl degeneration.
In patients receiving thiamine treatment, many axons were seen via electron
microscopy. These axons were not always surrounded by Schwann cytoplasm. No
regenerating axons were seen in untreated patients’ sural nerves. In two patients,
signs of remyelination were evident, although 91 and 30 days elapsed between treat-
ment and biopsy. The improvement of the scene observed with electron microscopy
in treated patients correlated well with improvement in all signs and symptoms.
This study represents the first to look at neuropathy in well-documented beriberi
patients. The light and electron microscopy changes in the sural nerves corre-
late well with similar changes in thiamine-deficient rats. Changes seen in the two
patients showing remyelination were noted to be similar to other patients undergo-
ing repeated myelination and remyelination. Large myelinated fibers seemed to be
more affected in these patients than unmyelinated fibers. Since thiamine plays such
an important role (thiamine pyrophosphate) in oxidative metabolism and energy for-
mation, large fibers may be more sensitive to depletion of thiamine.
In another study, hemodynamic studies were carried out in four alcoholic patients
with beriberi heart disease who did not have complicating cirrhosis, anemia, or other
known heart disease (Akbarian, M., 1966). All four were chronic alcoholics with
poor diet, and they all showed varying degrees of peripheral neuropathy. These four
cases were well established by both clinical and hemodynamic data, and by the rapid
response to thiamine treatment. All showed biventricular congestive heart failure
and enlargement with pulmonary congestion. One patient showed atrial fibrillation
and another patient had a complete left bundle branch block. These changes have
been seen in other cases of beriberi heart disease and in thiamine-deficient rats. In
all four cases, the presenting cardiac abnormality was high output failure. There was
low vascular resistance with low left ventricular filling. The cardiac output was dis-
proportionally elevated compared to the metabolic rate. Blood pressure was elevated
in two of the four cases. There were associated high plasma and blood volumes in
these patients. Clinically, the patients were confirmed as having left ventricular fail-
ure.
Treatment of the patients resulted in measurable improvement in only 37 min-
utes. In this case peripheral vascular tone and decreased left ventricular work
resulted in low cardiac failure. Three patients received long-term thiamine treat-
ment and had marked improvement clinically and physiologically. An increase in
vascular resistance led to a decrease in heart rate, stroke volume, cardiac output,
and oxygen consumption. The authors state that in addition to thiamine treatment,
patients received overall good nutrition, bed rest, diuretics, etc. Patients are often
treated as above, and controlled clinical studies looking at thiamine alone without
bed rest, good food, etc., should be done in relation to beriberi heart disease. This
type study has been done in terms of neuropathology in chronic alcoholics (see
Wernicke’s disease chapter).
It is noted by the authors that hard physical work acts to precipitate beriberi heart
disease due to maintenance of a high cardiac work load. The peripheral neuritis
so often seen in thiamine-deficient patients acts to prevent physical work, thereby
lowering the incidence of beriberi heart disease in those patients. In keeping with
that is the observation that adrenalin, which acts as a stimulator of cardiac output,
can produce symptoms associated with beriberi heart disease (Shimazono, J., 1927).
This was used as a “test” for beriberi disease in the clinical evaluation of patients
suspected of beriberi. It is also noted that a tropical (hot) climate may exacerbate the
development of beriberi. This explains the geographical location of high incidence
of the disorder. The tropical climate is known to result in vasodilation and a high
output state.
The strong link between the decarboxylation of pyruvate and cardiac muscle
work has been demonstrated (Olson, R. and Piatnek, D., 1951). It was of interest to
determine if the link between the decarboxylation of pyruvate and cardiac myofib-
rils was accomplished with or without the presence of mitochondria (DeArce, H.
et al., 1966). In this study, glycerinated cardiac myofibrils were prepared from dog
heart, separated from mitochondria, then incubated at various time intervals in the
presence or absence of pyruvate, NAD, ATP, etc., and contractility assessed.
Results from these studies showed that the decarboxylation of pyruvate was
linked to the contraction of myofibrils from the heart. This contraction occurred
in the absence of mitochondria. The contraction was thiamine and calcium depen-
dent. The separation of the carboxyl group from pyruvate provides energy that can
be used for myofibril contraction. It appears that ATP is responsible for the tension
aspect of cardiac myofibril contraction. It is possible that more than one source of
energy is available for normal cardiac myofibril contraction. The close link of this
process to thiamine availability explains how in beriberi heart disease, the absence
of thiamine in the diet can lead to fatal cardiac pathology.
Early studies emphasized the vasculature aspect of beriberi, supposing that car-
diac changes such as edema, fatty degeneration, and cloudy vacuolization of myofib-
rils, were secondary and were a result of increased work load and high output due to
peripheral vascular pathology. However a case with clear evidence of beriberi heart
disease was described with left ventricular pathology (Attas, M. et al., 1978).
This patient was a chronic alcoholic with significant cardiac signs and symptoms.
Symptoms included a grade 4/6 systolic ejection murmur, and a grade 2/6 diastolic
sound at the left sternal border. An EKG showed sinus tachycardia and non-specific
T wave changes. Chest X-ray showed cardiac enlargement as compared to earlier
chest X-rays. The patient soon showed a worsening biventricular failure. A diagno-
sis of severe acute (Shoshin type) beriberi was made, and 200 mg IV thiamine was
administered. As usual, the patient responded rapidly to thiamine treatment.
Beriberi 45
The authors state that in this case there was a left cardiac pathology that was not
secondary to vasculature alterations. They also state that the response to thiamine
lags behind the vasculature response, leading to low cardiac output. The impair-
ment of myocardial function could result from alterations including impaired car-
diac energy metabolism, altered coronary blood flow, and/or the presence of other
diseases known to have a cardiac component. Significant evidence for an alteration
in cardiac energy metabolism has already been described (McCandless et al., 1970).
The case described is the only reported case of measured left ventricular function
to date. Upon thiamine treatment, vascular resistance increased, with only trivial
decrease in ventricular ejection. Therapy therefore resulted in a higher stroke work
index, and a lower left ventricular end-diastolic pressure. This is seen as strong
evidence for a depression in myofibril contractility during the acute phase of this
patient’s beriberi heart disease. This patient also developed severe lactic acidosis
and an elevation of pyruvate. Elevation of these two metabolites is associated with
beriberi. The control of lactic acidosis in this patient was complicated until thiamine
treatment was initiated, emphasizing nutritional etiology of this disorder.
Not all cases of beriberi respond to thiamine treatment. In one case a chronic
alcoholic was admitted to a hospital with dyspnea, paresthesia in the lower extremi-
ties, edema, ascites, and sensory disturbances (Betrosian, A. et al., 2004). The EKG
was consistent with hypokinetic heart failure with ejection fraction less than 40%.
Blood thiamine was less than half normal. By the third day the diagnosis of beriberi
was made and thiamine was administered (100 mg IM/day). Some improvement was
noted, but on the sixth day the patient expired. At autopsy the heart was small, and
the myocardium was thin and atrophic. Microscopically, the heart revealed extensive
fibrosis and edema, hyaline degeneration, and asbestosis. Peripheral nerves showed
axonyl degeneration and myelin pathology.
The authors note that this is a case showing a “mixed” state, which included high
output cardiac failure and polyneuropathy. The patient died despite prompt initiation
of thiamine therapy. The authors speculate that the initial dose of thiamine might
have met the needs of the peripheral nervous system, but not that of the cardio-
vascular system. The patient developed low cardiac output failure after treatment,
which resulted in his demise. Clinicians should be conscious of the possibilities of
this complication to thiamine deficiency and beriberi.
In another single case study, a 24-year-old pregnant patient with hyperemesis
gravidarum was described (Indraccolo, U. et al., 2005). This patient claimed not
to have consumed alcohol, but had been vomiting for at least 40 days and could
eat only very small amounts of food. She was IV rehydrated and received IV glu-
cose and antiemetics. She developed confusion, horizontal nystagmus, aphasia, and
ataxia despite treatment. She lapsed into coma, and thiamine was administered.
The diagnosis of Wernicke’s disease was made in addition to the presence of wet
beriberi. After 5 days of thiamine therapy the coma was gone, and the patient could
partly feed herself. She continued to improve and gave birth to a normal baby.
This case of thiamine deficiency due to hyperemesis gravidarum had characteris-
tics of both wet beriberi and dry (neuropathy) beriberi. Previous cases of hypereme-
sis gravidarum in pregnancy have been treated with thiamine doses ranging from 50
to 200 mg of thiamine per day. Doses over 400 mg/day IV are accompanied with
risks such as nausia, anorexia, and ataxia. The pregnancy may be associated with
an increased thiamine requirement. The dose used in this study, 300 mg of thiamine
per day, then reduced to 200 mg/day when symptoms start to improve (fifth day),
could prove beneficial in such cases.
In another case report (Tran, H., 2006), an elderly woman with chronic alco-
holism was admitted with dyspnea. She had bilateral crepitations with chest auscul-
tations, and a chest radiograph showed cardiomegaly and frank pulmonary edema.
Blood thiamine was decreased 90% and transketolase activity also decreased. A
diagnosis of wet beriberi with cardiac failure was made, and thiamine therapy was
initiated. Thiamine treatment resulted in improvement of clinical signs in the fol-
lowing 2–3 days. The patient received regular thiamine supplementation and was
discharged.
This case shows an under-diagnosed cause of congestive heart failure – wet
beriberi and thiamine deficiency. Thiamine deficiency can be confirmed by blood
thiamine determination and transketolase measurement, but these assays may not be
available in many hospitals. Therefore, the diagnosis may have to be based on clini-
cal suspicion, and the treatment empirical. In this case, thiamine treatment reversed
the decreased left ventricular function.
Wernicke’s Disease
The clinical entity Wernicke’s disease was first described by Dr. Carl Wernicke in
1881. In this manuscript, Wernicke described both clinical features and autopsy find-
ings (Wernicke, 1881). The description was of three patients, and clinical features
consisted of ataxia, strabismus, nystagmus, diplopia, disorientation, and stupor. The
findings at autopsy were similar and consisted of small hemorrhages located in peri-
aqueductal gray. Wernicke interpreted this to be an acute inflammatory disease pri-
marily of the ocular nuclei.
In 1887, Korsakoff published a paper describing a unique amnesic state that
occurred in patients who were both alcoholic and non-alcoholic (Korsakoff, 1887).
Korsakoff made the correlation between the alcoholic neuritis of Wernicke’s disease,
and the memory defect and confabulation of Korsakoff’s psychosis. He considered
that these were two facets of the same disorder (Korsakoff, 1889). Korsakoff also
realized that these two features of the same disease might manifest in different pro-
portions in different patients.
In another manuscript, Korsakoff (1889) describes the amnesia of this disorder
as being of recent events, with a relative sparing of memory for more distant past
events. Korsakoff notes that this type of amnesia occurs after prodromal agitation
and confusion. The confusion may last several days, then resolves, but the amnesia
remains. There exists a broad range of manifestations of these symptoms. Some-
times distant memories are also affected. Recent memories may not be clear, for
example, a patient not being able to remember whether or not dinner had been taken.
Another example is that a patient might say that he went to town yesterday, when in
fact he was bedridden.
Pathological changes in Korsakoff’s psychosis were first noted in 1896 (Gudden,
et al., 1896). Brain lesions were described in the mammillary bodies, as well as in
the periaqueductal gray. In a later neuropathological description (Gamper, 1928),
the similarity between lesions in Wernicke’s disease and Korsakoff’s psychosis was
noted. Another (Kant, 1932) investigator found brain lesions of Wernicke’s dis-
ease in all of his fatal cases of Korsakoff’s psychosis. Several subsequent stud-
ies described brain lesions which were very similar in both Wernicke’s disease
and Korsakoff’s psychosis and included hemorrhagic changes in the periaqueduc-
tal gray, mammillary bodies, brain stem nuclei, and thalamus. These lesions could


explain the symptoms. It was Girard et al. (1956) who drew the conclusion that Wer-
nicke’s disease and Korsakoff’s psychosis were similar in symptoms, neuropathol-
ogy, and outcome. From this observation, the two disorders came to be called the
Wernicke–Korsakoff syndrome. As time progressed reports of large (70 cases) stud-
ies of patients diagnosed with Wernicke’s disease and/or Korsakoff’s psychosis were
published. These confirmed earlier descriptions of symptoms and location of neu-
ropathological lesions. Studies of Korsakoff’s psychosis patients described lesions
identical in nature and location as those seen in Wernicke’s disease. These authors
all suggested that this represented variable clinical expression and outcome of the
same disease.
It should be pointed out that there was some variation in different studies. Some
authors, for example, found changes in neurons of the cerebral cortex, while oth-
ers did not. Another example is that many neuropathologic studies did not men-
tion cerebellar changes, whereas many recent studies have shown characteristic
cerebellar changes (Figs. 1, 2 and 3).
Wernicke’s Disease—Gross Brain Lesions
Source Mammillary Thalamus Cerebellum Cerebral Periaqueductal Hypothalamus

Bodies Cortex Gray
Victor
Adams & + + + – +
Collins
Malamud &
Skillicorn + + + –
(Korsakoff’s)
Liu, et.al.
(MRI Study) + + +
Jagadha, er.al.
(non alcoholic + + +
Wernicke’s)
Kornreich, et.al.
(infant Wernicke’s) + + +
Fig. 1
Wernicke’s Disease—Microscopic Lesions
Source Thalamus Hypothalamus Cerebellum Mammillary Bodies Pons Medulla Fornix
Victor,
Adams, & + + + + + + +
Collins
Fig. 2
Specific neuropathological changes have been described in Wernicke’s dis-

ease (Victor et al., 1989). These observations resulted from the neuropathological
Wernicke’s Disease 49
Wernicke”s Disease—Type of Microscopic Lesions
Source Cell Necrosis Decreased Increased Hemorrhage Loss of

Neurons Astrocytes Myelin
Fibers
Victor
Adams & + + + + +
Collins
Malamud & + + + + +
Skillicorn
Jagadha,
Et.al., + + + + +
(nonalcoholic
Wernicke’s)
Fig. 3
examination of 62 cases of Wernicke–Korsakoff syndrome over several years. In

these studies, the surface features of the brains were normal, as was the weight.
Coronal sections were examined grossly and showed a symmetrical increase in
the size of the lateral ventricles (26% of cases). The thalamus showed characteristic
and symmetric areas of gray-brown discoloration in about 75% of cases. In addition,
these lesions were described as spongy or granular in appearance. Hemorrhages
were not frequently (less than 10%) seen grossly, and when present were usually
in the mammillary bodies. Lesions were observed in the midbrain, as well as in
the floor of the fourth ventricle, confined to the medulla. The cerebellum showed
gross lesions in 36% of cases examined and was described by the authors as an
atrophy of the folia of the vermis. The bilateral symmetry of the lesions was again
noted (Fig. 4).
Fig. 4 Schematic cartoon at the level of the pons, showing the cerebellum
Microscopically, lesions from this study were noted to have a symmetric consis-
tent location in the thalamus, hypothalamus, midbrain, pons, and medulla. Lesions
were also noted in the cerebellum, as well as, to a lesser degree, in the cerebral
cortex, hippocampus, and fornix.
The actual microscopic lesions showed variation due to the duration and course
of the disease process. The most severe extent of the microscopic lesion was nearly
complete cell necrosis in affected areas. Where lesions were not so severe both
neurons and myelin were mostly absent. The most common degree of change con-
sisted of some decrease in neurons, but many were spared. In this state, myelin
was destroyed more than neurons. Also present was an increase in astrocytes and
macrophages. Focal areas of hemorrhage were only occasionally seen. Blood vessel
changes were a common and prominent change in affected areas, but were consid-
ered by these authors to be of a secondary nature. In some cases macrophages con-
taining hemosiderin, indicating previous bleeding, were seen. These microscopic
lesions were bilaterally symmetric.
Generally speaking, within certain damaged areas, such as the parafascicular
nuclei of the thalamus, neurons were spared. There was a highly regional distri-
bution of damage, for example, the dorsal motor nucleus of the vagus nerve was
damaged in the upper medulla, whereas in the lower medulla, the vagus nucleus
was unaffected. Most areas affected with microscopic lesions were gray matter. In
all lesioned areas with damaged neurons, the myelinated fibers of the nerve cells
seemed more affected than did the cell bodies.
Vascular and glial changes in lesioned areas were quite similar from case to case.
The changes seen in neuropathology in the Wernicke–Korsakoff syndrome are dif-
ferent from those seen in either ischemia or edema. The relative sparing of neurons
seemed to these authors to suggest a metabolic alteration. The exact reason under-
lying the unique specificity of the lesion pattern is unclear, but certainly this also
occurs in Leigh’s disease. The peripheral nerve changes in Wernicke’s disease are
also identical to those seen in beri-beri.
In another comprehensive study (Malamud and Skillicorn, 1956), the relation-
ship between Wernicke’s disease and Korsakoff’s psychosis was studied. This was a
clinical and neuropathological study of 70 patients in mental hospitals who had been
diagnosed with Korsakoff’s psychosis. The clinical progress was followed through-
out hospitalization, and autopsies were performed upon death.
Autopsies revealed a variety of diseases such as cardiovascular, cirrhosis, carci-
noma, etc., consistent with the mean age at death (61.6 years). Gross examination of
the brains revealed significantly small mammillary bodies with a yellow-brown dis-
coloration. In addition, the ventricular system showed dilatation in a small percent
of cases.
The neurohistopathologic findings were revealing. The hypothalamic and thala-
mic periventricular areas had a diffuse gliosis in all cases using a Holzer stain. The
mammillary bodies showed bilateral atrophy and gliosis in all but three cases. In
most of these, the alteration was a chronic response of astrocytes and glial fibers.
While hemorrhages were rare, there were frequently proliferative changes in the
endothelium of blood vessels and an increase in reticular fibers. In some instances,
there was a focal necrosis, with decreased numbers of neurons. The medial thala-
mus showed degenerative changes in the dorsomedial nuclei in over one half of the
patients, where as the pulvinar was affected in only three cases. In these lesions,
unlike those in the mammillary bodies, the neurons had nearly all disappeared and
were replaced by glial fibers and astrocytes (Figs. 5 and 6).
Fig. 5 Schematic cartoon at the level of the mammillary bodies
Fig. 6 Brain slices at the level of the mammillary bodies. Normal on left; Wernicke’s disease on
right. Figure courtesy of Dr. Roland Auer (see acknowledgments)
In the brainstem, there was gliosis in the periaqueductal areas in over one half
of cases, and similar changes in nuclear areas such as the occulomotor nuclei, infe-
rior olives, ventricular nuclei, etc. In the cerebellum, degenerative changes were
present in one-third of the cases. Changes were located in the Purkinje cell layer,
wherein Purkinje cells were absent. There was an increase in glial cells in the molec-
ular layer (Bergmann) of the cerebellum, and marked gliosis of white matter. These
degenerative changes were most pronounced in the vermis.
The cerebral cortex showed no specific alterations. There were changes in a
small number of cases which included a diffuse proliferation of large cells called
Alzheimer’s glia. Several other cases had diffuse neuronyl involvement including,
swelling, central chromatolysis, and nissl bodies concentrated at the periphery of the
cells. The spinal cord showed pathology in five of 12 cases examined. These cases
had diminished numbers of anterior horn cells. Those present had axonyl changes,
which were characteristic of peripheral neuritis.
This study of clinical and pathological changes in the Wernicke–Korsakoff syn-
drome represents the second and largest such study. The pathological changes con-
sisting of lesions localized in the periventricular gray matter, mammillary bodies,
and medial regions of the thalamus (dorsomedial nuclei) are nearly identical to
the findings reported much earlier in the only other comprehensive study (Gamper,
1928). Intracerebral hemorrhage was rarely found in either study (Figs. 7, 8, 9, 10,
and 11).
Fig. 7 Close-up of Wernicke’s disease mammillary bodies in Fig. 6. Figure courtesy of Dr. Roland
Auer
The authors conclude that the lesions seen in their 70 cases of Korsakoff’s psy-
chosis correspond exactly with lesions described in Wernicke’s disease. This sug-
gests the two disorders are identical, except that one is acute, the other chronic,
and that they represent diseases that result largely from thiamine deficiency. Since
Fig. 8 Drawing of cresyl violet stain section through the mediodorsal thalamic nucleus: control
Fig. 9 Wernicke’s brain slice at the level of the mammillary bodies. Note lesion bilaterally sym-
metrical. Figure courtesy of Dr. Michael Norenberg (see acknowledgments)
autopsy data are increasingly difficult to obtain, recent neuropathological data can,
in some measure, be obtained from MRI studies, and these studies do show a
good correlation with autopsy gathered data (Liu et al., 2006). These MRI stud-
ies show bilaterally symmetric lesions in paraventricular regions, hypothalamus and
thalamus, mammillary bodies, and periaqueductal areas of the midbrain (Doherty
et al., 2002). It should be remembered that neuroimaging studies may not always be
Fig. 10 Wernicke’s photomicrograph showing spongy degeneration and presence of blood. Figure
coutersy of Dr. Michael Norenberg
Fig. 11 Brain slice showing mediodorsal thalamus in Wernicke’s disease. Figure courtesy of
Dr. Jeffrey Joseph (see acknowledgments)
definitive and that they may miss some lesions because of resolution considerations,
and because early pathology is biochemical in nature (Celik and Kaya, 2004).
The hemorrhagic component of Wernicke’s disease has been examined (Rosen-
blum and Feigin, 1965). These authors looked at 43 cases they had seen, and two
of these had large intracerebral hemorrhages in the vicinity of the third and fourth
ventricles. These two cases were classical Wernicke’s disease in terms of clinical
presentation and other neuropathological findings. The exception was that large
(6–8 mm) hemorrhages were located on the side walls of the third ventricle and
in the floor of the fourth ventricle. Smaller petechial hemorrhages were located in
close proximity. Microscopically, there was vascular proliferation and astrocytosis.
The hemorrhagic areas appeared as large lakes of red blood cells. In both cases,
the mammillary bodies were affected in the classic way – degeneration including
rarefaction, and infiltration of macrophages.
Clinical findings in this paper are of interest in that they showed 90% of patients
were chronic alcoholics, and all showed memory impairment and confusion. One-
third or more showed ataxia, peripheral neuritis, and findings suggestive of chronic
Wernicke’s disease. In one case there was a clinical picture which was mixed, with
symptoms of both Wernicke’s and Korsakoff’s diseases. Another case was one of
an acute attack of Wernicke’s disease which left the patient with loss of memory,
disorientation, and confabulation, which lasted until his death, and was reminiscent
of Korsakoff’s psychosis. These cases show the “overlap” of clinical symptoms,
and of pathological lesions. Of the 70 cases, the authors state that 31% showed, in
addition to Korsakoff’s psychosis, symptoms of a complete or partial Wernicke’s
syndrome. The pathogenesis of the mental symptoms is not well understood; how-
ever, the lesions in the hypothalamus and thalamus have been implicated (Hess and
Akert, 1955) as involved. While intellectual functioning occurs in the cerebral cor-
tex, these activities may be integrated in the limbic system, which includes the
thalamus and hypothalamus. This could explain many symptoms described in the
Wernicke–Korsakoff syndrome.
In the large study of Wernicke–Korsakoff syndrome described earlier (Victor
et al., 1989), mental confusion was the most common presenting symptom in two-
thirds of the cases (108 of 163). Ataxia was present in over 50% of cases and ocular
symptoms were present in 40% of cases. Other symptoms included memory loss,
polyneuropathy, exhaustion, etc.
Findings on physical exam in 245 patients with Wernicke–Korsakoff syndrome
included tachycardia (51%), fatty hepatosis (49%), redness of tongue (36%), skin
and mucous membrane disorders (29%), etc. Hypothermia was observed in only two
of the patients in this series; however, a higher incidence of hypothermia has been
noted in other reports (Harper, 1979) (Wallis et al., 1978).
Findings related to brain lesions consisted of confusion, defective memory, ocu-
lar abnormalities, and ataxia. In this series, confusion could be accurately assessed
in 229 patients. Ten percent of these 229 exhibited no alteration in mentation and
lacked any confusion. The remainder of patients showed significant changes in con-
sciousness and in mentation.
In this study, 16% presented with symptoms of delerium tremons. These symp-
toms were usually mild, but clearly related to alcohol withdrawal. Fourteen per-
cent of patients showed little or no confusion or agitation when first examined. The
patients did have severe memory and learning disorders characteristic of Korsakoff’s
psychosis. The majority of patients (56%) had mental confabulation and disorienta-
tion that the authors termed “global confusional state.” The major characteristics of
this were apathy, drowsiness, memory loss, and disorientation. These patients were
inert and indifferent to people, questions, and their surroundings. Other symptoms
included memory disorder, alcohol abstinence state, stupor, and coma.
Ocular symptoms occurred in 96% of patients. These symptoms consisted of nys-

tagmus, lateral rectus palsy, and paralysis of conjugate gaze. Each patient had more
or less frequency of these symptoms; only 27% had all three symptoms present.
Other ocular symptoms were uncommon. A sluggish reaction of the pupils to light
occurred in 19%, yet it was relatively mild.
Nystagmus as a presenting symptom was noted in 85% of patients. The most
common type was horizontal nystagmus to both sides. This was seen in 97% of
patients with nystagmus. One half of patients with horizontal nystagmus showed
vertical nystagmus. Treatment of nystagmus with thiamine cleared nystagmus in as
little as 3 days. In 54% of cases, a lateral rectus palsy was present on first exami-
nation. The majority of these patients had incomplete lateral rectus palsy. Weakness
of conjugate gaze was present in 44% of patients examined. Eighty-six percent of
those with gaze palsies presented with a horizontal plane gaze palsy.
Of the patients examined initially, only 13% showed no alteration in ability to
walk or stand. At the other end of the spectrum, 21% of patients could neither walk
nor stand unaided. These patients took uncertain lurching type steps and wanted
only to return to bed. Mild ataxia consisted of only being noticeable when the patient
was asked to perform tasks such as heel to toe walking. A characteristic tremor of
the lower extremity was noted, as was an incoordination of the upper extremity. A
few patients showed a combination of ataxic gait and ataxia of both upper and lower
extremities (Figs. 12, 13, and 14).
Fig. 12 Cerebellum in Wernicke’s disease. Figure courtesy of Dr. Roland Auer

Fig. 13 Photomicrograph of normal control cerebellar layers showing Purkinje cells and
Bergmann glia. Figure courtesy of Dr. Roland Auer
Fig. 14 Photomicrograph of cerebellar layers – note the absence of Purkinje cells. Figure courtesy
of Dr. Roland Auer
In 82% of patients, polyneuropathy was present. The legs were affected in most
cases, but in some, both upper and lower extremities were affected. In symptomatic
cases, complaints consisted of weakness, pain, and paresthesias. Symptoms were
usually noted first in distal portions of the extremities and then progressed prox-
imally. In some cases, the polyneuropathy was notable only on physical exami-
nation. The exam found tender leg muscles, decreased ankle jerks, and in some
instances, loss or decrease in knee-jerk reaction. In a few cases, the upper extrem-
ities were affected as well. These patients had loss of tendon reflexes and loss of
motor strength. Muscle weakness often took the form of foot and/or hand drop. Not
only were distal muscles affected but proximal muscles were also weakened. Total
paralysis was rarely observed. Tenderness to pressure was almost always noted, usu-
ally in the calves and feet. Muscle weakness in the legs and feet were almost always
associated with diminished knee and/or ankle jerks. Edema and thinness of skin
were common findings over the lower legs and feet in severe cases of polyneuropa-
thy in the lower extremity.
In terms of cranial neuropathy, except for occulomotor and vestibular function,
other cranial nerve involvement does not occur. Bulbar palsy also does not occur in
the Wernicke–Korsakoff syndrome. A few cases of dysphonia and dysphagia have
been described (Novak and Victor, 1974). These were cases where there was degen-
eration of the vagus nerve, and in which the patients also had a severe nutritional
deficiency component in their history. For a recent review, see Kashi, M. et al.,
(2009).
Abnormal laboratory findings were also noted at hospital admission. Abnor-
mal liver function tests were present in 43% of patients, and anemia was found
in 39%. Other laboratory chemistry abnormalities included leukocytosis, elevated
BUN, hyperglycemia, and hypovalemia. EEG results were available in a low number
of cases, and in half, there was a diffuse decrease in the frequency of brain waves.
This change was only mild to moderate. Remarkable was the finding that in spite of
obvious neurological involvement, half of those tested had normal EEGs. Some of
these “normal EEG” patients had extensive diencephalic lesions, and reduced cere-
bral blood flow, with reduced O2 and glucose consumption. Examination of CSF
revealed elevated protein levels in 23% of patients. Other than protein levels, all
CSF values were within normal values.
As stated above, recovery following thiamine treatment produced a rapid recov-
ery of function of some deficits. In the case of abducens nerve palsy, recovery of
function was noted as soon as 6 hours after thiamine administration. In another
group, recovery was noted in less than a day, and in all cases, improvement was
seen in less than 4 days. Complete recovery had occurred in most patients in less
than 7 days. Vertical gaze palsy showed improvement within 1–24 hours in half of
the patients. Other ocular symptoms showed similar modes of recovery.
Ataxias were somewhat slower to show recovery. In most cases of ataxia that
could be observed properly, recovery was noted in 2–6 days. In cases where recovery
could be determined, complete recovery occurred in less than 1 month in almost
half of the cases. Most recovered cases occurred in less than 2 months, but only
38% showed complete recovery. No significant improvement in ataxia occurred in
27% of patients. The remaining group made an incomplete recovery, plateauing at

a stage of lessened ataxia. These improvements were usually completed by the end
of the second month.
Thiamine administration had a positive effect on the recovery from the con-
fusion and confabulation state in basically all patients who were properly exam-
ined. The onset of improvement in mental symptoms occurred by 17 days after
thiamine administration. In some few cases, complete recovery from the confused
state occurred in only 6 days; others took up to 2 months. In patients with amnesia in
Korsakoff’s psychosis, about 25% in each of four groups showed no recovery, slight
recovery, significant recovery, or complete recovery. The mechanism by which some
recovered completely, and others not at all, is unclear, but may be related to the
length of time the patient had been sick. It should be noted that, in this study, 10%
of patients had no mental symptoms. Also, in about 10% stupor/coma was present,
making further assessment difficult.
Well-controlled treatment studies have been conducted in some measure because
of the large number of available Wernicke-Korsakoff patients. In one such study
(Phillips et al., 1952), 51 patients were studied, and 29 served as well-fed, well-
nourished control subjects. Experimental subjects received no food or nourishment
until their clinical status had been carefully assessed. The experimental patients
were divided into four groups: (1) glucose plus salt water, (2) an unfortified rice
diet, (3) a vitamin-deficient diet, and (4) a group who received the above three plus
whiskey.
Results showed that in the first group, there was no improvement in the clin-
ical picture until thiamine was added to the diet. The addition of a full supple-
ment of vitamins excluding thiamine had no effect. In the second group (unfor-
tified rice diet) symptoms remained unchanged until thiamine alone was added.
The third group also had no improvement. A subgroup of these patients was also
given whiskey, and after thiamine administration, they also improved. The memory
loss was slow to recover compared to recovery from nystagmus, but improvement
of memory was evident by 8 weeks in most patients. This is in keeping with the
slower recovery of mentation seen in nearly all patients with Wernicke–Korsakoff
syndrome.
These studies (Phillips et al., 1952) establish beyond doubt that the symptoms
of Wernicke–Korsakoff syndrome can be directly related to thiamine deficiency.
Similar studies by others (Victor and Adams, 1961) confirm these findings. The
studies emphasize the importance of instituting thiamine therapy as soon as pos-
sible in order to decrease the biochemical lesion, and prevent as much structural
damage as is possible. In some cases in which unsupplemented glucose is given,
without thiamine, Wernicke’s symptoms may be precipitated in a very short time.
Patients in whom thiamine is greatly reduced may be pushed into overt disease by
a high-carbohydrate diet, thereby depleting whatever residual amount of thiamine
was present.
In Korsakoff’s psychosis, symptoms are slower to recover and may be more
apparent when Wernicke’s symptoms clear. Weeks to months may elapse before
Korsakoff’s psychosis symptoms begin to improve, but eventually most patients
with Korsakoff’s do have some improvement. As many as 21% will make a complete
recovery, and a higher rate has been recorded in other studies (Rosenbaum and
Merritt, 1939).
In the early stages of the Wernicke-Korsakoff disorder, treatment with thiamine
can show almost immediate benefits, emphasizing the biochemical lesion character-
istic of this disease. The actual dose, and length of treatment of patients, however,
is not established (Morcos, 2004). The usual dose is of an initial load of 100 mg IV,
followed by a maintenance dose of between 50–100 mg by mouth until the patient
is eating well (Lacasse and Lum, 2004).
In a paper (Grunnet, 1969), changing patterns of incidence, distribution, and neu-
ropathology in Wernicke’s patients were examined. Results showed that the time
of death seemed to have lengthened over a period of 25 years, probably due to
improved health care. There appeared to be an increase in the numbers of cases, pos-
sibly due to increased awareness and availability of health care. Finally, the authors
describe a somewhat more chronic form of Wernicke’s disease in which the lesions
were more widespread, and to some degree, the lesions were more at a microscopic
level than at a gross level.
Another study examining the incidence of Wernicke’s disease in Australia in
131 cases has been published (Harper, 1983). This study showed that the Australia
incidence was (2.8%) higher than any other country. The exact reasons for this are
unclear; however, the author may have been more efficient in looking at autopsy
material. He points out that many cases are missed at autopsy and that clinical
features may not always be obvious. In this study, both the gross and histological
changes were quite similar to those described previously (Victor et al., 1989).
In a neurochemical study, (McEntee et al., 1987) the authors had noted that the
lesions of Korsakoff’s psychosis corresponded with the location of monoamine con-
taining neurons. The lesions and monoamine containing neurons are both located in
periventricular and periaqueductal regions of the brain. A total of 25 Korsakoff’s
patients were studied. CSF was obtained from these patients and controls, and 3-
methyl-4-hydroxyl phenylglycol, homovanillic acid, and 5-hydroxyindolacetic acid
were measured. Psychometric analysis just prior to obtaining CSF showed a nor-
mal intelligence, but a significantly impaired memory function. The CSF levels of
the three compounds measured were statistically lower than that of controls. These
compounds are a measure of cerebral noradrenergic activity. The compounds are
also altered in disorders such as Parkinson’s disease and Alzheimer’s disease. The
possibility exists that these other diseases may have been present in some patients,
contributing to the results. However, this study shows a possible link between dam-
age to noradrenergic neurons and the memory losses seen in Korsakoff’s psychosis.
In a second study (McEntee et al., 1987), the CSF concentrations of nore-
pinephrine, dopamine, and serotonin were examined in a group of Korsakoff’s
patients. Measurements were made and correlated with the learning of two motor
tasks. Results showed that improvement in learning correlated with CSF levels of
dopamine metabolites. These studies indicate that dopamine might be involved in
learning motor tasks and that a depletion of CSF dopamine metabolites could hinder
learning motor tasks in patients with Korsakoff’s psychosis.
In a retrospective study of clinical signs (Harper et al., 1986), careful analysis

of signs initially noted in the Wernicke–Korsakoff syndrome are discussed. These
authors note that in their study, if the criteria previously described were strictly
adhered to, the diagnosis would have been missed.
The authors state that mild cases of Wernicke’s disease surely exist and that a
less rigid criteria should be applied. Attention should also be placed on suspicion
for the diagnosis based on history. It is also noted that many mild episodes may,
over time, produce Wernicke’s disease in a chronic form with a somewhat different
presentation of symptoms. The problem is that if a rigorous standard is applied, then
cases of the Wernicke–Korsakoff syndrome may be missed, and treatment delayed,
or not applied. If more relaxed criteria are utilized, and suspicious cases identified,
then adequate thiamine therapy can be administered. The proof of this awaits a
neuropathologic study in which pathologically confirmed cases are identified who
had minor symptoms and therefore were not identified.
Wernicke’s disease can occur in non-alcoholic patients. This has been described
(Jagadha et al., 1987) in a group of five patients receiving dialysis. In all five, there
was no indication of alcohol abuse, but each had end stage renal disease. In addi-
tion, they were all receiving IV therapy and were very sick. They had repeated dial-
ysis, either peritoneal dialysis or hemodialysis. Infection was another feature, which
could compromise thiamine metabolism.
Upon death, neuropathological studies were performed. Brain examination
revealed lesions characteristic of Wernicke’s disease. These lesions were present
in the inferior olivary nuclei, hypothalamus, thalamus, and mammillary bodies. The
lesions were bilaterally symmetrical. Microscopic examination revealed nerve cell
death, axonyl swelling, gliosis, myelin loss, brainstem vascular proliferation, and
hemorrhages. In one case, the hemorrhages were confluent and prominent. Some
of the microscopic findings were in areas which showed no gross abnormalities,
emphasizing the need for a thorough pathological exam. It should be emphasized
that many factors can modify thiamine metabolism and produce Wernicke’s disease
in very ill patients, such as those undergoing dialysis (Figs. 15 and 16).
The actual levels of activity of key thiamine enzymes have been examined in
the brains of patients with Wernicke–Korsakoff syndrome. In this study (Butter-
worth et al., 1993), control, alcoholics without Wernicke’s, and Wernicke-Korsakoff
patients’ cerebellar vermis were studied. At autopsy, vermis samples were removed
and stored at –80◦ C until analysis. Neuropathological analysis showed that tissue
from nine patients that were alcoholic/non-Wernicke’s did not show evidence of
Wernicke’s disease. The two patients with Wernicke–Korsakoff syndrome had small
mammillary bodies. Microscopic examination showed the usual spongy character-
istic capillary proliferation and gliosis. The gliosis was present in the periventricular
region of the third and fourth ventricles.
Results from biochemical analysis of the three key thiamine-requiring enzymes
in the cerebellar vermis showed decreases in all three enzymes in the Wernicke-
Korsakoff patients. The greatest drop was in the activity of alpha-ketoglutarate
dehydrogenase. The non-thiamine requiring enzyme glutamate dehydrogenase was
unaltered, serving as an internal control for fixation artifact. These data suggest that
Fig. 15 Brain slice showing lesioned mammillary bodies and mediodorsal thalamic nucleus.
Reproduced from Krill (1995) with permission from Springer
significant decreases in thiamine-requiring enzymes may be essential in the disease

process, and their alterations may in turn help explain symptoms in these patients.
In another study examining treatment modalities (Ambrose et al., 2001), thiamine
was administered to various groups of alcoholic-dependent patients. These patients
were undergoing detoxification and were administered thiamine at different dose
levels. This was done without heavy dependence on the classic triad of symptoms:
ataxia, ocular disturbances, and confusion. These strict symptoms, when used, may
exclude some patients with Wernicke’s disease.
The investigators examined patient’s performance on a mini-mental state exam-
ination. Of the various treatment groups, the group receiving the highest dose of
thiamine (200 mg) performed highest. Other factors such as education, years of
drinking, etc., were not significant contributors to exam performance. These data
reinforce the concept that patients with mild, less symptomic forms of Wernicke’s
disease benefit from thiamine therapy. These patients should be actively looked for,
recognized, and treated.
A recent study was undertaken to evaluate MRI findings in light of signs and
symptoms in a group of Wernicke’s patients (Zuccoli, G. et al., 2007). This was a
retrospective study of 14 female and 12 male patients. Inclusion was based on a clin-
ical diagnosis of Wernicke’s disease, and of significant improvement after initiation
of thiamine treatment. All 26 patients had MRI performed, which included long rep-
etition time and short echo time spin-echo sequences, and contrast-enhanced short
repetition images in multiple planes.
Fig. 16 Photomicrographs of lesioned areas shown in Fig. 14: mammillary bodies and mediodor-
sal thalamic nuclei. Reproduced from Krill (1995) with permission from Springer
Results showed that one half of the patients had alcohol abuse, whereas one half
did not. In the non-alcohol abuse group, Wernicke’s was associated with GI tumors.
Other causes included voluntary starvation, anorexia nervosa, and poverty. The most
prevalent symptoms (81%) were changes in the level of consciousness. This ran
a range of mild disorientation to coma. Twenty patients (77%) had ocular symp-
toms, while 14 (50%) patients were ataxic. Only the ataxic patients had a statistical
correlation with the alcoholic group.
Results from MRI studies showed that 23 patients had high signal intensity
changes on long repetition time spin-echo sequences, and eight patients had low sig-
nal intensity changes on short repetition time spin-echo images. Twenty-two patients
had symmetric lesions in the medial thalamic nuclei and periventricular region of
the third ventricle. Fifteen patients had lesions in the mammillary bodies. Only one
patient showed lesions in the cerebellum.
The authors speculate that although the clinical triad of Wernicke’s disease is fre-
quently met (nystagmus, ataxia, and confusion), incomplete neurological presenta-
tion also occurs. The anatomic regions most frequently affected in this retrospective
study were the medial thalami and the periventricullar region of the third ventricle.
The authors suggest that these areas are where the maintenance of osmotic gradients
is closely related to thiamine levels, and so decreased thiamine might be expected
to produce altered physiology, and onset of lesions. Alteration of structure in the

mammillary bodies correlated with the alcohol-abuse group of patients (Fig. 17).
Classical Clinical Findings in Wernicke’s Disease

Nystagmus
Confusion
Ataxia
Fig. 17
Wernicke’s disease can occur in infants and has recently been described in six
infants who were fed a soy formula which had no thiamine content (Kornreich, L.
et al., 2005). These infants all were admitted at about the same time. Blood lactate
was increased, but there was no evidence of infectious agents. MRIs were similar,
and soon a connection between the MRI details and a certain soy formula was made.
After analysis of the formula showed no thiamine, thiamine treatment was initiated
and improvement was noted.
MRI examination showed T2-weighted areas of hyperintensity in the mammil-
lary bodies, periaqueductal gray, thalami, basal ganglia, and brain stem. All lesions
were bilateral and symmetric. These results emphasize that the Wernicke’s diagno-
sis showed lesions in the same general areas as are seen in adult patients, especially
lesions in the mammillary bodies. MR images demonstrated a characteristic lactate
peak.
Central Pontine Myelinolysis, Alcoholic
Cerebellar Degeneration, and
Marchiafava–Bignami Disease
Central Pontine Myelinolysis
Central pontine myelinolysis (CPM) is a distinct disorder with characteristic CNS

neuropathology and associated with thiamine deficiency. This is not a common
disorder and is often associated with some other severe illness independent of
alcoholism.
In one series of cases described in detail (Adams, R. et al., 1959), a patient was
admitted with a long history of chronic alcoholism. Upon admission, the patient
was seriously ill, had been drinking heavily, and was confabulating. He had chest
pain and a cough. Culture revealed Diplococcus pneumoniae type 5. He was given
penicillin, saline, carbohydrate, and thiamine. For 3 days, he improved and then
became tremulous and confused. He again improved and then by day 9, he became
unable to swallow or move his extremities. By day 11 he was flaccid and did not
react to painful stimuli. On day 12, he was responsive to stimuli. He showed no
nystagmus and could open and close his mouth, but could do little else. There were
no deep reflexes at the wrists, knees, or ankles. By day 14, there was improvement
in reflexes, but no other improvement. A chest X-ray showed diffuse mottled density
in both lungs. Respiration became shallow and gasping, and death from respiratory
arrest occurred on the 22nd hospital day.
Neuropathological features included convolutional atrophy and thin lep-
tomeninges. Major cerebral blood vessels were normal. The external surface of the
pons was gray and soft. Cortical parietal lobes were swollen and had congested
vessels.
The pons showed a striking symmetric punctate area upon myelin staining. This
lesion extended through the middle and upper pons, affecting almost the entire basis
pontis. Superiorly, the lesion extended into the tegmentum just to and into the medial
lemnesci. The lesion edges were clearly demarcated, but were irregular. Bundles
of myelinated fibers which entered the lesion became thinned and pale. Scattered
in the lesion were remnants of macrophages containing hemotoxyphilic granules.
The largest macrophages were at the periphery of the lesion. Nerve cells and axis
cylinders were largely unaffected. No oligodendrocytes could be seen in areas of
demyelination. In these areas, macrophages were fat laden. Some astrocytes were


present. Within the lesion, blood vessels seemed engorged. No other lesions or cells
similar to those in the lesion were found elsewhere in the brain. A few other lesions
thought to be inconsequential included those found in the spinal cord columns of
Goll, in the medulla in the floor of the fourth ventricle, and in the dorsal motor
nuclei of the vagus nerve. There were small hemorrhages in one mammillary body
(Figs. 1 and 2).
Fig. 1 Photomicrograph of central pontine myelinosos showing massive demyelination. Figure

coutersy of Dr. Michael Norenberg
In another case, described by the above author, a malnourished alcoholic woman

was admitted to the hospital dehydrated and anemic. Over time she developed
flaccid quadriplegia and paralysis of bulbar and respiratory musculature. The neuro-
logical symptoms continued to get worse, and she expired of respiratory paralysis.
Neuropathologic findings showed a normal external appearance. The basal por-
tion of the pons was soft and gray. Microscopically, lesions were present in the
rostral portion of the basal pontis. The lesion extended into the tegmentum and
involved the medial parts of the medial lemnesci. The lesion was sharply demar-
cated, but irregular. Again, myelinated fiber bundles approaching the lesion were
normal, then quickly thinned out, and became pale. Nerve cells and axis cylinders
were mostly spared in the lesion. Neuron numbers were slightly decreased. Some
neurons were smaller and/or pale. Macrophages were engorged with fat. Oligoden-
drocytes were not seen, while a few large astrocytes were noticed. Minor changes
were seen in other areas, including the anterior horns of the spinal cord, inferior
olivary nucleus, and nucleus ambiguous.
A third case was a malnourished alcoholic 43-year-old man who developed a
flaccid weakness of both legs. He also had advanced pulmonary tuberculosis. On
admission, he had an absence of knee and ankle jerks. He was anorexic, and an X-
ray showed a pneumonic density in the right lung. The patient’s temperature and
pulse remained high, and he expired on the 11th hospital day.
CPM, ACD, and MBD 67
Fig. 2 Photograph of central

pontine myelinolysis showing
demyelination, but with
Sparing of the cortical spinal
tracts. Figure courtesy of Dr.
Michael Norenberg
Neuropathological findings were normal regarding superficial appearance, and

the meninges were also normal. The basal area of the pons was gray. The lesion was
largely restricted to the central area of the pons, but was smaller in size than the
first two cases. There was a sharply demarcated focus of pallor in the basis pontis.
The lesion had many macrophages loaded with granules. The lesion was only about
4 mm in diameter, was sharply demarcated, and within the lesion no normal myelin
sheaths could be seen. Axis cylinders were decreased in numbers. Nerve cells were
normal in appearance and numbers. A few normal astrocytes were seen, and also
some enlarged astrocytes were noted. Engorged capillaries and venules were seen
in the lesion.
A few other alterations were seen in other areas, such as “loosening” of tissue
matrix and an increase in number and size of astrocytes. Neurons and blood ves-
sels were normal in the mammillary bodies. Other areas with changes included the
walls of the third ventricle, medial nucleus of the thalamus, third nerve nuclei, red
nucleus, and in the periaqueductal gray matter. The lesions in this case were consid-
ered characteristic of Wernicke’s disease. This case also had prominent astrocytes
in the cerebral cortex and dentate nucleus.
The authors commented on the remarkable localization and symmetric distribu-
tion of lesions in these three cases. In the first two cases, the majority of the basis
pontis was involved, whereas in case 3 a smaller portion of the basis pontis was
involved. Each case was one of demyelinative nature. Within the area of the lesion,
almost all medulated sheaths were involved. Thus, deyelination was present in all
tracts running through the lesion. In spite of an advanced degree of myelin destruc-
tion, axis cylinders, neurons, and blood vessels were preserved.
Most neurons were preserved, although some were swollen, and in some areas,
neurons seemed to be reduced in numbers. These neural changes were consistent
with those seen in areas with interruption of axis cylinders (axonyl reaction). Blood
vessels were patent in the lesions, but small vessels may have had increased cellu-
larity in the walls. There was a lack of oligodendrocytes, only being found in the
very margins of the lesions. There was an increase in astrocytes, characteristic of
myelin degeneration type lesions. Macrophages were seen containing large refrac-
tive non-hematoxyphilic particles.
The authors state that the lesion must have originated near the center of the pons
based on the appearance of cavitation at the center. Also, the most recent areas of
neuropathology were at the periphery. These lesion features consisted of large pieces
of degenerating myelin, and phagocytes. This recent lesion activity was around the
complete periphery. The authors speculated that the central (oldest) portion of the
lesion was about 6–8 weeks old. At least some of the advancing lesion correlated
with symptoms.
The authors made some comparisons between symptoms and lesions. All three
cases had a history of severe alcoholism lasting many years. Many symptoms could
be attributed to the lesions in the pons. Thus, flaccid quadriplegia, difficulty in
speaking and swallowing, decreased response to painful stimuli, and loss of corneal
responses could all be caused by the lesion. Cortico-bulbar tracts pass through the
pons, and their loss would affect speech and swallowing. Loss of corneal reflexes
would be attributed to trigeminal nerve loss.
The authors reject the idea that this is merely an unusual example of Wernicke’s
disease because of the characteristics of the lesion. In all cases of Wernicke’s, the
mammillary bodies had been affected, as well as the walls of the third ventricle,
peri-aquaductal region, and the floor of the forth ventricle in most cases. Also, his-
tologically, the lesions were more generalized and not limited to demyelination.
The lesions of CPM also differ from those of Marchiafava–Bignami disease in
that the location of lesions is most likely seen in the corpus callosum. Microscop-
ically, the lesions of Marchiafava–Bignami disease are those of demyelination and
some effect on axis cylinders, and the lesion can progress to cavitation. One other
feature of Marchiafava–Bignami disease is the significant role of nutrition, in that
several cases have been confirmed in patients with no history of alcoholism.
The neuropathological features of symmetry and constancy of location seem to
indicate a biochemical base for etiology. This is meant to say that some toxin or a
deficiency of some vital nutrient essential to the well-being of the brain is at the seat
of the disorder. Examples of these include thiamine deficiency, as well as cyanide
poisoning, anoxia, Wilson’s disease, hepatic encephalopathy, and hypoglycemia.
Several years later after the first description of CPM, the rarity of this disorder
was recognized in that only 18 cases had been described in the world, and none were
diagnosed before death (Cole, M. et al., 1964). These authors suggested that malnu-
trition is a critical feature and present two new cases to support that contention.
The first patient had a long term history of alcoholism and irregular dietary habits.
Just before hospital admission, the patient had a sore throat and mouth and facial
swelling. After admission, her condition remained unchanged when she complained
of diplopia. She was found to have weakness of all eye movements. A couple days
later, she choked while eating lunch and expired.
Necropsy showed an acute moderate pulmonary congestion. The brain was
grossly normal with no evidence of arterial disease. Microscopically, the brain
showed rounded and pale Betz cells in the cerebral cortex. The Betz cells seemed
normal in number. The hippocampus, hypothalamus, and mammillary bodies and
midbrain were normal in appearance. However, centrally located in the basalis pon-
tis was a sharply demarcated bilateral lesion with an almost total lack of any myelin
staining fibers. Beyond the periphery of the lesion where the tissue was normal, the
myelin was intact and stained well. Axis cylinders were present within the lesion,
but were poorly stained. Neurons were absent at the center of the lesion, while at
the periphery, neurons were swollen and poorly stained. Nuclei in the reticular for-
mation were rounded, swollen, and had central chromatolysis.
The second case was a man admitted with slurring of speech, diplopia, and weak-
ness. The patient had a history of heavy drinking. Several years earlier he had had
a pulmonary lobectomy for pulmonary tuberculosis. He had a severe drinking spree
just before hospitalization; he then refused to eat. When admitted, he was confused
and dysarthric. Diplococcus pneumonia was cultured from sputum. He was fed IV
and given multivitamins including B1. He continued to deteriorate and expired 1
month after admission.
Upon autopsy, the brain was normal externally. Histology showed normal cere-
bral cortex, cerebellum, and basal ganglia. The mammillary bodies had a decreased
number of neurons along with an increase in microglia and capillary prominence.
Centrally in the basalis pontis was a sharply demarcated symmetric lesion in
which there was an absence of myelin. As in previous cases, the myelin was normal
right up to the lesion’s edge. Neurons were normal in appearance in the lesion, but
there was a decrease in oligodendrocytes.
Both of these two cases were in chronic alcoholics, but in both cases, there was
significant malnutrition. There were neurons in the cerebral cortex and brain stem
consistent with a nutritional deficiency called “central neuritis,” and seen in brains
from pellagra patients. These neuronyl abnormalities have been seen in other cases
of severe prolonged malnutrition.
The authors also point out that in 17 cases of CPM presented by Berry and
Olszewski (Berry, K. and Olszewski, J., 1963), ten had evidence of malnutrition.
Most other patients had evidence of mild malnutrition. Some cases of CPM have
been seen in which there is no evidence of malnutrition. The exact nature of the
malnutrition seen in these cases is not clear, and there is also some variation in
presenting the clinical picture. Symptoms suggestive of brain stem lesions such as
nystagmus, quadriplegia, dysarthra, dysphagia, extraocular palsies, and stupor and
coma should alert clinicians and facilitate diagnosis.
By 1976, and the paper by Tomlinson, B. et al. (1976), the number of cases pub-
lished was 120. Some cases described demyelination in other areas besides the pons.
Several concurrent diseases were described including Wilson disease, amyloidosis,
and liver disease, as well as alcoholism and malnutrition. The paper reviews the
literature and presents two more cases of CPM.
The first case was of a 54-year-old female admitted in a confusional state with a
several-day history of vomiting, perhaps due taking to phenybutazone. Little other
coherent history could be obtained. The patient’s conscious level declined; she
became unresponsive to deep stimuli and had generalized hypertonicity and exag-
gerated reflexes. She continued to deteriorate, became stuporous, and died 5 weeks
after admission.
Necropsy showed normal brain exterior except for slight granularity over the
surface. There was severe demyelination throughout the dorsal pons to the middle
cerebellar peduncles. The area of the lesion appeared gray. There was a bilaterally
symmetric lesion around 1 mm each in the putamen.
Histologically, the main lesion in the pons showed extensive pontine demyelina-
tion. Moderate astrocytic proliferation was present in the lesion, and oligodendro-
cytes were not seen. The lesions microscopically were well delineated. The thalamus
had demyelination, although not as severe as in the pons. Lesions were also located
in the caudate nucleus, claustrum, putamen, hypothalamus, and scattered cortical
areas. White matter in cerebellar folia was demyelinated and exhibited astrocytic
and microglial infiltration.
The second case in this paper was a 57-year-old woman admitted because of
drowsiness. She had had a 3-week period of vomiting. She later developed seizures
and expired. At autopsy she showed terminal bronchial pneumonia and bilateral
suppurative pyelonephritis.
The surface of the brain showed slight convolutional atrophy. Vessels at the base
of the brain appeared normal. On section, a central pontine gray area extended from
the inferior colliculi to the mid pons. Other sites included the cerebral cortex, basal
ganglia, and thalamus.
Microscopically at mid pontine level, there was almost total demyelination.
There were neurons present, but they were reduced in number. There were cavities
of total tissue destruction with debris and a few macrophages. This lesion ceased
abruptly in the lower pons, and the medulla was normal. A few lesion areas were
present in the cerebral cortex, but most neurons were preserved. The cerebellum
also had a few areas of necrosis, along with marked astrocytic infiltration (Fig. 3).
The authors point out that the lesions in other reported cases are strikingly simi-
lar, while the actual size of the lesion varies from case to case. Some central pontine
lesions are so small (less than 6 mm) that they could easily be missed on section-
ing. There is also a wide variation in the size of the lesion and the clinical symp-
toms. Some patients had large lesions and minor symptoms, and some patients had
small lesions and major symptoms. Many had evidence of other CNS disorders,
especially Wernicke’s disease. Many cases have also been associated with other
chronic illnesses, especially pulmonary. These two cases did not present with other
major diseases. These two cases did have lesions outside the pons, as have others.
Central Pontine Myelinolysis—Microscopic Lesion
Source Convolutional Soft pons Myelin thin Nerve cells Increased glia Blood vessels
atrophy and pale affected engorged
Adams,
Victoe, + + + – + +
& Mancall
Cole,
Richardson, + + +
& Segarre
Tomlinson,
Pierider, + + + + +
& Bradley
Fig. 3
There is estimated to be about 10% of cases with more widely distributed lesions.
The two patients in this study had strikingly electrolyte disturbances (hyponatremia,
hypokalemia, and hypochloremia). These electrolyte disturbances have been noted
in most cases of CPM in the terminal phase, suggesting a contributory effect.
Speculative reasons for the highly selective nature of lesions in CPM are dis-
cussed. One reason stated is that the venous drainage could be insufficient, such
that poor drainage could predispose for disease. Others have suggested that the
oligodendrocytes and astrocytes were susceptible to disease and toxic/metabolic
problems. This could be a problem with the proximity of nerves (demanding good
circulation) and fiber bundles. This is strengthened by the location outside the pons
of lesions. These invariably involve brain areas where there is a close associa-
tion of neurons and fiber bundles such as the claustrum and internal capsule. The
authors also say that the close proximity of oligodendrites and neurons could be
detrimental in that neurons in stress could “steal” blood supply from oligodendrites
and myelinated fibers, decreasing their ability to remain healthy, and leading to
demyelination.
In another paper on CPM, the authors relate cases associated with burns and
serum hyperosmolality (McKee, A. et al., 1988). This was a retrospective study
over 16 years looking at CPM in patients dying during treatment for burns. Results
showed a frequency of 10 CPM patients out of 139 burn deaths compared to 10
cases out of 3528 autopsy cases in the general population. Data presented suggest a
causal relation between a prolonged episode of extreme serum hyperosmolality and
CPM. Evidence sited was hyperosmolality was present in all burn CPM patients and
seemed to correspond with the onset of CPM. Other published cases had hypona-
tremia and/or hyperosmolality.
The sequence of events could be that hyponatremia causes brain swelling. The
brain compensates by losing intracellular particles. Water is then lost, reducing
swelling. The brain is exposed to a relatively hypotonic insult. Because of hypona-
tremia and hypoosmolarity, burn patients are 25 times more likely to develop CPM
than non-burn populations.
Alcoholic Cerebellar Degeneration

A very detailed and comprehensive study of alcoholic cerebellar degeneration
(ACD) appeared in 1959 (Victor et al., 1959). In this study, 50 patients were fol-
lowed over many years. This group of alcoholic patients had a very specific and
uniform collection of symptoms, which were limited to gait ataxia, and ataxia to
the lower extremities. This ataxia evolved over a short period of time and was pre-
ceded and followed by relative stability. Cerebral lesions were largely located in the
cerebellar vermis.
Fifty patients with ACD were followed. There was little doubt that each patient
had a long previous history of alcohol consumption, exceeding 10 years duration
in most cases. Most had disruptive social lives, and many had DTs, and hallucina-
tions. Careful history and physical examination determined that severe nutritional
compromises had occurred in most patients. More than half presented with physical
evidence of nutritional deficiencies, such as loss of muscle bulk. Almost half of the
patients had liver disease, usually cirrhosis. Eleven of the patients no signs or had
histories of poor nutrition.
The average age of onset of cerebellar symptoms was 46. The range was 31–63
years of age. Over 60% had an onset of symptoms before the age of 50. The ini-
tial complaint of all patients, except two, was ataxia. This was variously described
as weakness, staggering, unsteadiness, or loss of balance. Signs upon exam were
nearly all ataxia of gait, and incoordination of legs. Symptoms such as nystagmus
were either mild or absent. Besides walking, leg movement disorders were present
in all but three cases. The leg symptoms were characteristic cerebellum incoordi-
nation, which was best demonstrated on heel to shin testing. Upper extremity coor-
dination involvement was minimal, with several cases showing no effect. A few
patients had a rhythmical tremor which appeared with certain postures, but was
not the usual cerebellar intention tremor. Cerebellar catalepsy was absent in these
patients.
Blood analysis, blood chemistries, and urinalysis were all normal, except if there
was a secondary disease such as cirrhosis. CSF was mostly normal in all cases
examined. Pneumoencephalograms were only obtained in three patients, but each
showed a dilatation of the fourth ventricle and cisterna magna and a small cerebel-
lum as compared to normal. Decreased intellectual capability was only present in
three cases in which there were earlier symptoms of Wernicke–Korsakoff syndrome.
Alcoholic cerebellar degeneration in this study had a variable course. In about
half (23) of the cases, the symptoms of ACD evolved rapidly, with the severity
of symptoms occurring in a few weeks to months. After this worsening period, the
progression stabilized. The stabilization seemed to correlate with improved nutrition
and abstinence.
In another group of patients, progression of symptoms continued to progres-
sively evolve over many years. Unless cessation of drinking and improved nutri-
tion occurred, these patients continued to get worse. A few of these patients did
stop drinking and eating poorly, and their deterioration plateaued. In some cases the
period of deterioration lasted as long as 15 years.
A third group showed a relatively stable period which was suddenly character-
ized by a rapid onset of deterioration. This group then showed plateauing in the
advance of symptoms. The rapid advance in symptoms could be measured in weeks
to months. An attempt was made to correlate the increase in severity of symptoms to
drinking or poor nutrition. In most cases, drinking had been more or less steady pre-
ceding the rapid worsening of symptoms. In a few cases, there had been a decrease in
alcohol consumption prior to an increase in symptom severity. Plateauing of symp-
toms was frequently associated with decreased drinking and improved nutrition.
Only six patients showed stabilization of symptoms without concurrent improve-
ment in nutrition and decreased alcohol consumption. When the ataxia became
established, even cessation of drinking and improved nutrition did not improve
symptoms. Only a few patients had any improvement, and this was thought to be
due to long-term better health and retraining in walking. This no doubt reflects the
more or less permanency of the cerebellar lesion.
Neuropathological examination was possible in 11 ACD patients described in
the above study. Grossly, brains from these ACD patients had a more or less normal
appearance. Coronal sectioning showed occasional dilatation of the lateral ventri-
cles, but otherwise seemed normal, except for the cerebellum. In some cases, cere-
bella showed a separation of the sulci and shrinkage of the vermis. Microscopically,
the culmen seemed to be most severely affected in most cases. In this region, there
were essentially no Purkinje cells. Astrocytic proliferation was present in the Purk-
inje cell layer in the affected cerebellar area. The molecular layer was only 1/2–1/3
the thickness of unaffected areas. Other areas affected included the flocculi, anterior
areas of the anterior lobes, and dorsal paraflocculus. Here too, Purkinje cells were
absent and there was astrocytic hyperplasia (Bergmann’s astrocytes). The molecu-
lar layer in these affected areas was also reduced in thickness, with accompanying
astrocyte proliferation. In the affected areas, many fibers were gone, as well as Purk-
inje dendrites. It was noted that the transition from “affected” to “non-affected” was
sharp, occurring over a linear span of a couple of cells. The occasional Purkinje cell
in the affected areas was pale and shrunken.
In areas in which destruction was somewhat less, there was still a reduction in
Purkinje cell numbers and the remaining cells were pale and small. Even in these
areas, the molecular layer was thinned and granular cells were reduced in number.
A reduction in axons and dendrites was seen in affected areas, and myelinated fibers
were pale.
Deep cerebellar nuclei were also affected. Thus, the dentate, globose, embo-
liform, and fastigial nuclei were usually described as having a loss of neurons and
shrinkage of those present. There was a concomitant proliferation of microglial cells
and astrocytes in these nuclei. Lipofuscin deposits were present in remaining neu-
rons. Fibers in the deep nuclei were decreased in numbers.
Certain other areas of the brain usually showed some involvement. For example,
the corpus callosum was thinned in some areas as much as 75%. Very mild pathol-
ogy was seen in a few instances in the cerebral cortex. The olives and superior
vestibular nuclei were involved in a few cases. These areas showed some reduced
numbers of neurons and occasional inclusions in remaining neurons. Myelin loss
was noted in a couple instances in spinal cord columns. The vagus nuclei and the
nucleus pontis showed minor involvement in a couple cases. Fibrose gliosis was
present in some areas; however, the mammillary bodies appeared normal.
The authors of this paper undertook a Purkinje cell count in ACD patients
as compared to control patients dying of unrelated causes. Four patients were
controls; two patients were alcoholic but had no neurological symptoms. Results
from controls were not significantly different from each other. Purkinje cell
counts in these controls ranged from 3.5 to 7.0 cells per millimeter. Counts
from patients with ACD showed, in highly affected cerebellar areas, less than 2
cells per millimeter. These numbers were of Purkinje cell presence or absence
and did not consider the size of cells or size of the shrunken area of the cere-
bellum. If the cerebellar area was itself shrunken, that would falsely elevate
the number of Purkinje cells per millimeter. Some control subjects were over
70 years of age and may have had minor age related decreases in Purkinje
cell numbers, but regardless, there was significant reduction in cells in ACD
patients.
The authors attempted to correlate clinical findings in each case to neuropatho-
logical findings at autopsy. In some cases, lesions were relatively mild, and patients
showed no signs of cerebellar symptoms. In other cases, where symptoms were
severe, some had widespread cerebellar lesions and others had lesions restricted to
the superior vermis and anterior lobes only. This is interpreted to mean that the
symptoms are generated by lesions of the superior vermis and anterior lobes. The
neuropathological basis for explaining both nystagmus and speech difficulties could
not be determined. The localization of symptoms to the above structures is not in
complete agreement with the functions of these cerebellar areas as determined by
other authors (Dow, 1938). Certainly the highly specific nature of the lesions, as
well as being bilateral, gives credence to the assigning of function by the authors.
More recent studies of the functional organization of the cerebellum (Henneman,
1956; Dow and Moruzzi, 1958) have presented data more in keeping with the
clinico-pathological results. These physiological studies involve recording and map-
ping evoked potentials from cerebellar areas, as well as defining motor effects
elicited by the electrical stimulation of similar cerebellar sites. This has resulted in a
map of topographical localization of represented body areas on the cerebellum. This
representation of body sites on the cerebellum correlates well with similar results
described in the ACD patients. For example, experimental physiological/anatomical
results show that cerebellar areas shown to be involved in lower extremity function
are precisely those lesioned by ACD. This could easily explain the most promi-
nent clinical characteristic, namely ataxia of gait, and leg incoordination. While the
culmen was almost always involved, the posterior area of the anterior lobes was
infrequently involved, thus the sparing of arm involvement. Head and neck involve-
ment was not seen clinically, and the declive was spared neuropathologically. The
site in the cerebellum responsible for nystagmus could not be determined.
The authors state that in terms of etiology, many potential causes are present.
Several patients had co-existing diseases such as carcinoma, but serious abuse of
alcohol occurred in all patients. In some patients, however, the cerebellar symptoms
occurred several years after abstinence started. Some of the abstinent patients were
under close supervision, and there was no drinking. This renders the alcohol toxic
possibility as an etiologic factor remote.
Another possibility is undernutrition. In fact, 75% of the patients in this study
were undernourished. Again, however, several patients had had long periods of nor-
mal nutrition. Scant few animal studies on nutritionally induced lesions in the cere-
bellum exist. Other conditions such as anoxia produce cerebellar lesions, but in
different areas than those seen in ACD. Some similar cerebellar lesions are seen in
hyperthermia, but the overall picture is of diffuse cerebellar involvement, not the
highly focal lesions seen in ACD. Other possible causes, including toxins such as
lead, methyl mercury, thiopene, DDT, carbon tetrachloride, nitrogen chloride, man-
ganese, dilantin, and injection of eosinophils, independently can produce cerebellar
lesions. For various reasons, such as location and nature of both lesions and symp-
tom production, all of these potential contributors can be excluded.
The authors state that it is speculative to suggest that the cerebellum in these
patients is damaged or weakened. The idea is that over time, the cerebellum was
brought to a condition such that it was “on the edge” and vulnerable. Arguing against
this concept is the fact that there never was a mild or “preclinical” cerebellar state
described. There are no childhood cases which might have predicted later develop-
ment of ACD.
There have, however, been descriptions of a biochemical feature of the cerebel-
lum which might help explain this concept. In most brain regions, there is an over-
abundance of enzyme capability. Thus, enzymes such as pyruvate dehydrogenase
may be present in 2–4 times the amount needed for normal function. The cerebel-
lum, however, seems unique in this regard. Studies have been described (Reynolds
and Blass, 1976) showing only a small excess of activity of pyruvate dehydroge-
nase in the cerebellum over what is required. The idea has evolved that such a small
reserve capacity might place the cerebellum in jeopardy more quickly than other
brain regions. This unique biochemistry could easily, over many years, be compro-
mised to the point that a small challenge, even after periods of abstinence, could be
precipitously detrimental. This concept, hard to prove, awaits verification. It could
however be the basis for the rapid deterioration of symptoms in ACD patients.
Marchiafava–Bignami Disease
Marchiafava–Bignami Disease (MBD) was first described in Italy and was thought
to only occur in Italian males drinking red wine (Marchiafava, E. and Bignami, A.,
1903). Soon it was found that MBD was a disorder primarily seen in alcoholics
regardless of ethnic background. MBD is a relatively rare disorder, with only about
300 cases reported in the literature, and nearly all have been chronic alcoholics. The
pathological feature which distinguishes this disorder is the symmetric demyeli-
nation of the corpus callosum. Lesions may infrequently extend laterally and may
involve the cerebral cortex.
In some very rare instances, complete recovery from MBD may occur (Tobita, M.
et al., 1997). In this case, T2-weighted MRI showed a high intensity of the splenium
and central corpus callosum. The patient was treated with doses of vitamin B com-
plex and a complete recovery was seen 3 weeks later. At this time, MRI showed
a nearly complete resolution of the T2-weighted high intensity seen earlier. The
authors state that this was an acute early example of MBD, quickly treated with B
complex vitamins.
While most cases of MBD are diagnosed at autopsy, increasingly a diagnosis is
made from clinical presentations and MRI. In a recent report (Arbelaez, A. et al.,
2003), two cases of MBD are reported in alcoholics who were diagnosed by both
clinical criteria and MRI. Both patients were chronic alcoholics, one consuming 1–2
bottles of wine per day for years. Clinical evaluation showed malnourishment, short-
and long-term memory loss, agraphia, ataxia, and a wide-based gait. Laboratory
data included hypoglycemia and decreased serum levels of thiamine. MRI in both
patients showed high T2 signal intensities in the corpus callosum, and in one case
extending into the white matter of the cortical frontal lobes. In each case, vitamin B
complex was initiated, which resulted in marked improvement in symptoms.
The authors point out that MBD is a disease almost exclusively of alcoholic and
malnourished patients and is thought to be a B complex (thiamine) related disorder.
This disease usually affects the body of the corpus callosum at first and then can
reach the genu and splenium. Other white matter areas may be affected in later
stages. The corpus callosum lesion may be layered, then show cavitation. During
acute phases, macrophages infiltrate. Demyelination may lead to loss of axons.
Patients with MBD may also have Wernicke’s disease, or even central pontine
myelination. Chronic MBD is characterized by dementia, and patients may live
many years. There may be a connection in some cases to hyponatremia, the cor-
rection of which can lead to myelinolysis. Diagnostic differentiation between MBD
and Wernicke’s disease is relatively easy due to the involvement of the mammillary
bodies in Wernicke’s disease.
Another case of MBD showed the usual involvement of the corpus callosum,
as well as symmetrical lesions in the putamen (Hayashi, T. et al., 2002). The
patient was a 65-year-old male who was a chronic alcoholic. When admitted, he
was comatose and had increased muscle tonis. Laboratory data revealed a decrease
in serum thiamine. At this point, thiamine treatment was started and in 24 hours
improvement began. An MRI on admission showed swelling with hypointensity on
the T1-weighted images from the corpus callosum. The corpus callosum and cere-
bral white matter showed hyperintensity on the T2-weighted image. In addition, the
posterior part of the putamina on both sides was involved with hypointensity on
T1 and hyperintensity on T2 images. Fifteen days later (and after 2 weeks of thi-
amine therapy) the corpus callosal swelling was diminished, but cystic lesions had
appeared. The lesions in the putamina were almost gone, and the cerebral white
matter lesions were completely absent.
The authors state that this represents the first report of lesions in the putamin
completely resolving. This case is also unique in that although the patient had coma,
a predictor of poor outcome, he recovered. The authors point out that in MBD cases
in which thiamine deficiency can be proven via serum thiamine testing, the defi-
ciency may be reversible with thiamine therapy. In this case, the above was true,
and the patient with MBD and comatose recovered. Thiamine treatment might well
be initiated in MBD patients when the diagnosis is made.
Clinical symptoms might be rather clear cut as in the above case, or more com-
plicated (Ferracci, F. et al., 1999). In this case, a confirmed alcoholic was admitted
with altered speech and gait. MRI shortly after admission showed hypointensive T1-
weighted, and hyperintensive T2-weighted images from the anterior/central areas of
the corpus callosum. Similar lesions were also seen in the centrum semiovale around
the lateral ventricles.
Clinically, the patient showed drowsiness, altered gait, absent ankle and plantar
reflexes, lack of speech, loss of reading ability, and loss of writing ability, especially
with the non-dominant hand. After the initial acute phase had passed, the patient
was mostly mute. Mutism has been described previously in cases of MBD. Features
of a hemispheric disconnect due to lesions in the corpus callosum manifest as alexia,
as in this case. The agraphia shown by this patient was worst in the non-dominant
hand. These symptoms can be explained by the results from MRI.
In another published paper (Demaerel, P. and van Paesshen, W., 2004), a 45-
year-old man who was a chronic alcoholic lapsed into unconsciousness. Clinical
exam showed peripheral nerve sensory neuropathy, dyspraxia, dysarrthria, and dis-
orientation. MRI showed diffuse swelling of the corpus callosum as well as some
degeneration of the cerebral cortex and cerebellum. Laboratory tests showed a low
level of serum thiamine. IV thiamine was administered (100 mg/day), with resultant
improvement in the clinical picture.
Yet another case of MBD involving successful treatment with thiamine has
recently been published (Kuo, C. et al., 2004). This case involved a 39-year-old
chronic alcoholic with a history of poor eating. Two weeks before admission, the
patient had slurred speech and alterations in mentation. Examination revealed the
patient was drowsy, had a wide-based gait, showed ataxia, and had mild bilateral
hyperreflexia. MRI of the brain showed hypointensive T1-weighted images and
hyperintensive T2-weighted images in the genu to the splenium of the corpus callo-
sum. Diffuse swelling was a feature of this lesion. The cerebellum, cerebral cortex,
and centrum semiovale appeared normal.
A diagnosis of MBD was made, and vitamin B complex plus thiamine was initi-
ated. During 12 days of hospitalization, his clinical signs and symptoms improved,
and he was discharged. The authors state that imaging studies can easily distinguish
between Wernicke’s disease and MBD, even though the clinical symptoms and his-
tories are similar. The edema seen in acute onset MBD abates, and is replaced by
demyelination and necrosis. The authors emphasize that with imaging techniques,
the diagnosis can be made, and vitamin B complex and thiamine instituted. This
aggressive and rapid treatment with thiamine has been responsible for stopping
the progression of MBD, and reversing some symptoms, as occurred in the case
presented.
The utility of MRI as a key diagnostic technique has been investigated
(Meneegon, P. et al., 2005) This was a retrospective study in which six patients were
identified as having MBD. Five of the six were chronic alcoholics and all had MRI
during the acute phase. Results of MRI showed hyperintensity of the corpus callo-
sum. Cerebral cortex lesions were found in the fronto-parietal cortex in threee of
the six patients. Of the three who did not have cortical lesions, improvement or full
recovery occurred. Of the three with cortical lesions, two died and one was left with
a severe cognitive impairment.
The authors state that low apparent diffusion coefficient and lesions in the cere-
bral cortex are associated with a poor prognosis. This is the benefit to MRI in that
it can serve to predict outcome. Immediately after diagnosis, thiamine treatment
was initiated, which was beneficial in some of the patients. The authors state that
diffusion weighted MRI enables an improved capability for diagnosis in MBD. If
diagnosis can be made accurately at an earlier stage, then treatment regimes have a
better opportunity for success.
In keeping with the above case is a description of MBD in a chronic alcoholic
(Haas, L. et al., 2006). This patient was found unconscious and with dysarthria.
Examination showed the patient was comatose, with symmetrical pupil reactions.
While slight diverging strabismus was present, most other findings were normal.
His serum thiamine was low. CT scan showed no abnormalities. An MRI showed
high intensity T2 lesions in the corpus callosum and internal capsule. These images
represented edema and demyelination.
The authors state that if MBD is mild or has a recent onset, CT scans may be neg-
ative. The clinical manifestations may be difficult to interpret as regards a diagnosis.
Before MRI, the diagnosis was made at autopsy. Use of MRI can reveal lesions in
the corpus callosum, which almost always indicates a diagnosis of MBD. Although
an uncommon occurrence in chronic alcoholics, making the diagnosis with MRI at
an early stage will justify thiamine treatment. In this patient, clinical treatment was
noted 2 days after thiamine treatment (100 mg/day) was initiated.
Occasionally MBD can be associated with “overlapping” findings of some other
alcohol associated thiamine deficient disorder. In this case, lesions similar to those
in CPM are described (Goswami, P. et al., 2008). This patient was admitted in
a stuporous state, and had a 15-year history of alcoholism. The patient had mild
hyponatremia and decreased serum thiamine levels. MRI findings showed areas of
T1 hypointense and T2 hyperintensities in the corpus callosum. These lesions were
in the central areas of the corpus callosum. Also, the sagittal image showed a T2
hyperintense area in the pons. Based on the MRI a diagnosis of MBD was made,
and thiamine treatment initiated, and a gradual correction of the hyponatremia was
begun. Improvement in symptoms was noted within a few days, and a fairly good
recovery was achieved.
The authors note that the MRI was characteristic of MBD in that the lesion in the
corpus callosum showed the main lesion located in the middle layer, with the dorsal
and ventral layers spared. This produces a “sandwich”-like appearance. The body
of the corpus callosum is usually affected first, followed by the genu and splenium.
Over time these lesions may spread to the putamin and cerebral cortex.
While patients were formerly diagnosed at autopsy, MRI has speeded diagnosis
and the initiation of thiamine therapy, which may be beneficial. Many patients with
an acute form of MBD may become comatose and die. At the other end are more
mild, chronic cases which may be amenable to treatment and survive many years.
The use of MRI enables the physician to estimate the extent of the disease. Lesions
limited to the body of the corpus callosum may be early, whereas lesions extending
to the cerebral cortex and putamin may represent a longer lesion process and a less
favorable outcome.
The finding in this case of CPM is particularly rare, but points out that one must
be alert to variations in not only clinical presentations, but also in lesion distribution.
In another case report (Fukumoto, J. and Suzuki, T., 2007), a 60-year-old alco-
holic was admitted in a stuporous condition. Examination showed polyneuropathy
with diminished deep tendon reflexes, hypertonic limbs, distal muscle atrophy, gait
disturbances, and amnesia. MRI showed T1 hypointensities in the corpus callosum
and T2 hyperintensities in the same areas. There were no other lesions noted in the
cerebellum or brain stem. There was significant cortical atrophy. Laboratory data
showed low serum thiamine levels. The patient also had signs of disconnection syn-
drome, namely apraxia, acalculia, and dyslexia. Thiamine, vitamin B complex, and
mianserin hydrochloride (for depression) therapies were started.
A gradual recovery of many earlier symptoms occurred. Thus, motility improved,
gait deficiencies decreased, dysarthria disappeared, and socialization improved. By
40 days, apraxia diminished. By day 60, the corpus callosum lesions seen earlier
on MRI had clearly decreased in size. Later still, his disconnection disorder had
completely resolved.
The authors note that MBD can be an acute inflammatory disorder in which the
splenium of the corpus callosum is affected. The disorder occurs almost always in
alcoholics, and is linked to disordered thiamine metabolism. Usually patients are
subject to one of three types of MBD: acute, subacute, and chronic. Both acute and
subacute tend to have poor outcomes, but the chronic form may be amenable to
thiamine treatment. The authors note that in a study where asymptomatic chronic
alcoholics had MRIs, two-thirds showed a reduced size of the corpus callosum
(Harris, C. et al., 1993). It is not known whether MBD represents an end stage of
years of deterioration of the corpus callosum, or represents an increased suscepti-
bility of some patients to corpus callosum damage. An acute phase might argue in
favor of the latter.
Hemorrhages are probably a feature of the lesions since hemosiderin is character-
istically found in the walls of the callosal cystic lesions. Such hemorrhagic lesions
are also seen in lesions in Wernicke’s disease
Mianserin hydrochloride was used as an antidepressant drug and also for its abil-
ity to decrease delirium. The authors suggest that the mental features of MBD should
be treated as well as using vitamin B complex and thiamine to treat the physical
aspects of MBD.
Thiamine treatment does not always work. In a case report (Kim, M. et al., 2007),
a patient was admitted stuporous and had had seizures, altered mental status, and
had consumed large quantities of alcohol for at least 10 years. MRI showed T2-
weighted hyperintensities, which were localized to the splenium and body of the
corpus callosum. Lesions were also found in the frontal cerebral cortex.
This patient was diagnosed and placed on vitamin B complex treatment plus
1000 mg/day of thiamine and corticosteroid delivered IV. In spite of this, the patient
remained stuporous. MRI at 15 days showed decreased signal density in the cere-
bral cortex lesions, and corpus callosum. Twelve weeks later, MRI showed corpus
callosum atrophy, with multiple lesions in the splenium, body, and genu. The patient
remained in a semi-vegetative state.
The authors note that the diagnosis can be difficult, but the lesions in the corpus
callosum and lack of lesions in the mammillary bodies were strongly suggestive of
MBD. This patient was unique in that he had diffuse frontal cortex lesions. These
have been seen before and their generation is unclear. They might be from a long
time existence of corpus callosum lesions, or have arisen due to some other condi-
tion such as hyponatremia and its rapid treatment.
An unusual case as regards symptoms was presented (Hirayama, K. et al., 2008).
This was a chronic alcoholic patient who showed on MRI symmetric demyelination
and necrosis in the genu, body, and anterior splenium of the corpus callosum. The
unusual symptom related to the coordination syndrome seen in most cases of MBD.
In this patient, the dyspraxia took a form in which both of the patient’s hands could
cooperate in a sequence of bimanual actions. For example, his right hand might
start an action with the left hand cooperating. Striking was that once the action was
completed the left hand would commence undoing the action in an antagonistic
fashion. In this antagonistic action, the right hand cooperated with the now “dom-
inant” left hand. These “back and forth” actions continued until the patient was
forcibly stopped. This occurred only in bimanual tasks and not at all in voluntary
actions. They only occurred in response to verbal commands or in actions imitating
someone else.
The authors speculate that this antagonistic action could result by conflict
between the two hemispheres due to inter hemispheric disinhibition caused by the
partial demyelination of the corpus callosum. Any bimanual coordination is still
possible because the corpus callosum is not completely disrupted.
As a sort of plea for MRI and early diagnosis is the report of two cases in which
the presentation was an acute onset coma (Vazquez, C. et al., 2008). MRI revealed
significant lesions in the corpus callosum. As many previous studies have shown, an
acute onset with extensive corpus callosum (and possible involvement of the frontal
cerebral cortex) is an ominous feature. These are most frequently associated with
death, whether or not thiamine therapy is initiated.
MBD has gone from a disorder solely diagnosed at autopsy and not treated to one
frequently diagnosed by MRI and amenable to treatment. This behooves the clini-
cian to be alert to the possibility of MBD in chronic alcoholics. When the diagno-
sis is made, thiamine treatment should be immediately started. Vitamin B complex
treatment should also be started. The possibility that quick treatment might lead to
improvement – even complete resolution – of some or all symptoms and signs places
MBD into a classification of “treatable,” which was not considered before modern
imaging permitted the rapid ability to diagnose this disorder.
Leigh’s Disease
The disease subacute necrotizing encephalomyelopathy (SNE) was first described

in 1951 by Denis Leigh and therefore is also called Leigh’s disease (Leigh, 1951).
This disorder, although afflicting mainly children, has many features similar to those
of Wernicke’s disease. These neuropathological features were so striking as to have
been noted in the first case description by Dr. Leigh. Many studies have evolved
over the years describing various biochemical features of Leigh’s disease. Most
recently, genetic and mitochondrial alterations in brain, blood, and tissue of SNE
patients have been described. This chapter will examine results from these studies
in a chronological order and try to reach some valid conclusions and note where
information is lacking.
Denis Leigh described the first case of SNE in 1951. This was a case of a
7-month, 3-week-old boy admitted to the hospital with pupils unreactive to light
and limb spasticity, and he appeared to be deaf (see Table 1). Over a period of 3
days, the patient became rapidly worse, lapsed into a coma, and died.
An autopsy was performed nearly 3 days later and was unremarkable except for
CNS findings (see Table 2). Upon gross examination of the brain, it was noted that
mid-thalamic regions showed areas of brownish discoloration in the dorsomedial
nuclei, nucleus submedius, substantia nigra, and the pulvinar. In the midbrain, sim-
ilar lesions were noted in the entire tegmentum. In the pons, lesions were found
centrally, and also in the trigeminal motor nuclei. The medulla showed the same
brownish lesions in the inferior olivary nuclei, and in the central gray. The posterior
columns were similarly affected in the spinal cord. The lesion became less con-
spicuous at lower spinal cord levels. There was a striking bilaterally symmetrical
characteristic to these neuropathologic lesions.
When sections were stained with Perdrau’s silver stain, the regions which
appeared to have a brownish color grossly were all sites of proliferation of small
blood vessels, including arterioles, pre-capillaries, capillaries, and venules (see
Table 3). The proliferation was not associated with hemorrhage or wall thicken-
ing. Neurons in close proximity to these vascular alterations were damaged. Neu-
ronal damage included cell body swelling, chromatolysis, and nuclear swelling.
Microglia were increased and were concentrated at the periphery of the lesion.
Oligodendrocytes were reduced in number. Sections stained with Heidenhain’s stain


for myelin showed widespread demyelination in affected areas. Microscopic dam-

age followed that seen on gross inspection. In the midbrain, the oculomotor nuclei
were severely damaged. Striking again was the bilateral symmetry. Also noteworthy
was the complete sparing of the mammilary bodies. At the level of the medulla, the
destructive process was limited to the inferior olives only. In the spinal cord, micro-
scopic changes were limited to the dorsal columns. These changes diminished with
descending levels of the spinal cord.
In the discussion of this case, Dr. Leigh comments on the similarity of the lesion
in this case to that seen in Wernicke’s disease. He also comments on the striking
bilateral symmetry of the lesions and on the highly localized, punctate nature of
the lesion. In considering a diagnosis, the author mentions some form of poisoning
(arsenic), equine encephalitis, and subacute necrotizing myelitis as possible diag-
noses, but ultimately he rules them out.
Finally, the author is striken by the similarity of the lesions in this case to those
seen in Wernicke’s encephalopathy. These similarities include a vascular response,
nerve cell damage, and proliferation of microglia and astrocytes. There are differ-
ences in regional sites between this case and Wernicke’s disease, most notably, the
mammillary bodies, damaged in Wernicke’s disease, were spared in this case. Also,
spinal cord lesions, as seen in this case, are not commonly noted in Wernicke’s
disease.
Leigh points out in his 1951 paper, that very similar CNS lesions were pro-
duced in earlier studies by administering quinoline compounds to monkeys (Richter,
1949). The lesions in these experimental monkeys were vascular in nature, highly
specific in location, and bilaterally symmetrical. The suggestion was offered that this
pattern was closely matched to lesions seen in Wernicke’s disease. In this first case
description of SNE, Leigh considers the diagnostic options of an unknown toxin
or virus, but seems to favor an infantile form of Wernicke’s disease as the possible
cause of this new encephalopathy.
Another early comprehensive description of Leigh’s disease Richter (1957)
details symptoms and pathological changes in three cases. All three had similar
symptoms, which began between 7 and 13 months. All three died after periods of
illness ranging from 4 to 14 months. Symptoms in these three cases were quite sim-
ilar and included reduced food intake, diminished to no response to light and sound
stimuli, retardation, seizures, spasticity, stupor, and death (see Table 1).
Upon gross examination of the brains of these three cases, many similarities were
noted, compared to the earlier cases (see Table 2). Areas showing gray-red discol-
oration and a softer texture included the caudate nucleus, putamen, basal ganglia,
striatum, substantia nigra, tegmentun of the pons, and the medulla. There appeared
to be a decrease in the relatively sharp demarcation between brain regions. The
bilateral symmetry of these gross lesions was noted.
Microscopically, lesions were generally similar to those of earlier studies (see
Table 3). These consisted in the first case of areas of degeneration of the periaque-
ductal region and substantia nigra. The degeneration consisted of necrosis involving
the ground substance, as indicated by incomplete staining. There were notable vas-
cular changes which included an increase in dilated pre-capillaries and capillaries.
Leigh’s Disease 83
In these areas, there was some limited neuronyl degeneration, but the neurons
were largely spared. There was demyelination and astrocytic proliferation. In the
pons, case 1 showed extensive demyelination throughout the tegmentum, and in
the medulla, there was increased vascularity, gliosis, and demyelination throughout
the reticular formation.
In the second case of Richter, many histological sections took the various stains
poorly, signaling degeneration of ground substance. In lesioned regions, there was
the consistent finding of increased prominence and/or numbers of pre-capillaries
and capillaries. In affected areas such as the occulomotor nuclei, the neurons were
normal, but there was significant astrocytic proliferation throughout the nucleus.
The trochlear nuclei were similarly affected. The myelin pattern in the brainstem
was affected (demyelinated) in the tegmentum and substantia nigra. The pontine
tegmentum showed vascular proliferation in the locus ceruleus, caudate, putamen,
globus pallidus, and subthalamic nuclei.
In Richter’s case 3, the tegmentum of the brainstem was spongy on sectioning,
and lesions were found in the trochlear nucleus. Neurons in this case were relatively
spared, but there was capillary and pre-capillary proliferation (prominence). Lesions
in the substantia nigra were the same, i.e., neurons were relatively spared. Lesions
in the pons also consisted of pre-capillary and capillary prominence, with gliosis
prevalent. In the caudate nucleus, most neurons were gone and the spaces were
occupied by a framework of proliferating capillaries. Microscopically, lesions were
noted in the putamen and the thalamus. Myelinated fibers in these areas were greatly
reduced or absent.
There were striking similarities in Richter’s cases as compared to previously
described cases (Tables 1, 2, and 3). First, was the bilateral symmetry of the lesions,
noted in all cases. Also noted were the highly specific and localized nature of the
destruction. Finally, in each case of SNE was the similarity of cerebral lesions to
those seen in cases of Wernicke’s disease, except for the sparing of the mammillary
bodies in SNE. Richter speculates that given the relative vulnerability of the new-
born brain, it is odd that in SNE, the mammillary bodies are spared, while almost
always involved in Wernicke’s disease. Richter suggests, however, that a clear patho-
logical relation exists between Leigh’s disease and Wernicke’s disease. Richter, at
this early time, suggests that SNE, unlike Wernicke’s disease, may have a genetic
base. Although Wernicke’s disease cases usually show some level of nutritional dis-
order, Leigh’s cases may have “feeding difficulties,” but overall, nutritional status
is about normal. From this, Richter suggested that there might be a “poisoning” of
metabolic enzymes, or that there could be an inherited enzymatic defect responsible
for the clinical disorder of Leigh’s disease.
In his discussion and conclusions, Richter noted the similarity of pathological
findings in his three cases to those described by Feigin and Wolf (1954). Feigin and
Wolf (see Tables 1–3) noted similarities of Leigh’s disease lesions to those seen
in Wernicke’s disease. While Feigin and Wolf pointed out the feeding problems in
cases of Leigh’s disease, these authors felt that this was a result of the disease as
opposed to a cause or contributor to the disorder. The fact that these early descrip-
tions of Leigh’s disease were so similar as regards to pathological changes, leads to
the conclusion that they collectively form an identifiable pathological entity which
is unique from other general infantile degenerative disorders, and from Wernicke’s
disease.
There followed papers describing cases of Leigh’s disease in combination with
other symptoms and findings not ordinarily associated with SNE (Kamoshita et al.,
1970). These authors found infantile spasms and hyperarrhythmia, and an abnormal
EEG. The uniqueness of these combined findings is discussed, but no conclusions
are drawn. In other studies, the association of Leigh’s disease and renal and arterial
lesions are described (Crome, 1970), and ocular findings have also been examined
(Grover et al., 1970).
In 1967, Yashon and Jane (1967) described the 21st–23rd cases of Leigh’s
disease. Symptoms in these three cases were overall similar to those described in
previous cases (Table 1).
SNE SYMPTOMS
VISUAL HEARING SPASTICITY STUPOR/COMA SEIZURE ACTIVITY
D. LEIGH + + + +
1 CASE
R. RICHTER + + + +
3 CASES
YASHON + + + +
& JANE
3 CASES
FEIGIN & + + + +
WOLF
3 CASES
DEVIO, + + + +
ET.AL
1 CASE
Table 1
However these authors point out that varying degrees of pathologic involvement
influence the expression of symptoms in each individual case. Seizures were present
in one of the three cases described by Yashon and Jane and in only two previously
reported cases. Malnutrition, present in some cases, was absent in one of these three
cases. In the three cases of Yashon and Jane, there was no evidence of any genetic
link, as was noted in previous studies, which had shown a high incidence of perinatal
deaths in family histories.
Regarding the cerebral lesions, Yashon and Jane (1967) point out that prominent
capillary vascularity and breakdown of ground substance, giving a loose appear-
ance, are so characteristic that classification of a case of Leigh’s disease without
these features would be impossible. Again, Yashon and Jane comment on the sym-
metry of lesions, the relative sparing of neurons in lesion areas, and the striking
similarity to the lesions of Wernicke’s disease. Unusual in these cases was the find-
ing of involvement in the mammillary bodies. Mammillary body lesions are almost
always present in Wernicke’s disease, and involvement of mammillary bodies in
one case of Yashon and Jane is only the second report of its kind in the literature in
Leigh’s disease. Again, in relation to Wernicke’s disease is speculation that, because
of lesion similarity, some sort of thiamine metabolism alteration must be considered
in Leigh’s disease. The paper of Yashon and Jane concludes with an interesting
discussion about the relation of SNE to other thiamine-related disorders besides
Wernicke’s disease. For example, the cerebral lesions in Chastek paralysis, which
are very similar to those seen in both SNE and Wernicke’s disease, are examined.
Chastek paralysis (in foxes) was eliminated (reversed) by thiamine treatment. There
is speculation of possible enzyme antagonist activity in human cases which might
be responsible for the symptoms and lesions (Tables 2 and 3).
SNE GROSS BRAIN FINDINGS
THALAMUS PULVINAR INF. PONS OLIVE MEDULLA POSTERIOR SUBSTANTIA CORTEX CAUDATE
COL. CENTRAL COLUMN S.C. NIGRA
GRAY
D. LEIGH + + + + + + +
1 CASE
R. RICHTER + + + + + + +
3 CASES
YASHON + + + + + + +
& JANE
3 CASES
FEIGEN & + + + + + +
WOLF
3 CASES
DEVIO, + + + + + +
ET. AL
1 CASE
Table 2
SNE HISTOLOGICAL FINDINGS
VASCULAR NEURAL ACTIVE DEMYELINATION GLIOSIS SOME

PROLIFERATION DEGENERATION MICROGLIA (ASTROCYTIC) NEURAL
SPARING
D. LEIGH + + + + +
1 CASE
R. + + + + + +
RICHTER
3 CASES
YASHON + + + +
& JANE
3 CASES
FEIGEN & + + + + + +
WOLF
3 CASES
DEVIO, + + + + + +
ET. AL
1 CASE
Table 3
In 1968, a paper was published in which lactate and pyruvate were studied in
a child with supposed Leigh’s disease (Hommes et al., 1968). Results showed that
both lactate and pyruvate were elevated in this patient, who showed symptoms sim-
ilar to those of Leigh’s disease, and who had three siblings die of symptoms said to
be those of Leigh’s disease (Clayton et al., 1967). A liver biopsy showed an almost
complete absence of pyruvate carboxylase, a gluconeogenic enzyme. Data from this
patient were suggestive for a defect in gluconeogenesis. With this in mind, this
patient was treated with lipoic acid. As described before (Clayton et al., 1967), lipoic
acid had a significant improvement in symptoms, and lactate and pyruvate blood
levels dropped. The authors emphasized that this case was suggestive of Leigh’s
disease; at the time of publication, the patient was alive, and there was no patho-
logical conformation of the diagnosis. This paper did reconfirm possible beneficial
effects of lipoic acid treatment.
In 1969, a “factor” was described in the urine of a patient with Leigh’s disease
which inhibited the enzyme thiamine pyrosphate-adenosine triphosphate phospho-
transferase (Pincus et al., 1969). The reaction is as follows: TTP + ATP → TTP
+ ADP. Postmortem biochemical analysis in Leigh’s patients yield TTP values in
brain, liver, and kidney near zero as compared to normal control tissue. In these
patients, activity levels of pyruvate decarboxylase, alpha-ketoglutarate dehydroge-
nase, and transketolase (all thiamine-requiring enzymes) were slightly decreased or
not changed at all (alpha-ketoglutarate dehydrogenase). Not only urine but blood
and spinal fluid also contained the inhibitor.
Because of the striking similarity of cerebral lesions in Wernicke’s disease and
Leigh’s disease, investigators were looking for some alteration in the known neuro-
chemical reactions involving thiamine. In the Pincus et al. paper, the finding of an
inhibitor in body fluids of the reaction forming TTP seemed exciting because of the
following: (1) this could explain the development of the disorder; (2) this could also
explain normal levels in vitro of thiamine-requiring enzymes; and (3) this finding
could become the base for a relatively simple urinary test for Leigh’s disease. In this
case, brain tissue was obtained 45 minutes after death, and samples were analyzed
for TTP. Results showed a virtual absence of TTP in brain samples, although TTP
was present in liver and kidney. This finding suggested that the inhibitor is only
active in brain tissue. As was pointed out by the authors, thiamine monophosphate
constitutes 10% of thiamine in the body, TPP about 80%, and TTP about 10%. The
physiological role of TTP in brain function is not clear, but may play a role in nerve
conduction.
In a subsequent paper (Cooper et al., 1970), description is presented of results of
screening samples of urine from cases of possible SNE, as well as normal controls.
These studies showed five positive cases of SNE, and three probable cases based
on the urine test for inhibitor. There were three false positive tests – two in siblings
of SNE patients and one in a case of metachromic leukodystrophy. Results are also
described stating that the inhibitor is a labile glycoprotein with a M.W. of around
30,000. The lability makes isolation and characterization difficult. It is stated that
several patients diagnosed with SNE were at that time being treated with thiamine
and have shown improvement.
In a somewhat later paper (Pincus et al., 1971), the effects of treatment of a case
of Leigh’s disease with thiamine and thiamine propyl disulfide were described. Ini-
tially, treatment with the above compounds was associated with a dramatic improve-
ment in neurological symptoms. Later, however, thiamine levels in CSF dropped,
and symptoms became worse. Following death, tissue from the patients’ brain
showed low thiamine levels, and TTP was absent in samples taken from the pons.
The reasons for these results are unclear.
Another interesting paper by Simopoulos et al. (1972) details changes in SNE

not heretofore described, and reflects the complexity of this disorder. This paper
describes a case of SNE in which the patient displayed many of the symptoms
of SNE, including visual symptoms, poor muscular development, and ataxia. The
patient was found to be hyperpneic and had hypomagnesemia and hypokalemia.
Both blood and urine were analyzed for the inhibitor, and the results were positive.
Based on this, and the clinical findings, SNE was suspected. Anemia was an early
feature, and a bone marrow aspiration showed pronounced vacuolization. A renal
biopsy was normal. Because of the close relation of SNE and Wernicke’s disease,
a treatment regime of thiamine was tried. Administration of thiamine resulted in a
disappearance of the inhibitor, and 3 months after initiating thiamine treatment, a
bone marrow aspiration showed that the vacuoles seen previously had disappeared.
Upon discontinuation of thiamine treatment, the vacuoles had reappeared in the
bone marrow. Symptoms seemed to improve such that the patient was able to attend
kindergarten. Her IQ at that time was 90.
Subsequently, the patient had headaches, refused to eat, refused medications,
and was readmitted to the hospital. She was dehydrated, emaciated, semi-comatose,
and showed opisthotonic posturing. The patient was clearly in a severe relapse, and
despite IV thiamine administration died.
Postmortem exam revealed no gross abnormalities in the cerebrum, basal ganglia,
brainstem, or cerebellum. The spinal cord had discoloration in the dorsal columns.
Microscopically, there was a striking vacuolization in the white matter. Myelin loss
was most conspicuous in the dorsal columns. Some vacuolization was seen in sub-
cortical fibers, and mild neuronyl loss was noted. Mild vacuolization was present
in the basal ganglia, thalamus, hypothalamus, and mammilary bodies, but there was
no neuronyl loss or vascular prominence. This same vacuolization was noted in the
reticular formation and periaqueductal gray.
As pointed out by the authors, respiratory alkalosis, hypokalemia, hypomagne-
semia, aldosterosemia, and vacuolization of bone marrow have never before been
described in Leigh’s disease. Regarding brain pathology, the mammillary bodies in
this case showed changes rarely seen in other cases of Leigh’s disease. In addition,
other brain lesions, gross and microscopic, do not duplicate those characteristically
seen in Leigh’s disease. Even the course of this patient’s disease, drawn out over
several years, is not consistent with other cases. This patient did have the inhibitor
present in both blood and urine.
Thiamine has been previously reported to have beneficial effect in the treatment
of megaloblastic anemia and in this case resulted in the decrease in bone marrow
vacuoles. Thiamine also resulted in the disappearance of the inhibitor in blood and
urine. When thiamine treatment was stopped, these diminished changes returned.
The bone marrow alterations, coupled with the presence of the inhibitor, suggest a
variant of SNE not yet described. The lack of characteristic clinical course and of
neuropathological changes argues in favor of the above hypothesis. This case would
probably not been called Leigh’s disease 10 years earlier.
In a paper not directly associated with a symptomatic Leigh’s patient, results
are presented of the measurement of the urinary inhibitor in the parents of seven
cases of Leigh’s disease (Murphy, J. et al., 1974). The inhibitor of the TPP+ATP
→ TTP+ADP reaction was similar in parents to that in confirmed Leigh’s patients
in that (1) thiamine treatment reduced the levels of the inhibitor, (2) the inhibitor
becomes inactive after three weeks of frozen storage, and (3) the inhibitor is
destroyed by boiling water. Because the inhibitor in parents was found at the same
levels as in symptomatic patients, this assay cannot be used to identify patients
before onset of symptoms. Also of interest is the finding that parents with inhibitor
are asymptomatic.
In another paper (Hommes et al., 1973), the effect of thiamine administration
in rat liver mitochondrial pyruvate dehydrogenase activity was examined. In this
study, rats were injected IP with thiamine dichloride and after five days, liver mito-
chondria were prepared and analyzed. Results showed that inhibition of the enzyme
by ATP was reduced in treated as compared to non-treated rats. The implication is
that thiamine administration “improves” the ability of pyruvate dehydrogenase to
metabolize pyruvate, thereby increasing the efficiency of the TCA cycle. This could
serve to explain the beneficial effects of thiamine treatment in Leigh’s patients.
Further studies of inhibitor characteristics (Murphy et al., 1974) showed that all
obligate carriers of Leigh’s disease have inhibitor in their urine. Further, it was found
that the inhibitor could be stabilized by 50 mM mercaptoethanol. That the inhibitor
was not found in an active form in controls suggested that it is in a unique easily
measured marker for SNE. The inhibitor has a M.W. of around 37,000. This paper
also demonstrated that the inhibitor is specific for the brain reaction of TPP + ATP
→ TTP + ADP; the inhibitor had no effect on the liver reaction. The question of
why do carriers have similar inhibitor activities, yet are asymptomatic is discussed.
It is suggested that (1) the level/activity of inhibitor in brain tissue is not the same in
patients vs. carriers or (2) that the sensitivity of dietary thiamine in carriers is more
“efficient” in mediating the inhibitor effect than it is in patients with overt Leigh’s
disease (Pincus et al., 1973).
Pincus et al. (1973) published a paper in which thiamine treatment was exam-
ined in a large group of 21 SNE patients. Two groups of patients were studied: (1)
12 patients with a secure diagnosis of Leigh’s disease, largely by autopsy, and (2)
nine patients in a “presumptive” category. Results showed that in the group of 12
patients, 9 died. Why this occurred is discussed in light of the fact that with treat-
ment, several enjoyed periods of remission. The remissions in the treated group of
patients were more significant and lasted longer than those of untreated cases from
the literature The ultimate result of mortality was discussed. Three reasons are cited
to explain why, when treated, 9 of 12 patients died The first possible reason is that
treatment was initiated in at least three cases when the patient was in extremis and
had already developed irreversible brain structural damage. A second reason could
be the development of an adaptive mechanism, which results in the progressive low-
ering of thiamine levels, despite treatment, and the third possibility is that SNE is
not a simple vitamin deficiency, but involves direct inhibition of thiamine-related
reactions.
The fact that carriers of the inhibitor are asymptomatic (Murphy et al., 1974)
and the aforementioned suggestion of the development of an adaptive mechanism
to offset thiamine treatment are interesting. It could be that thiamine or its ester in
carriers is somehow more efficient in mediating the inhibitor effect than is the case
in patients with overt SNE.
The question of the specificity of the urine inhibitor as a test for Leigh’s disease
has been carefully examined (Pincus et al., 1974). In this study, the urine of over
500 subjects was tested for the presence of the inhibitor. Subjects were arranged
in a variety of categories, including autopsy-proven cases, living cases of Leigh’s
disease, patients thought to have Leigh’s disease, patients once suspected of Leigh’s
disease, normal controls, etc. Results showed the false positive rate was 6.4%. More
than one-third of parents of Leigh’s cases tested positive for the inhibitor. In 16
cases of Leigh’s disease confirmed at autopsy, or living, with siblings confirmed
at autopsy, all had the inhibitor present in their urine. In 42 patients suspected of
Leigh’s disease, but later proved to have some other disorder, 39 tested negative
for the inhibitor. In this study, no false negatives were found in untreated patients
when the urine samples were handled properly. The authors are careful to state that
the test is more definitive in ruling out the diagnosis of Leigh’s disease than it is in
establishing it due to the 6.4% false positive rate.
In a somewhat more thorough autopsy examination report, pathology in other
organs besides the brain have been described (Crosby and Chou, 1974). This was an
autopsy-proven case of Leigh’s disease, with a somewhat unusual course in that the
patient had suffered a traumatic brain injury, had a drawn out course of symptoms
(over 4 years), and died at the age of 10. At autopsy, the brain showed typical gray
brain discolorations in the globus pallidus, hypothalamus, substantia nigra, tegmen-
tum of the medulla/pons, and lesions were noted in the spinal cord. Microscopic
examination was characteristic of Leigh’s disease. The main feature was a prolif-
eration and hyperplasia of capillaries in affected areas, and demyelination. In most
areas, neurons were spared.
In addition to the cerebral changes, light and electron microscopy were per-
formed on biopsy samples and on samples taken at autopsy. Findings included the
presence in muscle biopsy samples of red staining (trichrome stain) clumps of mate-
rial in about 5% of fibers. These clumps were called “ragged-red fibers.” Most other
features of the muscle fibers were within the normal range, except mitochondria,
which appeared enlarged. They contained rectangular paracrystalline arrays located
in the crystal membranes. There were also intramitochondrial dense bodies. These
characteristic mitochondrial changes are associated with several mitochondrial dis-
orders. In Leigh’s disease, biochemical features such as elevated lactate and pyru-
vate, decreased pyruvate dehydrogenase activity, and altered amino acid metabolism
are all noted. The authors of this paper speculate that previous autopsy examination
has focused on the brain, and perhaps missed muscle and heart changes seen in this
case. The idea is advanced that Leigh’s disease may primarily be a mitochondrial
disease, with variations, and that further studies should also examine muscle and
heart. The existence of cardiomyopathy and mitochondrial myopathy in an autopsy-
proven case of Leigh’s disease stresses the importance of a full thorough investi-
gation and suggests that some cases of mitochondrial myopathy may actually be
Leigh’s disease.
In another study, the cerebral levels of thiamine, thiamine monophosphate,

thiamine pyrophosphate, and thiamine triphosphate were measured in brain tissue
from proven Leigh’s cases and non-Leigh’s controls (Murphy and Craig, 1975).
Results showed that levels of thiamine triphosphate were significantly decreased in
Leigh’s patients as compared to controls. In addition, thiamine pyrophosphate lev-
els were elevated in Leigh’s brains as compared to controls. This suggests that the
inhibitor is active in inhibiting the phosphotransferase activity. Brain tissue from
Leigh’s patients and controls were able to catalyze the reaction ATP + TPP →
TTP + ADP in vitro, demonstrating that the enzyme was present, but blocked in
Leigh’s patients in vivo. The inhibitor did not block the reversal of this reaction.
The authors state that these data suggest that the inhibitor plays a major and sig-
nificant role in the etiology of Leigh’s disease. This study does not rule out other
possible contributory (mitochondrial) defects. The authors speculate that since both
Wernicke’s disease and Leigh’s have many pathological changes in common, and
both have a deficiency in TTP, then TTP must play an important role in normal
cerebral metabolism. There is no inhibitor present in Wernicke’s patients, yet lev-
els of TTP are low due to overall thiamine deficiency. Thus the depleted levels of
TTP result from a different mechanism, yet there are impressive neuropathologic
similarities.
In a brief paper published in Lancet (Blass et al., 1976b), the authors measured
pyruvate dehydrogenase in fibroblasts from two Leigh’s patients. These patients
had classic symptoms, as did siblings who died with similar symptoms. And one
of the two had autopsy-confirmed Leigh’s disease. Results showed a 75% decrease
in activity of pyruvate dehydrogenase activity in these two patients’ fibroblasts. The
point is made that Leigh’s disease may have multiple biochemical etiologies and that
care should be exercised in defining the specific cause in each case, as the therapy
may vary depending on this single characteristic.
In another paper published at about the same time (Murphy, 1976), many cases
of Leigh’s disease were reviewed. Most cases fell into two groups: (1) those with
lactic acidosis, and with a focus on Krebs cycle intermediates, and (2) those in
which thiamine and its esters were studied. In this review, in contrast to the previous
paper (Blass et al., 1976b), studies were cited in which pyruvate carboxylase was
not changed in several Leigh’s patients. Murphy states that in all Leigh’s patients
in whom the inhibitor was tested, it was present. Furthermore, when thiamine phos-
phate esters were compared in the brains of controls to the levels in Leigh’s patients’
brains, the levels of thiamine diphosphate were significantly elevated. Further, the
levels of thiamine triphosphate were significantly decreased. These changes seem to
be specific for brain, as liver levels are not altered in patients. This author speculates
that either the increase in thiamine diphosphate or the decrease in TTP is significant
and has some as yet undefined role in CNS function.
Because of the supposed involvement of altered thiamine metabolism, numer-
ous therapeutic attempts have been tried over several years. The most common
is to administer thiamine in high doses in the form of thiamine disulfide or thi-
amine tetrahydrofurfuryl disulfide. Initially, there is usually a more or less dra-
matic improvement in symptoms, weight gain, and a lowering of inhibitor levels
in the urine. Later, however, in spite of thiamine administration, the symptoms and
inhibitor return, and the downward course continues.
Reasoning that a block in thiamine metabolism relating to its role as a coen-
zyme could slow the Krebs cycle, leading to decreased energy production, the Krebs
cycle was fed from a different site (McCandless and Hodgkin, 1977). In this study,
a brother and sister, both with classic Leigh’s symptoms, and one with a positive
urine inhibitor test were studied. These two Leigh’s patients were administered
L -glutamine, which was designed to feed the Krebs cycle through glutamate and
alpha-ketoglutarate, thereby partially bypassing thiamine-requiring steps. Results
from this study were equivocal, and similar to those using thiamine treatment.
There was a significant weight gain, and small improvements in symptoms, which
lasted several weeks. Then occurred a slow return to the pretreatment condition
and continued deterioration. These two patients were far along in their course when
this therapy was attempted, and treatment with L-glutamine would be worth trying
immediately after suspicion of diagnosis.
In 1978, a paper was published (Devivo et al., 1978) in which a careful study
of a Leigh’s patient’s biochemical changes were studied. In this case, there was
also careful postmortem documentation of the neuropathological lesions. A distur-
bance of the oxidative decarboxylation of pyruvate was suggested by lactic acidosis,
hyperglycemia in response to IV alanine, and an elevated flux rate for glucose.
During two hospitalizations, blood lactate levels were elevated. Pyruvate was
also elevated, but not as dramatically. Administration of alanine produced a hyper-
glycemic response. Fibroblast enzyme activities were measured in the patient and
parents and were found to be similar to those in blood.
Postmortem measurement of liver and brain pyruvate dehydrogenase showed that
the patient had lowered levels of tissue enzyme as compared to controls. These
measurements were made regionally and showed a greater decrease in the brain
stem as compared to the cerebral cortex. These changes were shown to be localized,
as are the neuropathological changes in this and other cases of Leigh’s disease.
“Mixing” experiments were also performed on postmortem tissue. In these stud-
ies, liver homogenates from the Leigh’s case were mixed in varying amounts with
control liver homogenates to determine if there was any inhibitory activity present.
Results showed an incremental decrease in activity in the control enzyme activ-
ity with addition of Leigh’s homogenate. The decrease only reflected the “dilution
effect” as would be predicted, not a quick and total depletion expected if there was
an inhibitory effect.
Liver homogenates from the Leigh’s patient, while showing decreased activity,
were able to be activated in vitro by the addition of magnesium chloride and calcium
chloride to the homogenate. While the slope of activation was slow, as compared to
control values, the fact that it was capable of reactivation suggests that the catalytic
components of the pyruvate dehydrogenase complex are present and functional. The
“mixing” experiments also argue in favor of this, and argue strongly against the
inhibitor concept as pathognomic for Leigh’s disease, or even being important.
Results from neuropathological studies showed characteristic lesions as pre-
viously described in Leigh’s disease patients. Thus, lesions were found in the
putamen, caudate nucleus, hypothalamus, substantia nigra, and pontine tegmentum.

The mammillary bodies were spared. Microscopically, lesions showed diminished
neuropil, neuron loss, demyelination, and an enlargement and increase in numbers
of capillaries. In some areas, neurons seemed relatively spared as compared to sur-
rounding neuropil cells. The cerebral cortex was spared, and only minor changes
occurred in the spinal cord fasciculus gracilis. These characteristic changes are seen
in nearly all Leigh’s patients studied.
The authors state that this study suggests dichloroacetate should be considered
as a possible treatment. Dichloroacetate has been effective in previous studies in
treating lactic acidosis. This study is important because novel biochemical studies
were carried out before and after death in an autopsy-proven case of Leigh’s disease.
In another interesting report from Japan (Toshima et al., 1982), a case of Leigh’s
disease, documented at autopsy, was described. The patient was an 11-month-old
boy with muscle hypotonia, seizures, and retarded development. Upon admission,
the patient had lactic acidosis (elevated threefold), and pyruvate was elevated about
the same degree. Also elevated was alpha-ketoglutarate, and alanine was elevated
in the urine. Following a glucose tolerance test, lactate/pyruvate levels increased
abnormally. Many other enzyme levels were normal, as were CSF values.
A muscle biopsy, taken at 17 months of age showed no evidence of necrosis, or of
regeneration. There were no “ragged red” fibers noted. Lipid droplets were seen both
with light and electron microscopy. Mitochondria appeared normal. Blood pyru-
vate dehydrogenase activity was only about 25% that of controls, whereas alpha-
ketoglutarate levels were unchanged.
Treatment in this patient consisted of thiamine therapy and therapy with a
low carbohydrate–high fat diet. Following 3 weeks of thiamine therapy, there
was no improvement in either biochemical or clinical criteria. Therapy with the
low carbohydrate–high fat diet for 6 months resulted in a decrease in elevated
lactate/pyruvate levels, to nearly control values, and significant clinical improve-
ment. During this trial, weight gain was normal, and the patient’s muscle weakness
decreased. He became able to sit, and reach for toys. Later, his development quotient
dropped, and he died at 25 months from pneumonia.
Results from histopathologic examination at autopsy showed coagulation necro-
sis and capillary proliferation. The lesions were bilaterally symmetrical and sharply
demarcated. Lesions were found in the globus pallidus, caudate, substantia nigra,
thalamus, hypothalamus, reticular formation, and cervical spinal cord. Lesions not
as striking were present in the cerebral cortex, cerebellum, and putamen. The mam-
millary bodies were largely unaffected (Figs. 1 and 2).
Nerve cells were involved in some areas, but relatively spared in other areas.
Nerve cell changes included pyknosis, karyorrhexis, and eosinophilia. Also present
was a striking symmetrical demyelination of nerve cell fibers. The demylination was
most pronounced in the reticular formation throughout the brain stem, and cerebellar
peduncles. Less demyelinated areas were the internal capsule, cerebral peduncles,
and the optic nerve. Also noted was a non-specific meningitis, and it seemed to be
related to the lesions of Leigh’s disease. This patient was on a respirator for 3 weeks
before death, but no lesions of “respirator brain” were found at autopsy.
Fig. 1 Photograph of Leigh’s brain slice showing lesions in the caudate, putamin, and globus
pallidus, but with striking sparing of the mammillary bodies. Figure courtesy of Dr. Michael
D. Norenberg
Tissues sampled at autopsy (3 hours after death), were assayed for pyruvate
dehydrogenase/decarboxylase activity. Activity measured less than 10% of control
values. The enzymes were measured in liver, kidney, cerebral cortex, and the cere-
bellum.
The idea for treatment with a low carbohydrate–high fat diet is based on the fact
that acetyl-CoA can be formed from fatty acids, thereby generating molecules for
the TCA cycle which could enter past the diminished pyruvate dehydrogenase com-
plex. Since the patient did show some transient improvement is encouraging, but the
cycle did slow down later did occur, with the usual fatal outcome. Similar transient
improvement has been observed before. These investigators did demonstrate that
the defect in activity of the pyruvate dehydrogenase in brain and liver was similar,
whereas activity of pyruvate carboxylase in liver was much higher than that in brain.
The authors suggest that a key objective in therapy is to maintain low lactate levels
in CSF, indicating an absence of lactic acidosis.
The pyruvate dehydrogenase enzyme complex contains three components and
also contains several gene products which may be defective in Leigh’s disease
(Kerr et al., 1987). The first enzyme component, pyruvate dehydrogenase (E1) is
a thiamine-requiring enzyme, and has been found defective in several cases consis-
tent with Leigh’s disease.
In the case presented in this paper, a newborn baby was found to have metabolic
acidosis, which was corrected. Blood lactate was elevated about threefold. At the
time the patient was hypotonic and had diminished auditory response. At the age
Fig. 2 Photomicrograph of
Leigh’s brain, lower medulla,
fourth ventricle showing
symmetric cavitation, and
striking vascularity. This slide
has a reactive look, and it is
symmetrical. Figure courtesy
of Dr. Michael D. Norenberg
of 3 months the patient was hospitalized following a brief seizure. At that time, he
had strabismus and poor head control. His blood lactate was elevated. Seizures were
controlled with phenytoin and phenobarbital, and he had two normal EEGs. Because
of the metabolic acidosis, vegetable oil was added to the diet to provide 60% of
energy as fat. From 6 to 12 months, he gained weight slowly, and was lethargic.
Finally he became markedly lethargic, and had severe metabolic acidosis. His blood
lactate levels remained high. He subsequently had a respiratory arrest, was placed
on a respirator, and died 2 days later. An autopsy was performed immediately.
Autopsy results showed no evidence of cardiac or skeletal muscle involvement
either grossly or microscopically. The CNS showed symmetrical focal areas con-
sisting of cystic necrosis in the putamen and caudate nucleus. The hypothalamus,
mammillary bodies, midbrain, and pons appeared normal. Microscopically, neu-
ronal loss and astrocyte proliferation were not seen. There was mild hypomyelina-
tion in the cerebral hemispheres and in the optic nerves and posterior columns of
the spinal cord. There was also evidence of acute anoxic-ischemic damage. In bio-
chemical studies, PDC deficiency was noted in lymphocytes from blood samples.
The degree of deficiency was such that lymphocytes from this patient had activity
levels of only about 10% of controls. Tissue samples consisting of liver, heart, kid-
ney, and skeletal muscle were harvested within 1 hour of death, and mitochondria
were prepared. These mitochondria showed high respiratory control rates, indicat-
ing viability. However, the mitochondria from the four tissue types showed a lack of
ability to oxidize pyruvate as compared to controls. In the brain and kidney, whole
tissue homogenate’s PDC activities were about 1–4% that of controls.
The activity of the three enzymatic components of the PDC complex were mea-
sured in fibroblasts, liver, and muscle mitochondria and compared to controls.
Results showed activity of E1 in liver was almost zero, and only half in fibroblasts
compared to controls. Western blot analysis showed no immunoreactive protein for
E1 alpha or E1 beta as compared to controls in samples from liver or brain, and only
barely detectable in fibroblasts. By contrast, E2 and E3 components had activity
levels comparable to control tissue samples.
The almost complete absence of activity of the alpha/beta subunits of the E1 com-
ponent of PDC accounts for the overall loss of activity in this patient’s tissues. It is
the E1 component, which is the rate-limiting step in this complex. The authors note
that autopsy results are in keeping with a diagnosis of Leigh’s disease. There are
some variations in lesions between cases, and in this case, symptoms were present
very soon after birth. The rate of onset and rate of progression clearly affects neu-
ropathologic findings. The reason for the reduction of amounts of both alpha and
beta components of the E1 complex is not clear. One would expect that a single
point mutation would only affect one component. In other published cases, both
units have been found to be deficient (Ho et al., 1986), or only the alpha unit was
diminished.
Cytochrome c oxidase deficiency is another genetic defect which has been con-
clusively demonstrated in patients with Leigh’s disease (Miranda et al., 1989). This
enzyme is a complicated one, with at least 13 subcomponents, some of which are
encoded by mitochondrial DNA, others by the nuclear genome. The authors stud-
ied fibroblasts from a Leigh’s patient who had developed symptoms at the age
of 7 months. His parents were clinically normal, but a sibling died with autopsy-
confirmed Leigh’s disease. Cytochrome c oxidase (COX) activity in fibroblasts was
about 18% of normal.
Results showed fibroblasts from the patient after several tissue culture doublings
were flattened (instead of spindle shaped), and many contained cytoplasmic vac-
uoles. SV40 transformed cells from the patient also showed decreased COX activity,
similar to untransformed cells. The authors found normal COX activity in hybrids
derived from COX-deficient CX3 cells from a Leigh’s patient, and COX-positive
HeLacot cells, in which the HeLa mtDNA had segregated. The authors interpret this
to indicate that the defect in this patient was due to a mutation in nuclear DNA.
These results are in general agreement with another study showing no decrease in
COX in isolated mitochondria (DiMauro et al., 1987). This study also described
“mixing” experiments in which enzyme activity was unchanged when Leigh’s mito-
chondria were mixed with those of a normal control.
Further Leigh’s disease research revealed genetic transmission of defects in at
least three mitochondrial enzymes, including pyruvate dehydrogenase complex,
COX, and, less frequently, NADH Co Q reductase. A T to G point mutation at
nt 8993 in the ATPase 6 gene of mtDNA has been described in 12 patients with
Leigh’s disease (Santorelli et al., 1993). Symptoms of this defect express as neuro-
genic weakness ataxia and retinitis pigmentosa (NARP). Santorelli et al. determined
that when present in high percentage, the nt 8993 mutation resulted in the Leigh’s
disease phenotype. The mutation was present in all five mothers studied, but not in
maternal grandmothers or aunts. Respiratory chain enzymes were all normal in 11
of 12 patients. The mutation was present in similar amounts in all tissues tested.
A distinguishing feature of this mutation is the presence of retinitis pigmentosa.
This symptom is not associated at all with COX deficiency or pyruvate dehydroge-
nase deficiency. In spite of the progressive clinical course of Leigh’s disease, the
authors suggest that knowledge of this mutation is very important from the stand-
point of genetic counseling and possibly for prenatal diagnosis. It is necessary that
the phenotype differences between these patients and those with COX and pyruvate
dehydrogenase deficiencies are understood.
In a comprehensive paper by Rahmen et al., the specific etiologies of 35 cases
of Leigh’s disease, and 32 cases with some atypical features, were determined
(Rahman, S. et al., 1996). The goal was to determine what proportion of Leigh’s
patients might have maternal, X-linked, or autosomal recessive inheritance. Mus-
cle samples for analysis were usually taken by open biopsy, or from organ donors
(controls).
Results showed that ten patients had DNA mutations at nt 8993 in the mitochon-
drial genome. Of these, six had a T to G transversion and four had a T to C transition.
Enzyme analysis was performed in fibroblasts, and liver and muscle biopsy speci-
mens. Seven patients were found to have pyruvate dehydrogenase deficiency (six
had the mutation at the E1 alpha site), nine patients had COX deficiency, and 13
had complex 1 deficiency. No patient had pyruvate carboxylase deficiency. There
were some tissue specificity findings in that some pyruvate dehydrogenase deficient
patients had only slightly decreased enzyme in fibroblasts, but much lower levels
in skeletal muscle, Based on their study, the authors give an estimated incidence of
Leigh’s disease as 1 in 77,000 births. In an attempt to correlate the clinical picture
to the type of defect, no correlation could be made.
At this time (1996), not all Leigh’s cases could be assigned to a specific
cell/molecular defect, so further studies should reveal other mutations. The present
results showed that the complex 1 deficiency was a common cause of Leigh’s dis-
ease in this study, in contrast to other studies in which it was infrequent. Deficits in
several places in the “energy production system” can result in SNE. The actual phe-
notype in any given patient cannot be predicted by site of mutation; therefore, the
onset and severity of symptoms, and the course of the disorder must be related to the
degree of impairment of energy producing pathways, and ability of other pathways
to compensate. These results, of course, are based on muscle, liver, and body fluids,
which may not reflect regional cerebral conditions.
Yet another possible cause of Leigh’s disease is mitochondrial DNA depletion
(Absalon et al., 2001). In this paper, a patient is presented who was hypotonic at
birth and needed resuscitation for several minutes. Physical examination showed
poor spontaneous movement of facial muscles, extensor posturing of upper extrem-
ities, flaccid lower extremities, and absence of deep reflexes. There was a primary
metabolic acidosis. Brain CT scans were normal on days 1 and 7. An MRI on
postnatal day 35 showed signals consistent with a diagnosis of SNE. Muscle biopsy
was performed, and results showed normal structure with no “ragged-red” fibers.
Histochemical stains showed a pale staining of mitochondrial succinic dehydroge-
nase and COX. A mutation screen of mitochondrial DNA showed no deletions or
point mutations. A southern blot showed that the quantity of mitochondrial DNA
was less than 5% of normal control mitochondrial DNA. The patient died on day
40, and an autopsy was denied.
The authors comment that mitochondrial DNA depletion has not been care-
fully considered as a cause of SNE, and only one previous case has been reported
(Morris et al., 1996). The molecular basis of mitochondrial DNA depletion is as yet
undetermined. Some studies suggest that mitochondrial transcription factor A may
be lacking or non-functional, resulting in the mitochondrial depletion syndrome.
These authors again stress the importance of defining the exact molecular mecha-
nism responsible for Leigh’s disease in each individual case. Genetic counseling is
very important to parents of children with Leigh’s disease, and this process is facil-
itated by knowledge of the nature of the exact defect of the disorder. The authors
state that mitochondrial DNA depletion is clearly a cause of Leigh’s disease and
should be considered as a significant cause than has previously been thought.
Another mitochondrial DNA point mutation involving A 8344 G is of interest
because phenotypic expression can be quite varied (Berkovic et al., 1991). In a
short paper by Tsao (Tsao et al., 2003), a well-documented case of Leigh’s disease
was associated with the A 8344 G mutation. This patient was born with hypotonia
and postnatal developmental delay. He sat at 8 months of age, stood at 12 months,
and was able to say five words at 20 months. At that time, he was admitted with
asthma and breathing difficulties. He soon showed respiratory failure, lactic aci-
dosis, and progressive mental decline. Over 3 months, his neurological regression
led to the loss of all his motor skills. The patient developed Leigh’s symptoms,
including muscle weakness, external ophthalmoplegia, ptosis, deafness, and a high
lactate/pyruvate ratio. He became lethargic and stuporous. Pneumonia led to respi-
ratory arrest and death.
Cerebral MRI at 20 months of age demonstrated symmetric hyperintense T2-
weighted signals from the putamen, claustra, thalamus, midbrain, and medulla.
These signals were bilaterally symmetrical. The spinal cord was also affected. Mag-
netic resonance spectroscopy revealed increased lactate in the putamen. Two months
later, the lesion size was described as smaller in the putamen, claustra, and medulla,
whereas the lesions were larger in the thalamic nuclei, midbrain, and pons. Liver
and muscle biopsy tissue showed minimal COX decrease, no ragged red fibers, and
enlarged mitochondria. All other enzyme levels were normal, including pyruvate
dehydrogenase.
The A 8344 G mutation is usually associated with epilepsy and ragged red mus-
cle fibers. Increased lactate, deafness, and ataxia are also associated with this muta-
tion. The association of Leigh’s disease with the A 8344 G mutation represents an
atypical phenotypic manifestation. Several other non-typical phenotypes are associ-
ated with this point mutation, such as Charcot-Marie-Tooth disease and the Ekbom
syndrome. Leigh’s disease can result from a seemingly wide variety of genetic muta-
tions, all of which involve the Krebs cycle, or the respiratory chain. This case points
out the diversity of causes of Leigh’s disease, and stresses again the importance of
looking for the cause. Some of this patient’s relatives also had the A 8344 G muta-
tion, but had almost no symptoms. This points out the complicating fact that not all
individuals with a given point mutation will be symptomatic. How this dichotomy
can occur is a complete mystery.
Another paper (Benit et al., 2004) showing a late onset Leigh’s disease case
reflects the heterogeneity of both the causes and the phenotypes. This was a case
of a patient who had a single episode of febrile seizures at the age of 9 months, and
then at 9 years, the patient developed a persistent stiff neck. At that time, MRI, EEG,
and EMG were normal.
One year later, the MRI showed a high T2 signal intensity in the putamen, white
matter, and brain stem. CSF lactate was elevated, and a mitochondrial complex 1
deficiency was found in a muscle biopsy sample. Two years later, the patient devel-
oped acute pancreatitis and respiratory insufficiency. The patient expired 18 months
later following an acute multisystem failure (Figs. 3 and 4).
Polarographic studies showed low pyruvate plus malate oxidation in skeletal
muscle mitochondria from the patient as compared to controls. Activity ratios
showed a decrease in complex 1 dependent activities. RT-PCR and D-HPLC anal-
ysis of mRNA from skin fibroblasts further characterized the defect. These defects
occurred in the NDUFS3 gene subunit involved in the catalysis of electron trans-
fer from NADH to ubiquinone. Of the complex 1 subunits, some defects produce
Leigh’s symptoms; defects in other subunits can produce cardiac symptoms inde-
pendent of Leigh’s disease. It is possible that only some mutations of subunits
produce super oxide over production, which is known to stimulate cardiomyocyte
hypertrophy.
Fig. 3 Photomicrograph of Leigh’s brain showing hypervascularity. Figure Courtesy of

Dr. Michael D. Norenberg
Fig. 4 Photograph of Leigh’s brain slice showing brownish lesions in the midbrain. Figure cour-
tesy of Dr. Michael D. Norenberg
Fig. 5 T1-weighted axial

image showing swollen
hypointense putamina with
areas of hypointensity within
the lesions. A 1.5-year-old
girl with Leigh’s disease.
Reproduced from Valanne
et al. (1998) with permission
from the American Society of
Neuroradiology
Fig. 6 T2-weighted axial

image showing hyperintensity
and swelling in the putamina.
Myelination delayed
1.5-year-old girl with Leigh’s
disease. Reproduced from
Valanne et al. (1998) with
permission from the
American Society of
Neuroradiology
The above report of a defect in the NDUFS3 subunit represents the last subunit
of complex 1 to have a phenotypic reflection of the mutation. All 14 genes (seven
mitochondrial, seven nuclear) have disease-causing mutations, several of which may
result in Leigh’s disease. Other so-called “supernumerary” subunits (at least 28 in
number) are associated with complex 1. The functional role of these remaining
supernumerary subunits is unclear, but they may be involved in tasks such as reg-
ulation of activity of complex 1, protection of complex 1 from oxidative distress,
assembly of components, etc. This paper is another example of the very high het-
erogeneity of causes of Leigh’s disease and of the necessity for trying to determine
the specific cause of the disorder, at least for counseling purposes.
A recent paper (Hoefs, S. et al., 2008) reports a patient with a mutation in the gene
NDUFA2 of the NADH:ubiquinone oxidoreductase (complex 1), the first complex
of the respiratory chain. This complex consists of 45 subunits, and it has three func-
tional fractions. The patient in this study was a product of consanguineous parents.
An older sibling was normal.
This patient had delayed development from birth and at day 5 had cardiomyopa-
thy. By 4 months, cerebral atrophy and hypoplasia of the corpus callosum showed
on examination of MRI images. At 7.5 months the patient developed acidosis, gen-
eralized clonic-tonic seizures, and lapsed into a coma. The patient never regained
consciousness, and died 43 days later.
Multiple tests were done on this patient. Results showed a defect in mitochondrial
energy generating capacity in which there was a decrease in pyruvate oxidation and a
decrease in ATP production in the patient’s muscle tissue. Immunoblot data showed
a decrease in protein level of NDUFA2 in the patient as compared to controls. The
mutation in NDUFA2 was associated with the depolarization of mitochondria in
living cells. This decreased mitochondrial membrane potential could be reversed
by baculoviral complementation. This potential is directly coupled to the normal
function of the respiratory chain.
The authors state that complex 1 deficiency is one of the most common defects
of the oxidative phosphorylation system. Few treatments are available for these
defects, making prenatal diagnosis and genetic counseling very important. The mito-
chondrial membrane potential could be a hallmark of complex 1 deficiency, making
diagnosis more certain.
In another recent study (Sijens, P. et al., 2008), two patients with Leigh’s disease
were examined with MRI, and lactate was found in gray and white matter, and
high choline levels were in white matter only. In the first case, symptoms began to
appear at about 9 months. The patient had nystagmus and hypotonia with pyramidal
signs. Muscle biopsy showed decreased complex 1, 2, 3, and 4 enzyme activity.
This patient died at 3 years of age. The second patient had respiratory problems at 4
years, and examination showed about a 1.5-year developmental delay. Blood lactate
was high and he showed a reduced activity of complex 1 and 4, plus diminished
muscle ATP production.
Specific results from MRI showed inverted lactate doublet signals were increased
in both cases. Choline peaks were high in both patients’ white matter. The combi-
nation of biochemical data and MRI results led to the Leigh’s disease diagnosis in
each case. Control patients were those who were not suffering from Leigh’s disease,
but who had epilepsy, extrapyramidal symptoms, etc.
The findings in these two patients are in keeping with the symptoms and course
of Leigh’s disease. The authors state that the MRI findings imply altered energy
metabolism throughout the entire brain. While choline was elevated in both patients’
brains, only one had demonstrable lesions in white matter. While Leigh’s disease is
thought of as a gray matter disease, these two cases reaffirm the idea that white mat-
ter pathology may be occurring. The data suggest that these MRI findings (elevated
lactate and choline) could serve as useful diagnostic methods in determining the
presence of Leigh’s disease.
Leigh’s disease and MELAS (mitochondrial encephalopathy, lactic acidosis, and
stroke like episodes) have steadily increased in terms of diagnosis and reported cases
(Shanske, S. et al., 2008). In these disorders, the G13513A mutation is in the mito-
chondrial DNA NO5 gene. Results from studies of patients with these disorders
showed most but not all had decreased respiratory chain activities at complexes 1,
3, and 4. The ND5 gene is a locale for mutation, and the G13513A mutation was
found in 26 of 33 patients. This mutation must be considered in cases of MELAS,
or in Leigh’s disease. Keys to this mutation include late onset, lack of a family his-
tory, and normal or mildly decreased complex 1 activity. It is important to examine
multiple sites of maternal tissue in an effort to find the mutation so as to provide
good genetic counseling.
African Seasonal Ataxia
Mark 1:6 And John . . . did eat locusts and honey.
Since biblical times man has consumed insects as a primary food source
(entomophagy). Insects are, of course, also consumed as a delicacy in many cultures
and served in fine restaurants around the world. In Western Africa, and especially
in Nigeria, a silkworm larvae, Anaphe venata, is consumed in large numbers during
the rainy season of July–October. These larvae serve as a primary protein source for
hundreds of thousands of indigenous people during this period. Over several years,
it became apparent that there were neurological manifestations occurring during the
rainy period in those consuming Anaphe venata larvae and that this syndrome had
many features of thiamine deficiency.
The Anaphe venata is a larval stage of the butterfly Lepidoptera: Notodontidae.
This larvae serves as a defoliator of Triplochiton scleroxylon, which is a forest tree
growing in Western Nigeria. During the rainy season these larvae fall from the trees
and are easily collected, processed, and sold in the markets (Figs. 1and 2).
Before any correlation of the consumption of Anaphe venata and neurological
signs and symptoms, a study of the food value of the larvae had been done (Ashiru,
1988). The purpose of the study was to determine the nutritional value of Anaphe
venata larvae which were being eaten during the rainy season in Nigeria. The larvae
are collected and sold in small rural markets.
The investigator in this study had a dual purpose. The first one was to chemi-
cally measure the nutrients present in Anaphe venata larvae, and the second was
to determine the extent of human consumption of the larvae. The food value of
Anaphe venata was assessed by collecting several hundred larvae and removing the
stiff hairs (setae) on the body by passing the larvae through a flame. The larvae
were then dried, and powdered in a mortar and pestle and analyzed. Initially the
calorific value of the larvae was compared to that of eggs. Results showed that the
calorific value fell between that of egg yolk and white. A quantitative nutrient analy-
sis showed that the larvae contained crude protein, fat, ash, manganese, copper, zinc,
and phosphorus. When larvae were compared to beef, pork, lamb, and chicken, the
larvae had higher protein content than lamb and pork. Six of the eight essential
amino acids were found in Anaphe venata. Eight other non-essential amino acids


Fig. 1 Anaphe venata larvae generously supplied by Dr. Bola Adamolekun
were identified in the larvae. From these biochemical analyses of nutrients present
in Anaphe venata, it is clear that they are an excellent source of nutritional value.
In an attempt to determine awareness of the larvae and their use as a food source,
questionnaires and interviews were held with people in Nigeria. Results showed that
80% of those interviewed knew the larvae were edible, and nearly 70% had either
eaten them or had family members who had consumed the larvae. Before consump-
tion, the setae had been removed by drying the larvae in hot sand. They were then
dried in the sun for preservation. When eaten, they were frequently cooked in stew.
Entomophagy in the interviewed group had taken place for a range of 9–50 years.
This author concludes that Anaphe venata larvae are indeed an excellent source
of protein in an otherwise protein-deficient region of Africa. The larvae are only
available for a 4-month period, although after drying, they have excellent storage
African Seasonal Ataxia 105
Fig. 2 Anaphe venata larvae generously supplied by Dr. Bola Adamolekun
potential. The author suggests that they would make a useful commercial source for
protein, and mass rearing of Anaphe venata should be considered.
Descriptions of short-lived epidemics consisting of ataxia were first described
in western and north-western sections of Nigeria (Wright, F. and Marley, D., 1958;
Coakham, H., 1972; Osuntokun, B., 1972). These brief reports described clinical
features of what we now know as African Seasonal Ataxia (ASA), but before the
correlation with thiamine deficiency had been discovered. In the report by Osun-
tokun, it was emphasized that the disorder was not “flu-like,” but rather seemed to
have a rapid onset a few hours after a meal, usually consisting of yams. In 1971
there were hundreds of cases throughout Western and Northwestern Nigeria. The
disorder was described as self limiting, and the first symptoms were nausea and
vomiting. The most prominent symptom was a course tremor, with signs of cere-
bellar involvement. Altered consciousness was present in at least 40% of patients,
and ranged from confusion to stupor and coma. Other symptoms included echolalia,
myoclonic jerks, abdominal pain, cogwheel rigidity, pin point pupils, etc. There was
no evidence of viral infection. The etiology was suspected to be a form of intoxica-
tion resulting from a cholinergic stimulation by an unknown substance in the yams.
The author speculated that there could be a genetic link in the disorder and that there
was no evidence of vitamin deficiencies or intoxication by cyanate or cyanide.
In 1992, Adamolekun was the first to describe ASA in Western Nigeria, and
correlate it to thiamine deficiency (Adamolekun, B., 1992). His clinical observa-
tions were that the syndrome occurs acutely after patients consumed a carbohydrate
meal. Symptoms include dizziness, vomiting, ataxia, and altered levels of conscious-
ness. The affliction occurs in males and females equally, and in all age groups. The
impairment of consciousness can end up in overt coma.
Adamolekun states that he was impressed by the similarities between patients
presenting with ASA and the symptoms described by Wernicke in 1881 as acute
thiamine deficiency (Wernicke, 1881). These classical acute symptoms of thiamine
deficiency include ataxia, nystagmus, and alterations of consciousness. Adamolekun

noted that the symptoms responded well to Parenterovite 1 and 2, a preparation
containing high concentrations of vitamins.
Adamolekun speculates in his first paper on ASA that the patients developing
symptoms were marginally thiamine deficient due to the consumption of something
possibly containing a thiaminase. The seasonal nature of the affliction suggested the
possibility of a thiaminase in the larvae of seasonal insects. A large carbohydrate
meal could then trigger the symptoms by placing an unusually high demand for
already diminished thiamine reserves.
Soon after the above observations (Adamolekun, 1993), the correlation between
ASA and the Anaphe venata larvae was made. In those studies, the author inter-
viewed 29 hospitalized patients with a diagnosis of ASA who had developed symp-
toms during the 1991 and 1992 seasons. All patients were of a low socioeconomic
level, and were consuming high-carbohydrate meals, and were protein deficient
(mean albumin 32.5 g/L, normal range 35–40 g/L). All patients had consumed
roasted Anaphe venata larvae as part of their meals immediately before the onset
of symptoms. Adamolekun noted that the silkworm Anaphe venata moults only
once per year and that correlates with the rainy season. The larvae are easily col-
lected, and are sold in the local markets. The author further states that thiaminases
are present in many insects and that the interviews of patients suggesting Anaphe
venata entomophagy and the availability of the larvae as a protein source were
highly suggestive.
In 1993 there was an outbreak of ASA, again in Western Nigeria during the
rainy season, which was well documented (Adamolekun, B. and Ibikunle, F., 1994).
This outbreak took place in a small (population 60,000) town consisting largely
of farmers and merchants of low socioeconomic means. The authors verified the
diagnosis of ASA in 34 new admissions to five local hospitals. Careful histories were
taken and the meal eaten just prior to the onset of illness was noted. Symptomology
and a neurological exam were undertaken.
Of the 34 confirmed cases, there were 8 men and 26 women; median age was 29
years. Eight patients were children under 8 years of age. All adults were farmers or
small traders of a low socioeconomic class. Presenting symptoms included those
previously described, including nausea, vomiting, dizziness, tremor, ataxia, and
confusion. Clinical signs included nystagmus, cerebellar ataxia, intention tremors,
dysarthria, and impaired consciousness. All patients reported consuming roasted
Anaphe venata larvae. The registered hospitals in the area reported 1126 admissions
for ASA in the 1993 season, a rate of 1.87% in this town.
Patients were treated with IV Parentrovite, a multivitamin solution which con-
tains 250 mg thiamine per dose. Hospital staff were trained in recognizing the dis-
order and in educating patients and family members of the risks of eating the larvae.
It was hoped that the information would be spread among others consuming the
larvae, which might decrease the entomophagy.
There was a large increase in ASA during the rainy season in 1993 as compared
to previous years. This was attributed to a high inflation rate which made purchase
of other protein food sources more difficult. Clinically, all had cerebellar ataxia,
and some showed basal ganglia and autonomic dysfunction. There were no deaths
attributed to ASA during this reporting period. The incidence rate of 1.87% was
deemed falsely low since many afflicted people may have chosen “traditional” non-
hospital treatments. As it was, the epidemic accounted for over 70% of all hospital
admissions in the region for the month of August.
In another study, a double-blind placebo controlled study was performed exam-
ining the efficacy of thiamine as a treatment modality for ASA (Adamolekun, B.
et al., 1994). Fourteen patients were assigned randomly to either a placebo group
or to a thiamine hydrochloride group. Severity of clinical signs was assessed on
admission to the hospital, at the start of therapy, and at every 24 hours for 72 hours.
All patients were unable to stand or sit up in bed before therapy started. Patients
showing no improvement at 72 hours were treated with multivitamin therapy IV.
Eight patients were treated with thiamine hydrochloride, and six with placebo
(lactose). Results showed a dramatic and constant improvement in neurological
signs such that in five of the eight thiamine-treated patients, the neurological
signs were completely reversed by 72 hours. The remaining three were completely
reversed in the immediate post-study period. All patients on placebo deteriorated
slightly over the 72-hour period. The improvement in thiamine-treated patients, and
the differences between treated and placebo patients was statistically significant.
These results unequivocally implicate thiamine deficiency in the generation of
symptoms in ASA. The patients with this disorder consume diets with high levels
of carbohydrates which may contain thiamine binding cyanogenic glycosides. Then,
during the rainy season when protein-rich Anaphe venata larvae fall from the trees,
thiaminases present in the larvae are consumed in large amount. This effectively
pushes the patients into overt thiamine deficiency producing the ASA epidemics. In
fact, a history of monotonous carbohydrate diet and Anaphe venata entomophagy
was present in all patients examined. The rapid reversal of symptoms with thiamine
administration suggest the classical “biochemical lesion” of thiamine deficiency is
present before structural changes occur. This rapid reversibility is similar to that seen
in early Wernicke’s disease when symptoms can be reversed with thiamine treat-
ment. It would be expected that with prolonged thiamine deficiency and/or yearly
development of symptoms, structural damage might occur in the brain, as is seen in
the Wernicke-Korsakoff syndrome.
The epidemiology of ASA has also been studied (Adamolekun, B. and Ndububa,
D., 1994). Results from this study showed that of 37 patients, the ratio of males
to females was 1.5–1.0. This ratio could be due to males more likely to come to
the hospital, and more likely to consume Anaphe venata larvae. The peak month
of incidence was August, with some cases appearing in July and September. All
patients became symptomatic from 7 to 12 hours after the last main meal. All had
consumed heavy carbohydrate meals. All had also consumed a stew containing the
Anaphe venata larvae. There was no evidence or history of a familial link to the
disorder, or of transmission from one family member to another. There was no
history of preceding febrile illnesses. The symptoms were the same as described
previously, and consisted of cerebellar ataxia, tremor, nystagmus, dysarthria, and
impaired consciousness. CSF was normal, but nearly one half of patients showed a
diffuse slowing of the EEG. There was no evidence of alcohol or other drug abuse,
which might have influenced these patients. These data continued to point to a thi-
amine deficiency state, and indeed all responded positively to IV vitamin B complex
infusions.
In a review article on ASA (Adamolekun, B. et al., 1997), it was noted that
the high-carbohydrate diet of ASA patients in Western Nigeria was cassava
(Manihot esculenta), which contain the cyanogenic glycosides linamarin and
lotustralin (Jackson, F. et al., 1992). This same diet is consumed in Zaire, Tanzania,
Senegal, Uganda, and Mozambique. Interestingly, several neurological disorders
have been associated with ingestion of cassava/cyanogenic glycosides accompanied
by depleted intake of sulfur containing amino acids.
At the time of the Adamolekun review article, 100% of ASA patients had
consumed the roasted larvae of Anaphe venata regularly and in the last meal before
onset of symptoms. Entomophagy is a common socially acceptable way for low
socio-economic people to supplement their diets throughout Africa. Wars and civil
unrest may increase entomophagy in areas where it might not ordinarily be present,
i.e., Zimbabwe.
While the Anaphe venata larvae have several essential amino acids, it lacks
sulfur containing amino acids such as cysteine and methionine. The cyanogenic
glycosides consumed in high-carbohydrate foods need sulfur for detoxification.
Thiamine, which contains sulfur, could serve as a binding site for the cyanogenic
glycosides thereby rendering the vitamin inactive and further contributing to the
thiamine-deficient condition.
The exact nature of the mode of “action” of the Anaphe venata larvae was specu-
lative until the work of Nishimune et al. (2000). As described in this report, Anaphe
venata larvae were collected in Uganda. The dry pupae were extracted with sodium
potassium buffer by grinding in a motor and pestle. Following centrifugation, the
supernatant was used as a crude extract. The extraction was incubated with thi-
amine for 30 minutes and then assayed for thiamine content. Results showed that
thiamine decomposition was time dependent, co-substrate requiring, temperature
dependent, and heat labile. Co-substrates were several, including pyridoxine, ani-
line, and aminopyridine. Based on dialysis experiments, the authors concluded that
the thiaminase was a type 1-base exchanging type of thiaminase. The authors found
the highest enzyme activity was at 70◦ C, showing that it was somewhat heat resis-
tant. Heating for 15 minutes at 100◦ C essentially inactivated the thiaminase. The
molecular weight of the thiaminase was estimated to be around 200 kDa.
This study represents the first description of a thiaminase in insects, although
these enzymes had previously been described in fish and plants. This study con-
firmed the hypothesis advanced earlier by Adamolekun that the Anaphe venata lar-
vae contained a thiaminase. The fact that the thiaminase in Anaphe venata could be
largely deactivated at 100◦ C for 15 minutes suggests that with proper cooking, the
larvae could be a safe protein source.
Another study examined the sub-acute effects on behavior of injected extracts
of Anaphe venata larvae in mice (Iwalewa et al., 2005). Larvae were collected in
Western Nigeria, and prepared for consumption in the usual manor. The larvae were
crushed and extracted in an aqueous form (polar) and an ethyl acetate form (non-
polar). Results from these two extract forms were not significantly different, and can
be considered together. The idea was to describe and analyze such acute signs and
symptoms such as mild tremor, jerking, hyperactivity, grooming, rearing, sniffing,
etc. These signs and symptoms began to show after 7 days of IP injection of the
extracts of the Anaphe venata larvae.
Results showed that many so-called sub-acute neurotoxic behaviors were not
decreased by heat during the processing of the larvae. The internal organs of the
mice injected with larvae extracts were not damaged. Observations on organ weights
included brain. Results suggested that additional sub-acute toxic effects would pro-
duce the full ASA syndrome. This could be expected to occur over a period of
months during the rainy season when the consumption of larvae is high.
The observation that mice could be rendered “thiamine” deficient in about 7 days
has been noted in a separate study (McCandless, D., 2004). In this study, 25–30 g
mice were injected IP with an Anaphe venata extract made by pulverizing 1 g of
larvae in 2 ml of saline. This mixture was centrifuged and supernatant injected at a
dose of 0.1 ml per mouse per day. Mice were also injected with a low dose (50 µg)
of pyrithiamine per day. This low dose regime produces neurological symptoms
in mice in 10–12 days. In mice injected with both the low dose of pyrithiamine
and Anaphe venata larval extract, weight loss, ataxia, rolling over movements, and
ruffled fur were clearly apparent in 7–8 days. Mice became stuporous, and death
occurred by days 9–10. This suggested that a rodent model, similar in symptoms to
the human counterpart, was easily available. It is anticipated that future neurochem-
ical studies on brains of animals rendered thiamine deficient by extracts of Anaphe
venata larvae will show similar results to animals with thiamine deficiency of a more
“classic” nature (McCandless, D. and Schenker, S., 1968).
The incidence of Anaphe venata induced thiamine deficiency in Nigeria has been
reduced primarily due to the efforts of Adamolekun (1992). There was a signifi-
cant movement to educate the indigenous people as to the risks of Anaphe venata
entomaphagy. Handouts were prepared, posters placed, and verbal warnings were
given by healthcare workers as to the possible risks. These efforts were effective in
reducing the frequency of these entomophagy-related outbreaks. They still exist in
Nigeria to a lesser extent, and occur in other countries as well.
As an historical note, over 50 years earlier, another thiaminase was discovered in
silver foxes among those on the Chastek fox farm in Minnesota (Green, R., 1937).
Within a year, the disorder, called Chastek Paralysis, had been experimentally repro-
duced and the neurological symptoms attributed to thiamine deficiency (3&4 paper,
1942). Characteristically the outbreak occurs during the fall/winter months when the
foxes are switched to a diet consisting largely of fish. After a period of 3–6 weeks,
affected foxes start to eat less, then progressively develop neurological symptoms
consisting of abnormal gait, staggering, and appearances of weakness. Within a day
or two, affected foxes develop severe symptoms. These consist of spasticity, severe
ataxia, inability to stand, and hyperesthesia. Seizures frequently occur, as well as
stupor, and death occurs within about 24 hours following onset of the severe stage.
Pathologically, the heart showed some changes including hemorrhages, which

were also seen microscopically. Areas of degeneration were seen in association with
hemorrhages. Connective tissue proliferation surrounding vasculature of the heart
was also seen microscopically. These cardiac changes were however only seen in
about half of the cases. The development of brain symptoms occurred more rapidly,
and produced death before the development of most cardiac pathology.
Brain lesions were described as being symmetrical, highly focal, and distributed
in a manner reminiscent of Wernicke’s disease. In Chastek paralysis, there was pro-
liferation of endothelial cells and alteration of small vessel vasculature. There were
many small hemorrhagic areas. The vasculature changes appeared as tortuous ves-
sels which may have reflected the appearance of small vessels which were actually
there, but not originally apparent. The changes were strikingly similar to those seen
in Wernicke’s disease.
In the medulla of foxes suffering Chastek paralysis, lesions were invariably
present in the dorsal nucleus of the vagus nerve. Characteristic lesions were seen
in the medial vestibular nucleus, and these might extend into other vestibular nuclei.
Lesions may also be present in the restiform body and in the floor of the fourth ven-
tricle. Lesions were found in the superior portions of the nucleus gracilis and nuclus
cuneatus. Hemorrhages were noted in the reticular formation, but not proliferative
vascular changes. The inferior olives were affected in half of the cases examined.
The bilateral symmetry was noteworthy, and the degree of change remarkably sim-
ilar side-to-side.
In the cerebellum, lesions were almost always located in the folia around the
forth ventricle. Proliferation and dilation of vessels were prominent in the granular
layer, while hemorrhage was more prominent in the molecular layer. The changes in
the cerebellum, as in the medulla, were almost always localized in the gray matter
and bilaterally symmetric.
In the midbrain, lesions were frequently noted in the inferior colliculus, whereas
the superior colliculus was usually spared. The nucleus of the third cranial nerve is
usually affected, whereas the fourth cranial nerve nucleus is spared. The red nucleus
and substantia nigra were rarely affected. Sections through cortical areas such as
the occipital and temporal lobes were negative for lesions. Occasional lesions were
seen in the frontal lobes, and these were characteristic vascular changes seen in
other areas of the brain. In cortical lesions, as in cerebellar lesions, the most super-
ficial layers were most dramatically affected. Foci of degeneration and of vascular
proliferation changes were not seen in white matter.
The authors note the striking similarities between the neuropathology of Chastek
paralysis and that of Wernicke’s disease. The symptoms of ataxia, nystagmus, stu-
por, and death are comparable. The authors did not assess eye movement, but lesions
support the concept that there would be such changes. As in Wernicke’s disease,
no lesions were noted that developed outside gray matter, although they may have
occasionally extended from gray into white matter. The microscopic changes found
in the small blood vessels of Wernicke brains correspond exactly with those seen
in Chastek paralysis. These changes in Chastek brains are an irregular dilatation
of small blood vessels which give the appearance of vascular proliferation. In both
Chastek paralysis and Wernicke’s disease, hemorrhage is noted in the center of the
vascular lesion.
The authors speculate about the localization of lesions in Chastek foxes. They
state that since thiamine is required for oxidative metabolism in the thiamine-
deficient state, pyruvate would be elevated in tissue. Elevated tissue pyruvate levels
could in turn be toxic to gray matter precipitating the neuropathology noted. A small
number of nursing fox pups also suffered from an acute form of the disorder; how-
ever, they had not yet developed structural damage at the time of their demise.
Shortly after the description of Chastek paralysis, the nature of thiamine deacti-
vation in the fish fed to foxes was described (Krampitz, L. and Wooley, D., 1944).
In this study, thiamine was incubated with various amounts of carp viscera extract.
Following incubation, the amount of thiamine remaining was determined. A variety
of paradigms were used, such as varying the pH, using extract of carp viscera as
well as whole carp extract, etc.
Results showed that the destruction of thiamine by carp intestine is an enzymatic
process. The reaction was largely not affected by either temperature or pH. The
enzyme was isolated in 10% NaCl, and is composed of a non-dialyzable, heat labile
component, and a dialyzable heat stable component. The thiaminase is therefore
presumed to be the component of carp intestine which is responsible for the thiamine
deficiency seen in the foxes with Chastek paralysis.
In another thiaminase study, 14 cats were fed a diet of canned commercial cat
food which contained whole fish (Jubb, K. et al., 1956). Weight gain in affected
cats started to decline after the 11th day on the diet. By the 22nd day, animals
had lost 25% of their weight and were barely eating. Ataxia was present a week
later and suggested cerebellar involvement. When held by the rear legs, cats’ heads
were ventroflexed instead of the normal dorsilexed position. Righting reflexes were
abnormal, and occasional seizures were seen. Animals, if untreated, quickly become
stuporous and die.
Brain lesions were highly focal and bilaterally symmetric. The most severe
lesions were found in the inferior colliculus. Gross lesions were also found in the red
nucleus, medial vestibular nucleus, habenular nucleus, oculomotor nuclei, cuneate
nucleus, mammillary bodies, and other areas.
Microscopically, affected brains showed one or more of three characteristic
lesions: varicose dilations of small blood vessels with hemorrhaging, edema, and
vascular hypertrophy and gliosis. The vascular dilations were seen not only in
lesions but throughout the brain. Lesions were localized in gray matter, and when
white matter was involved, it was an extension of the gray matter lesion. Actual
hemorrhages were almost always confined to gray matter. In lesions recovering
from injury, the thickening of perivascular reticulum and hyperplasia of endothe-
lium gave the appearance of new vessel formation. Edema was confined to the gray
matter affected, and rarely seen in white matter. Myelin appeared normal. A glial
proliferation was noted in some nuclei which appeared to have been spared any
damage by the thiaminase containing cat food. Acute neuronyl changes were not
seen within the affected nuclei. Neurons in close association with edema and vascu-
lar hemorrhage were usually destroyed.
The authors note that many fish varieties contain a thiaminase which is usually
found in the fish viscera. Many different fish parts may be incorporated into cat food,
depending on seasonal variations in what is caught. Not all varieties of cat food
will contain the thiaminase. The authors speculate that possible focal brain damage
might be attributed to accumulating pyruvate due to decreased activity of pyruvate
decarboxylase. The striking appearance of the vascular response suggests that it is a
very important aspect of the neuropathological lesion. The lesions strongly resemble
those seen in both Chastek paralysis and to those of Wernicke’s disease. This in itself
strongly suggests a thiamine-based syndrome in these cats.
Inherited Ataxias
The inherited ataxias represent a group of disorders which have various etiologies,
but have ataxia as a constant and major symptom, and are related to thiamine defi-
ciency. The relation to thiamine deficiency is derived from the finding that many
cases have a genetic deficiency in the function of the enzyme pyruvate dehydro-
genase. It turns out that there are many variations of inherited ataxias, each with a
slightly different biochemical lesion, yet materializing in ataxia.
John Blass was one of the first to describe the inherited ataxias and has been
a pioneer in this area. In a case described by Blass, J. et al. (1970), an 8-year-
old boy who had several bouts of ataxia per year since the age of 16 months had
extensive biochemical tests aimed at elucidating his defect. Clinically, most attacks
followed a febrile illness. Symptoms consisted of ataxia lasting several hours to
a week. Examinations revealed a cerebellar and choreoathetoid syndrome during
acute episodes. In addition, the patient showed irregular eye movements and mild
dystonic posturing. These symptoms largely resolved after 4 days.
Pyruvate, alanine, and lactic acid were all elevated in the patient’s blood. Alanine
was also elevated in urine. Using white blood cells and cultured skin fibrobasts, oxi-
dation of radiolabeled pyruvic acid and pyruvate decarboxylase were all decreased
by 80% as compared to controls. Similar studies on the patient’s father showed val-
ues halfway between controls and the patient’s values.
The decrease in pyruvate decarboxylase activity remained low despite addition
of thiamine pyrophosphate to the incubation media. This implies that thiamine
metabolism in this patient was normal. There was no evidence for the existence
of any type of inhibitor of thiamine pyrophosphate. There was an increase in the
oxidation of pyruvic acid in white blood cells in both the patient and in controls.
This was attributed to an incomplete saturation of apoenzymes and coenzymes.
The authors attribute the symptoms in this patient to the decrease in pyruvate
decarboxylase. Elevated pyruvate was the most significant change in the patient’s
blood. The combination of clinical and biochemical alterations in this patient are
uniquely different from those in other similar disorders such as lactic acidosis,
Leigh’s disease, Hartnup disease, etc. Therefore this patient represented a new
description of an ataxia with a biochemical basis.


A follow-up study was published based on further investigation of this case

(Blass, J. et al., 1971). The patient continued having cerebellar ataxic attacks and
choreiform movement disorder. Between attacks he was described as mildly clumsy.
Otherwise, the patient developed normally, and there was no significant history of
head injury, anoxia, toxins, etc., which could have contributed to his condition.
In this particular hospitalization, he was observed to have two ataxic episodes.
Symptoms included an ataxic gait, intention tremor, slurred speech, and irregular
eye movements. Blood levels of pyruvate were elevated, while lactate levels were
usually normal except during some attacks. EEG studies during attacks showed
bisynchronous discharges of slow and sharp wave forms without obvious epilep-
tiform activity. Otherwise, EEG patterns were largely normal. Muscle biopsy was
normal in paraffin sections; in frozen sections, minor granularity was seen, but no
ragged red fibers. Most other lab results were normal.
Two treatment regimes were tried in this patient. First, dexamethasone seemed
to have produced a reduction in the duration and severity of ataxic attacks at an
earlier age, but not later. Thiamine treatment was associated with a lowering of
blood pyruvate. Also, the frequency of attacks decreased from about one per month
to a period of 5 months without an attack.
This patient had a clear inherited defect in the first thiamine-dependent enzyme
in the pyruvate dehydrogenase (PDH) complex. In other examples of a decrease in
the thiamine-dependent component of this enzyme complex, ataxia was present and
blood pyruvate levels were increased. While carbohydrates are the major source for
energy, ataxia was present, but mentation was normal. Not surprisingly, the defect
produced a selective symptom complex.
The authors point out that there seemed to be precipitating events such as fever
and stress which push an already marginal enzyme capacity into overt deficiency.
The thiamine treatment, which had a measurable positive effect, could be therapeu-
tic because of the incomplete binding between cocarboxylase and the enzyme com-
plex. This binding could be enhanced by saturating the system, thereby increasing
the levels of functional complex. The authors point out that a variety of other neu-
rological abnormalities such as Leigh’s disease, Wernicke’s disease, and Friedrich’s
ataxia are associated with many of the above symptoms and biochemical abnor-
malities. Even at this early date (1971), it was suggested that these mitochondrial
diseases may have multiple biochemical aberrations.
In another study of a patient with deficient liver pyruvate carboxylase activity,
the enzyme was measured in leukocytes and skin fibroblasts and found normal
(Brunette, M. et al., 1972). The patient had developed severe seizures 8 hours after
birth. Severe hypoglycemia was recognized, and the patient was placed on glucose
IV Lactate, pyruvate, and alanine levels were elevated. The patient continued to
progress poorly in terms of growth and mentation. Seizures continued, and sporadic
acidosis was noted. Thiamine treatment seemed to have an ameliorative effect by
decreasing metabolic acidosis. Free and total thiamine levels in blood were normal,
suggesting that the beneficial effect might be related to the increased binding of
coenzyme to apoenzyme produced by thiamine treatment. It should be remembered
that blood levels may not accurately reflect brain tissue levels in discrete regions.
Inherited Ataxias 115
In a review paper (Blass, 1972), the author presents data from over 60 patients
and family members showing that parents of patients with pyruvate decarboxy-
lase deficiency have enzyme levels about halfway between controls and patients.
The author states that not all hereditary ataxias are related to pyruvate metabolism
defects. Conversely, pyruvate metabolism is involved in a significant number of
ataxia cases. The author also notes that the properties of PDH appear to be sim-
ilar between organs, lending credence to using data from fibroblasts to estimate
biochemical characteristics of brain enzyme.
Finally, it is noted that biochemical defects of this nature can cause very dis-
crete changes in neural function. In keeping with this is that the significance of
biochemical changes is different in various brain regions. The nature of this varia-
tion is unclear, but could be related to the presence or absence of reserve capacity
of enzyme activity.
Yet another paper (Blass, J. et al., 1972) reports a patient differing from previ-
ously reported cases of inherited ataxias in that there was a partial defect in the
ability of the patient’s fibroblasts to oxidize metabolites through the TCA cycle.
This patient was a severely retarded 3-year-old girl. Development after birth was
delayed. On admission, the patient had muscular hypotonia, poorly coordinated
movements, possible nystagmus, mental retardation, and microcephalis. Blood lac-
tate and pyruvate levels were elevated and thiamine treatment did not lower these
metabolites. Citrate levels were also low. Results showed a possible defect in
the conversion of citrate to isocitrate, and the activity of PDH was lower in the
patient’s fibroblasts. Estimation of the first enzymatic component of the PDH com-
plex indicated it was normal. Therefore, the defect must have been in the other two
components.
This seems to be a case with severe neurological symptoms, which shows a low
conversion to CO2 of pyruvate, palmitate, and citrate. There was also a defect in
the alpha-keoglutarate dehydrogenase enzyme. Although the mechanism of inheri-
tance was not determined, the mother’s consanguinity and decreased PDH activity
indicates a recessive inheritance. The authors point out that a major oxidative defect
could certainly produce major neurological symptoms, as is seen in brief periods of
hypoxia/anoxia, or hypoglycemia.
Salient features of a review article (Blass, J. et al., 1975) describing 13 known
patients with demonstrated abnormalities of PDH were presented. Most patients
presented with a cerebellar ataxia. The movement symptoms were intermittent and
frequently precipitated by fever or other stresses. In 10 of the 13 patients, symp-
toms were clearly identifiable by 1 year of age. More severe neurological symptoms
included microcephaly, hypotonia, seizures, and four patients died before the age of
10. Most had elevated levels of blood lactate and pyruvate, or alanine.
In terms of treatment, a high-fat, low-carbohydrate diet seemed to help one
patient, but four others became worse. Ketone bodies have the potential to bypass
the PDH defect, and feed the TCA cycle. Thiamine treatment at a level of 600 mg
per day had little or no effect on the course of inherited ataxias.
Neuropathological studies on the patient who died in 1 year showed many
cerebral malformations including an absence of the corpus callosum. Three other
patients had diffuse myelin depletion to as little as 10% of controls, and there was
a diffuse depletion of neurons. Lesions associated with Leigh’s disease were not
seen.
Abnormalities in the PDH enzyme complex were identified in most patients.
Enzymes such as alpha-ketoglutarate dehydrogenase exhibited normal activities.
PDH levels of activity in parents were mostly halfway between controls and patients.
Studies of the subcomponents of the PDH complex show that a defect in the first
component (thiamine dependent) produces mild ataxia symptoms, whereas defects
in the second component produce more severe symptoms. This must relate to the
level of impairment of the total activity of PDH. One might expect that a defect in
the second component would produce a greater decrease in PDH complex activity.
There is also the issue of regional PDH reserve activity in areas such as the cerebel-
lar vermis. In this area it has been shown the reserve activity is lower than in areas
such as the cerebral cortex. In keeping with the obvious need for availability of PDH
is the finding that its activity in brain mitochondria is three- to fourfold that found
in liver mitochondria (Jope, R. and Blass, J., 1975).
Another study by the same group (Jope, R. and Blass, J., 1976) examined the
“activity” state of PDH under various conditions. Results showed that ATP con-
centrations or ATP/ADP ratios influence in an inverse ratio the proportion of PDH
in an active form. This was born out examining effects of decapitation, with resul-
tant ischemia. Treatment of mice with amobarbital increased the energy charge and
decreased the active form of PDH. These in vivo mouse studies confirm previous in
vitro mitochondrial studies (Jope, R. and Blass, J., 1975). The authors note that other
factors modulate the activity of PDH, such as the concentrations of Mg and Ca.
One other comment is that the activity of PDH is low relative to the flux of
pyruvate. The PDH activities are around 100–225 micromoles/gram tissue/hour
(McCandless, D. and Schenker, S., 1968; Jope, R. and Blass, J., 1975). Flux of
pyruvate is about 70–140 micromoles/gram tissue/hour. This provides an approxi-
mate threefold reserve, in most brain areas, but it should be remembered that some
areas have a much lower reserve capacity (Reynolds, S. and Blass, J., 1976).
There may also be a relation between PDH activity and acetylcholine synthe-
sis (Gibson, G. et al., 1975). When various inhibitors such as bromopyruvate were
added to minced brain, a decrease of pyruvate oxidation by 5% led to a decrease in
acetylcholine by 7%. This also held true for the inhibitor 2-oxobutyric acid (Blass,
J. and Lewis, C., 1973).
In a paper from France (Saudubray, J. et al., 1976), two sibs are described with
neonatal lactic acidosis associated with a defect in pyruvate carboxylase. Both sibs
showed symptoms within 24 hours of birth. These included hypothermia, hypoto-
nia, anorexia, and metabolic acidosis. Blood values of ammonia and lactate were
elevated. Pyruvate was mildly elevated, and the lactate/pyruvate ratio was ele-
vated. Several enzymes were decreased; most significantly, pyruvate carboxylase
was decreased to as low as 2% of controls.
Although autopsy was refused on the first sibling, permission was granted for
the second. Results showed a widespread diffuse and symmetrical demyelination
of both cerebral and cerebellar white matter. Symmetrical paraventricular cavities

were found associated with the lateral ventricle and also between the dentate nuclei
and the fourth ventricle. If there was any other neuropathological data, it was not
published. The authors state that these brain findings are not the same as is seen in
Leigh’s disease.
Clinically and biochemically, the authors suggest that their two cases of pyruvate
carboxylase deficiency are unique, and have a rapid clinical course, severe metabolic
acidosis, and hyperlactic acidemia. These are all different from Leigh’s disease and
so fall among the many variations of inherited ataxias.
Also under the category of inherited ataxias are a group characterized by
spinocerebellar degeneration, called Friedereich’s ataxia. This disorder may result
from several mutations, but a significant number involve the oxidation of pyruvate
(Blass, J. et al., 1976). And if to emphasize the diversity of symptoms and neu-
ropathological findings, at least one case of clinical Friedreich’s ataxia has been
published in which the neuropathological findings were those of Leigh’s disease
(Dunn, H. and Dolman, C., 1969).
In the description of five cases of Friedreich’s ataxia, all had ataxia develop
around the age of 12. Most patients showed speech slowing. Muscle strength was
decreased, and deep reflexes were decreased. Visual findings included nystagmus,
diplopia, and decreased corneal reflexes.
Results of biochemical studies showed that the activities of pyruvate and
alpha-ketoglutarate dehydrogenases were decreased in disrupted fibroblasts of the
5 patients as compared to 16 control subjects. PDH activity was 43% of controls,
while alpha-ketoglutarate dehydrogenase was 50% of control values. Increasing
substrate or cofactors in vitro did not change the activity rate. These above data
seem to define a biochemical lesion in cases with clear clinical evidence of Friedre-
ich’s ataxia.
The authors speculate that the one common factor in both the above enzymes
is lipoamide dehydrogenase. A defect in this third component of the enzyme com-
plexes could therefore affect both PDH and alpha-ketoglutarate dehydrogenase, in
turn producing the clinical picture of Friedreich’s ataxia. Since the brain relies on
oxidative metabolism for normal function, decreasing activity of these two critical
enzymes associated with the TCA cycle could easily be seen as deleterious to normal
function. Many previous studies support the concept that mild–moderate decreases
in TCA cycle enzymes are associated with neurological symptoms such as ataxia
(Figs. 1–7 are representations of slices and sections from a case of Friederich’s
ataxia. Courtesy of Dr. Roland Auer).
An important paper looking at methods for measuring activity of PDH in fibrob-
lasts has been published (Sheu, K. et al., 1981). This study examined activities
of PDH in normal fibroblasts and in two cell lines originally described by Blass,
from patients deficient in the enzyme PDH. It was shown that activation was pos-
sible by pretreating fibroblasts with dichloroacetate. The concept is advanced that
a wide variety of metabolic situations may have effects on the activity of PDH.
Pre-activation of enzyme activity in fibroblasts with dichloroacetate maximizes the
enzyme, so that low levels on the order of 3% of normal can be measured. Results
Fig. 1 Brain Slice at the level of the caudate nucleus. Courtesy of Dr. Roland Auer
have confirmed that PDH exists largely in an inactive state and needs activation in
order to be measured. The authors point out that in Friedreich’s ataxia, treatment of
fibroblasts aids detection of the deficiency.
The dichloroacetate activation procedure provides a simple reproducible method
to measure PDH activities in inherited ataxia, parents of patients, and normal con-
trols. This should help identify patients at an early stage. These assay techniques
may not be available at all hospitals, and any improvement in assay methodology
is welcome. The authors are working on techniques to use cells such as leukocytes
and amniotic cells.
Central nervous system dysfunction has been described in a case showing con-
genital lactic acidosis, associated with a defect in complex 1 of the respiratory chain
(Moreadith, R. et al., 1984). In this case, a defect in complex 1 on the inner mito-
chondrial membrane was documented in liver, kidney, heart, and skeletal muscle.
The patient was hypoglycemic, and in some respiratory distress within 24 hours
of birth. He had cardiac hypertrophy, and liver enlargement by 6 weeks. Neuro-
logically, the patient was unresponsive and no longer reacting to visual or auditory
stimuli. A computerized tomographic scan and sonogram showed decreased density
of frontal and parietal white matter. There was a concomitant lactic acidosis and
an elevated lactate/pyruvate ratio. Despite several therapeutic attempts, including
thiamine, the patient expired at 16 weeks of age.
Fig. 2 Friedreich’s ataxia case showing cerebellum. Figure courtesy of Dr. Roland Auer
The decreased activity of complex 1 in the mitochondrial respiratory chain

was confirmed in liver, heart, kidney, and skeletal muscle. The presence of cen-
tral nervous system dysfunction and tomographic evidence certainly suggest brain
involvement. Measurements of other potentially defective enzymes such as PDH
showed normal levels of activity. Tissues examined had fat droplets, probably rep-
resenting an inability to oxidize fatty acids. The authors state that ATP synthesis by
oxidative phosphorylation would be decreased and that depleted ATP would have
contributed to the demise of the patient. The authors state that autopsy results sug-
gest brain involvement, although scant neuropathological data are presented.
In another case of PDH deficiency, all three components of the enzyme complex
were measured and the defect localized to the first (thiamine requiring) component
(Ho, L. et al., 1986). In this study an immunoblot technique was used to assess
components of the PDH complex. E1 (the first component) had an activity level of
only 33% of control values.
In the over 80 cases described at this point, as many as 50 have been attributed
to defective PDH components. The previous measurements of the E1 complex were
difficult due to its contributing only about 10% to the overall total PDH activity. The
immunoblot method described should simplify and increase the number of defective
components of the mutation of PDH.
Emphasizing the complexity and continuing evolving discovery of additional
mutations involving inherited ataxia, elevated lactate/pyruvate ratios, and in these
Fig. 3 Brain slice at the level of the superior cerebellar peduncle from a Friedreich’s ataxia patient.
Figure courtesy of Dr. Roland Auer
cases, seizures may occur (Mollet, J. et al., 2008). Symptoms in these cases were
similar. Patients seemed to progress well for a period, but in one case became ataxic
by the age of 2. At the age of 2.5, seizures (generalized tonic) began and were
accompanied with increased CSF lactate. Sodium valproate and rivotril controlled
seizures for 5 years. His neurological condition deteriorated by 12 years of age,
such that he could not walk or speak. Seizures were increasing, and MRI revealed
a severe cerebellar atrophy along with cerebellar ataxia and ptosis. A brain MRI
showed cerebellar vermis hypoplasia. Some years later, he showed severe cerebel-
lar atrophy. Yet another patient in this study is a sister of case 2 and had a similar
course. At the age of 13, she had cerebellar ataxia.
Numerous molecular studies described in this manuscript showed a defect in
the CABC1 gene which is involved in the ubiquinone biosynthesis pathway. The
precise function of the ABC1 protein is unclear. The patients in this study were all
ubiquinone deficient. The deficiency was noted in skeletal muscle; no “ragged red”
fibers were noted.
A key clinical observation in those patients was the development of seizures, to
the point of “epilepsia partialis continua.” The authors looked for the CABC1 muta-
tion in patients with similar symptoms (such as Leigh’s patients) and the CABC1
mutation was not found. This indicates a new variation of cerebellar ataxia attributed
to the CABC1 gene, and not previously described. Although rather similar in clin-
Fig. 4 Photomicrograph of the olive in a Friedreich’s ataxia patient. Figure courtesy of Dr. Roland
Auer
ical presentation, the patients had varying degrees of respiratory chain dysfunction
and ubiquinone deficiency. For example, two patients had severely decreased muscle
ubiquinone content, whereas another patient had ubiquinone levels at lower control
values. The authors state that muscle values may not reflect metabolism in the CNS.
The exact function of the CABC1 protein is unclear, especially in brain. Ubiquinone
treatment trials had little beneficial effect.
The reason for variations in symptoms and in tissue responses to mutation of the
CABC1 gene is unclear. Probably different organs would have different ubiquinone
requirements and have different normal ability to metabolize ubiquinone. These
tissue differences would of course be reflected in the deficient state. Additional
hypotheses to explain different tissue responses involve expression levels of var-
ious proteins concerned with ubiquinone synthesis and metabolism. None of the
patients with CABC1, SCo2, or p53R2 mutations showed any evidence of apopto-
sis or tumor susceptibility. CABC1 is a target of the p53 tumor suppressor gene,
making it the third p53 gene in disorders of mitochondria (Iiizumi, M. et al., 2002).
To summarize, this manuscript reports the first description of severe neurolog-
ical symptoms, and cerebellar atrophy/ataxia attributed to a point mutation. Treat-
ment regimens were unsuccessful. Severe seizures to the point of epilepsia partialis
continua were a unique feature of seizures in these cases. The seizures seemed
intractable to drug therapy, but other seizure treatment regimes were apparently not
tried. Again, these cases point out the extremely complex features of these inher-
Fig. 5 Photomicrograph showing “dying back of neurons” in a Friedreich’s ataxia case. Figure
courtesy of Dr. Roland Auer
ited ataxias and the problems of both diagnosis and possible treatment modalities.
Correct diagnosis alone and delineation of the disorder biochemical features outside
of a major medical center seem daunting.
A recent review paper (Sedel, F. et al., 2007) has been published looking at inborn
errors of metabolism. Examined were a large number of inborn errors of metabolism
(IEMs) with emphasis on those amenable to treatment. These were characterized
and assigned to various groups including inherited ataxias. It is stated that nearly
all IEMs affect many organs such as liver, heart, kidney, skeletal muscle, but nearly
all also affect brain. IEMs are candidates for various treatment modalities, includ-
ing enzyme and gene replacement therapy. The review focuses on disorders that
are mild–moderate in symptom presentation and have late childhood or adolescent
onset.
Inborn errors of metabolism which involve chronic ataxias include the PDH defi-
ciency syndromes. In these, MRI of the CNS usually shows bilateral symmetric
lesions in some but not all ways similar to lesions seen in Leigh’s disease. These
disorders characteristically have cerebellar ataxia, ocular disturbances, deafness,
lactic acidosis, and elevated lactate/pyruvate ratios. Successful treatment of these
specific IEMs have been elusive. Often, acute episodes of cerebellar ataxia can be
triggered by stress or fever in PDH deficiency patients. Chronic cerebellar ataxia
also accompanies Friedreich’s ataxia and coenzyme Q10 deficiency. Seizures may
Fig. 6 Slice at the level of the thoracic spinal cord from a case of Friedreich’s ataxia. Figure
courtesy of Dr. Roland Auer
occasionally accompany inherited ataxias and may respond to typical anticonvul-

sant therapy. Seizures are characteristic in glucose transporter (GLUT-1) deficiency.
In the case of GLUT-1 deficiency seizures, treatment with a ketogenic diet may
be effective. These patients may also show loss of muscle tone and impaired con-
sciousness. There is usually a low CSF/blood glucose ratio, which can be useful in
diagnosing this IEM.
In a review paper (Berendzen, K. et al., 2006) the use of oral dichloroacetate in
patients with congenital lactic acidosis was examined. Much of the lactic acidosis
is a result of improved pyruvate dehydrogenase activity, and about 90% of cases
are due to defects in the thiamine requiring first component of the three-component
haloenzyme. At least 80 different mutations have been reported in patients with
lactic acidosis. Inhibition of this key enzymatic step is thought to decrease energy
production in brain.
Dichloroacetate increases pyruvate dehydrogenase activity by either inhibiting
pyruvate dehydrogenase kinase and/or by stabilizing the dehydrogenase enzyme or
slowing its rate of breakdown. Looking retrospectively at patients’ records it was
shown that those receiving a daily dose range of 20–135 mg/kg of dichloroacetate
was associated with a decrease in blood and CSF lactate by about 40%. Four patients
Fig. 7 Slice at the level of the lumbar spinal cord from a case of Friedreich’s ataxia. Figure cour-
tesy of Dr. Roland Auer
with pyruvate dehydrogenase complex deficiency have been treated for up to 5 years
and are still being evaluated.
The authors argue that this data supports the concept that long-term (5 years)
treatment with dichloroacetate is safe. The exact benefit for these disorders is
unclear, but controlled clinical studies seem warranted based on the above data on
dichloroacetate.
In a paper (Debray, F. et al., 2008), two siblings with ataxia were studied and
found to have an unreported mutation (G585C) in the pyruvate dehydrogenase A1
gene. These patients clinically exhibited intermittent ataxia. The clinical hetero-
geneity of pyruvate dehydrogenase deficiency is emphasized, and these two patients
may represent the first two cases in which CSF lactate levels were not elevated.
This disorder seems to have a broad spectrum of clinical progression. In the mildest
form there may be occasional episodes of ataxia, with complete resolution between
episodes. In the worst case, the patient has a course as seen in Leigh’s disease,
relentless and resulting in death.
This paper describes a heretofore unreported defect in the pyruvate dehydro-
genase A1 gene where the mutation c6585c leads to the replacement of glycine to
alanine. This change is in a thiamine pyrophosphate binding domain (TPP-BD). Thi-
amine treatment had no improving effect on the two patients. The authors emphasize
the large phenotypic variation in symptom presentation in intermittent ataxias and
suggest that the intermittent nature of these disorders could lead to an underestima-
tion of incidence.
In an interesting recent case description (Henwood, M. et al., 2006), a patient
previously diagnosed and well documented with pyruvate dehydrogenase deficiency
presented with severe diabetic ketoacidosis. She was being treated with a ketogenic
diet aimed at circumventing in part the impaired dehydrogenation of pyruvate with
resultant energy metabolism depletion. The dilemma was to find treatments for the
pyruvate dehydrogenase deficiency and diabetes.
The authors implemented a ketogenic diet and limited insulin therapy which
achieved a safe balance. It was shown that adequate ketosis and insulin treatment
could coexist such that each metabolic disturbance could be satisfactorily treated.
In fact, her progress under this new treatment regime has improved overall including
her mental status. In addition, her growth has improved, as well as developmental
milestones, such as sitting up, standing, and drinking from a cup.
In a recent study (Kang, H. et al., 2007) the characteristics of seizures were
examined in patients with mitochondrial diseases. This was a retrospective study
in which 142 patients were examined in terms of seizure phenotype and mito-
chondrial disorder. Of these patients, 45 were identified as having mitochondrial
disease, and 22 of these were epileptic. Of the 22, various clinical manifestations
were present including optic neuropathy, hearing loss, cardiomyopathy, develop-
mental delay, etc. Eight patients in whom the test was done showed significant
decrease in complex 1 activity. The epileptic phenotypes were diverse. Some had
severe myoclonic seizures, some generalized seizures, some had absence seizures,
etc. Seizures, however, were the first unequivocal symptom in 16 patients. All of
these patients had EEGs characteristic for the phenotypic type seizure. These find-
ings are not unusual in that it is recognized that mitochondrial diseases can lead to
brain damage conducive to seizures (Kunz, W., 2002). In some cases the diagno-
sis of mitochondrial disorders precedes seizures, and of course not all patients with
mitochondrial disorders ever have seizures.
Most of the patients described in this manuscript did not respond well to
antiepileptic drug treatment. The treatment of mitochondrial disorder patients with
ketogenic diets has been successful in some patients, but not all. The authors
point out that the retrospective data cannot provide definitive numbers concern-
ing the number of epilepsy cases in mitochondrial disorders. Data does suggest
that mitochondrial disorders were more common in epileptics than anticipated.
While the phenotype expression of these epileptic patients is diverse, the clinician
should always be alert to the possible coexistence of both disorders in the same
patient.
In a retrospective study, 73 patients diagnosed with mitochondrial diseases were
studied (Debray, F. et al., 2007). Results showed that the patients could be arranged
in seven phenotypes: (1) neonatal onset lactic acidosis (10%), (2) Leigh’s syndrome
(18%), (3) intermittent neurologic (5%), (4) Leber hereditary optic neuropathy
(5%), (5) mitochondrial myopathy (19%), (6) visceral (11%), and (7) nonspecific
encephalopathy (32%).
Initial clinical presentation occurred at a mean age of 7 months and included

metabolic acidosis, failure to thrive, seizures, ataxia, muscle weakness, psychomo-
tor delay, muscle weakness, etc. In the neonatal group, all presented shortly after
birth with lactic acidosis and neurological distress. All died in the neonatal group
with a median age at death of 4 days.
Biochemical studies showed that 72% of the patients had chronic lactic acidosis.
Blood alanine was elevated in a similar number of patients. Krebs cycle intermedi-
ates were elevated and ketone bodies were found in the urine of 70% of patients in
which these were measured.
MR images were available in 59 patients, and most frequently seen were basal
ganglia hyperintensities and cerebral atrophy. Other significant findings included
those localized to the cerebellum, brainstem, white matter, and corpus callosum.
These were all hyperintensities. Stroke like infarcts were occasionally seen. Mus-
cle samples examined with electron microscopy showed mitochondrial proliferation
or ultrastructural morphologic alterations in 80% of 40 patients’ samples. Staining
techniques showed mitochondrial proliferation in about 50% of samples examined
(Figs. 8, 9, 10, and 11).
Biochemically, an enzyme defect was identified in 44 patients, and complex 4
deficiency was the most frequent. Pyruvate dehydrogenase and complex 4 were the
most common defects, about 25% each. The overall mortality rate was 46% with a
median survival rate of 12 years. There was no association of mortality rate and the
Fig. 8 MRI of a Leigh’s

patient showing high signal
intensity of periventricular
white matter. Reproduced
from Schiff et al. (2006) with
permission from
Wiley-Blackwell

intensity from swollen
caudate nuclei and putamina
and thalamus at acute
deterioration. Reproduced
permission from
Wiley-Blackwell
Fig. 10 MRI of Leigh’s

intensity of cortex and
subcortex at a time of acute
deterioration. Reproduced
permission from
Wiley-Blackwell

patient 1 year after acute
deterioration showing
bilateral hypersignal and
atrophy of the basal ganglia,
and enlarged ventricles with
cerebral atrophy. Reproduced
permission from
Wiley-Blackwell
biochemical defect. Signs before 6 months of age of liver involvement, metabolic

crisis, etc., were indicators of high mortality rate.
The authors point out a high frequency in these patients of a complication of
some sort in the pregnancies. A low birth weight was also noted. Cardiomyopathies
were relatively low in these patients. Measurement of blood lactate proved to be a
non-specific test. Since most of these disorders are untreatable, the authors suggest
initially using the least invasive techniques, which can have a good yield in terms of
diagnosis. The clinical course of these disorders can be long term, so clinical follow-
up is essential. The authors state that many patients have only mild or no sequelae
after 5 years of follow-up. This paper again points out the “overlap” between Leigh’s
disease and the so-called inherited ataxias.
In a recent case (Sedel, F. et al., 2008), an adult with a deficiency of pyruvate
dehydrogenase was successfully treated with thiamine. The patient was a 26-year-
old man who from about the age of 2 had paroxysmal walking problems associated
with fever. He had optic neuropathy by the age of 20, and measurement of blood
lactate showed a mild increase. Measurement of lymphocytic pyruvate dehydroge-
nase activity showed a level of only 13% of control. Sequencing the PDHA1 gene
showed a mutation in the third exon (E75A). This is a mutation not recorded before.
No other mutation in the PDHA1 gene was found.
Treatment with thiamine at a dose of 1 gm/day was initiated. Eighteen months

later, visual acuity was improved, and no additional episodes of muscle weakness
or ataxia had occurred. Lactate levels were within normal limits. The effects of
treatment on ataxia and muscle weakness are thought by the authors to be a direct
effect of thiamine treatment. Adding thiamine to the patient’s fibroblasts in cul-
ture increased pyruvate dehydrogenase activity nearly fivefold. The observations are
significant because pyruvate dehydrogenase activity has only rarely been described
in adults, and especially because the deficiency has only rarely been improved with
thiamine treatment. It should be noted that the dose of 1 gm/day is a very high dose.
These positive findings are unique in that this mutation may be equally unique, rep-
resenting the first such description.
Thiamine Deficiency in Serious Illness
This chapter is entitled thiamine deficiency in serious illness and will examine
relatively recent descriptions of the development of thiamine deficiency in various
cancers, gastric bypass surgery, etc. Predating these discoveries have been thorough
descriptions in prisoners of the development of thiamine deficiency. These cases are
different from those in various cancers in that they are clearly based on dietary inad-
equacies. They are similar however in that these men were seriously ill and have no
other counterpart in clinical medicine. Therefore a comprehensive study and review
of this prisoner condition is synopsized here (Denny-Brown, D., 1947).
Beriberi
Dr. Denny-Brown served as a consultant in neurology in Southeast Asia immedi-

ately following WW2 and was in a position to examine thousands of released pris-
oners of war. There were various “releases” of prisoners and in one such release
there were 650 prisoners, and only 235 of them were sick. Of these there were a
surprisingly low number of overt vitamin deficiency cases. There were several cases
of beriberi, as well as cases of ataxia, edema, and polyneuropathy. The vast majority
had malaria, dysentery, and helminth infestations. The number of cases dying before
release from any of the above conditions, including beriberi, is unclear.
In another prisoner release, beriberi was recognized in a number of patients.
Peripheral neuritis, edema, and ataxia were all noted; however, the number of these
cases was lower than was expected. Other nutritional deficiencies were diagnosed
including pellagra. The findings of other neurological deficits such as retrobulbar
neuritis, spastic paraplegia, and acrodynia (burning feet) could have been indepen-
dent of thiamine deficiency. These conditions have been described as a component
of beriberi, but the author speculated that they occurred independently.
Most released patients were suffering from a variety of illnesses, some severe,
some not severe. In addition, thousands of prisoners died before release, and these
may represent many severe cases of nutritional deficiency coupled with other seri-
ous illnesses. Nevertheless, these released prisoners had nutritional deficiencies
(thiamine) coupled with diseases such as malaria. The interrelationship between
these two disorders remains unclear.


Although thiamine deficiency (beriberi) has been medically described since

1642, thiamine deficiency in serious illness (TDSI) is a relatively new clinical find-
ing. In addition to thiamine deficiency in prisoners, it has been described in patients
with breast cancer, hematologic cancers, gastrectomy patients, HIV/AIDS, and oth-
ers. In some instances the thiamine deficiency is mild and measured largely through
laboratory tests. These cases might have no overt symptoms, but the possible con-
tribution of the deficiency to morbidity and mortality is unknown. In other cases,
the thiamine deficiency can be overt and severe on its own and produce deleterious
effects such as cardiac failure.
Cancer
Another type of thiamine deficiency associated with serious illnesses are cancers
treated with 5-fluorouracil. This anticancer agent appears to be an antagonist to
thiamine derivatives, and thereby produces thiamine deficiency in some patients
being treated with the drug.
One of the first such studies (Basu, T. et al., 1974) of TDSI was of transketo-
lase activity in cancer patients undergoing 5-fluorouracil treatment. Transketolase,
a thiamine-dependent enzyme was chosen as an indicator of thiamine deficiency.
Other easily obtained test data such as blood pyruvate levels and urinary excretion
of thiamine are non-specific. The measurements of blood transketolase activity and
the thiamine pyrophosphate (TPP) effect (the increase in transketolase activity when
TPP is added to the assay) are, when taken together, excellent measures of relative
thiamine deficiency. These two assays were used in this study to measure thiamine
deficiency.
Results showed that in cancer patients (breast, respiratory system, bladder,
cervix, bowel, etc.) transketolase activity was not statistically significantly changed
as compared to controls. When adding TPP to the reaction, however, the values
of activity for transketolase were significantly elevated as compared to controls.
The elevation was about 65%. To further examine the possible cause of the effect
on transketolase, 5-fluorouracil was added to red blood cell hemolysates. Results
showed a graded effect in that the transketolase activity was decreased in proportion
to added 5-fluorouracil.
Even though the cancer patients were receiving multivitamins, many had signif-
icant values for the TPP effect, indicating a significant level of thiamine deficiency.
Clinical evidence of thiamine deficiency was absent. However, these data suggest
that cancer patients are susceptible to thiamine deficiency. The overall effect of min-
imal thiamine deficiency in cancer patients is unclear, but bears further investigation.
The effect of 5-fluorouricil is consistent with the concept that it is an antagonist
to TPP. The exact mechanism by which 5-fluorouricil acts on thiamine is not clear.
Further studies examining how 5-fluorouricil works on the transketolase reaction
and on the effects of other antitumor agents are warranted.
In another paper by Basu (Basu, T. and Dickerson, J., 1976), the thiamine status
is examined in patients with breast and bronchial cancers. A total of 42 patients with
Thiamine Deficiency in Serious Illness 133
early breast or bronchial cancer who had not yet received chemotherapy or surgery
were studied. In these patients, the TPP effect was measured in cancer patients
and non-cancer controls. Results showed that the TPP effect was increased over
twofold in bronchial cancer and threefold in breast cancer patients. When examined
for thiamine, results showed an increase of the vitamin in urine of the bronchial
and breast cancer patients. There was no evidence for any nutritional deficiencies
in any of the 42 patients examined, nor evidence of malnourishment due to loss of
appetite.
The increase in urinary thiamine could be due to a defect in absorption of thi-
amine in the gut, or it could also be due to a failure to convert thiamine to TPP.
The authors state that available data suggest that in these two cancers, there is a
decrease in the conversion of thiamine to the coenzyme form, TPP. Further study
should determine the exact cause of thiamine deficiency in these illnesses.
Intensive Care Units
In a paper (Cruickshank, A. et al., 1988), a retrospective study was done to examine

possible thiamine deficiency in critically sick patients. In this study, 152 patients
were divided into two groups: those who died in the ICU within 1 week and those
who did not die. There were an equal number of patients in each group. Transketo-
lase activity and the TPP effect were measured in each case.
Results showed considerable variation in transketolase activities within the two
groups. The activation of transketolase in RBCs in patients who died was signifi-
cantly increased. This value is derived from the initial transketolase value, and that
measured after TPP is added to the reaction. An increased TPP effect is taken as evi-
dence of thiamine deficiency. A few of the critically ill patients in the ICU continued
to have poor transketolase activation values, and died despite thiamine supplemen-
tation of 1.2 mg/day.
This study shows a correlation between mild thiamine deficiency and failure to
survive in the ICU. The exact cause for this is unclear. Severity of illness seemed
to be about equal between groups. The overall nutritional status of these patients
could not be fully assessed, but those in the group who failed to survive may have
had a lowered status. Of the group who were thiamine deficient and died, seven had
hemodynamic deterioration before death and one had neurological complications.
None of these showed any pathological evidence of either beriberi or Wernicke’s
disease. Beriberi can present clinically as either high- or low-output diseases, mak-
ing diagnosis in otherwise ill patients difficult. The presence of thiamine deficiency
in otherwise sick patients may influence the patients’ outcome, but exactly how and
to what degree is uncertain.
The approximate daily requirement for thiamine is about 1 mg/day, but this
is increased in conditions in which metabolic rate might be increased such as in
some cancers and other serious illnesses. In parenterally fed patients, the suggested
requirement could be as much as 3 mg/day (American Medical Association, 1975).
These levels of thiamine supplementation could probably be increased somewhat.
One way to assess this is by performing frequent transketolase assays, but this is
not practical. Evidence for clinical improvement using thiamine supplementation is
inconclusive, but such additional thiamine is certainly not detrimental, thereby jus-
tifying its use in critically ill patients. Further investigation is needed to evaluate
thiamine deficiency in otherwise very sick patients.
Hematologic Tumors
In another study (van Zaanen, H. and van der Lelie, B., 1992), thiamine deficiency
was diagnosed in patients with fast growing hematologic tumors. Six patients were
examined for thiamine deficiency, and among these, two were found. Both patients
had severe cardiovascular symptoms. These symptoms were treated with thiamine
supplementation, and the symptoms quickly reversed. In neither of the patients was
there any evidence of neurological symptoms commonly seen in Wernicke’s disease.
Neither of these two patients were abusing alcohol.
The authors speculate that because leukocytes have levels of thiamine-dependent
enzymes (pyruvate dehydrogenase, alpha-ketoglutarate dehydrogenase, and transke-
tolase), the hematologic tumors have even higher consumption of thiamine, thereby
leading to thiamine deficiency. The authors state that this finding in cases of hema-
tologic rapidly growing tumors suggests that clinicians be vigilant for other such
cases. The presence of thiamine deficiency in these types of cases should prompt
treatment with increased thiamine supplementation. The possible beneficial effects
from early recognition of the deficiency and early treatment are stressed.
Gastrectomy
Yet another class of patients have been described with thiamine deficiency
(Koike, H. et al., 2001). In this study, 17 patients were examined who developed
polyneuropathy following gastrectomy. Patients included in this study had had a
total or partial gastrectomy due to ulcers or tumors, and were thiamine deficient
and developed a polyneuropathy. The polyneuropathy was confirmed by sural nerve
biopsy. Thiamine levels were measured in blood, and transketolase activity was
measured in RBCs. Patients with other neurological disorders were excluded from
the study, including alcoholics. Sensory and motor conduction studies were per-
formed, and a psychiatric evaluation was done.
Results showed that the first symptoms were weakness in arms and legs in 11
patients. All 17 eventually developed extremity polyneuropathy with the lower
extremities being more affected. With time the trunk became involved in five
patients. Inability to walk was common to almost all patients. Burning and numb-
ness was present in all 17 patients. Cranial nerves were affected, as evidenced by
facial palsies and gaze palsies.
The Wernicke-Korsakoff syndrome could be diagnosed in four patients, and
heart failure was diagnosed in ten patients. Edema was present in the legs of most
patients. All patients had blood thiamine levels below that of controls. Studies of
nerve conduction showed significantly decreased amplitude of compound muscle
action potentials as well as sensory nerve action potentials. The density of myeli-
nated fibers in the sural nerves was significantly reduced. All myelinated fibers were
affected. Unmyelinated fibers were decreased as well.
Treatment of these patients with 100 mg of thiamine daily had a marked effect.
Heart symptoms such as cardiomegaly, arrhythmias, and edema were improved in a
few days. Improvement in muscle weakness could be noted in a few days, and the
patients’ ability to walk returned after 3–6 months of thiamine administration. Overt
Wernicke’s disease improved with thiamine treatment (not unexpectedly), whereas
symptoms of Korasakoff’s psychosis remained in one patient for at least 2 years.
Both thiamine levels and transketolase activity levels were decreased in all
patients in this study. Thiamine levels proved to be more reliable than transketolase
activity levels. Blood levels of most other vitamins were normal or only slightly
decreased, but no symptoms of other vitamin deficiencies were noted. The clini-
cal symptoms of these patients were comparable to those seen in beriberi. These
included polyneuropathy, cardiac manifestations, and edema. The rapid improve-
ment in symptoms following thiamine treatment further emphasizes the thiamine
deficiency characteristics of this disorder.
Gastrectomy was a common procedure in all patients in this study. A few reported
cases of thiamine deficiency in gastric resection patients had been reported earlier
(Abarbanel, J. et al., 1987). None of the present cases were operated on for obesity,
only for ulcers or tumors. Vomiting, common in surgery for obesity, was not a fea-
ture of these cases. Previous reports of Wernicke’s disease have appeared in cases
of gastrectomy (Shimomura, I. et al., 1998). Since thiamine is largely absorbed in
the intestine, which is not included in a gastrectomy, the exact mechanism of how
thiamine deficiency develops is unclear.
The authors further state that thiamine deficiency symptoms did not appear
immediately after surgery, but some did have an acute onset. Because of a possi-
ble decrease in eating prior to surgery due to ulcers or tumors, a mild subclinical
thiamine deficiency occurred. After surgery the patients might have an increased
demand for thiamine as is seen in cases where thiamine deficiency can be precip-
itated by fever, increased muscular tone and activity, pregnancy, or other causes.
Also, it is possible that treatment of tumors with agents such as 5-fluorouracil could
also produce thiamine deficiency by inhibiting thiamine activity as described earlier
in this chapter.
In another study (Seligmann, H. et al., 2001), the thiamine status was assessed
in 14 patients with B-chronic lymphocytic leukemia (CLL). This leukemia was
diagnosed based on the National Cancer Institute guidelines. Excluded from this
study were alcoholics, malnourished patients, malabsorption syndromes, patients
with vomiting, prolonged fever, liver disease, etc. In the 14 patients with CLL, and
14 controls, thiamine status was assessed by erythrocyte transketolase activity and
by the TPP effect.
Results showed thiamine deficiency, as measured by transketolase and the TPP
effect, in 5 of the 14 CLL patients. None of these five showed any overt clinical signs
or symptoms. Since the study had a rigorous exclusion criteria, it would seem that
CLL played a role in the thiamine deficiency. This could be due to an increased thi-
amine requirement by tumor cells, increased whole body requirements, or decreased
TPP. In keeping with these possible mechanisms was the finding that the thiamine
deficiency was present in patients with advanced disease. While the five patients
with thiamine deficiency were asymptomatic, the presence of a “mild” thiamine
deficiency in the CLL disease process is yet to be determined.
It is noted by the authors that treatment of this subclinical thiamine deficiency
in cancer patients should be approached carefully. For example, ribose synthe-
sis coming from the pentose phosphate pathway (a thiamine-dependent pathway)
could be increased by thiamine treatment. Increase in ribose for nucleic acid syn-
thesis could possibly be a promoter of tumor cell growth in various cancers. Thus,
excess thiamine therapy might promote tumor growth, and overall aid the disease
process.
Previous studies had shown that cancer treatment was associated with thiamine
deficiency, and the exact cause was speculative. Additional studies examined the
effect of various antitumor drugs on thiamine metabolism. In one such study, the
neurological signs and symptoms of 5-fluorouracil were described (Moore, D. et al.,
1990b).
5-Fluorouracil
5-Fluorourcil is an antimetabolite used widely as an antitumor agent against a num-

ber of solid tumors. It is known that 5-fluorouracil can produce neurological symp-
toms, although this is not common. The paper presents two cases of 5-fluorouracil
neurotoxicity. Both patients had genital tract malignancies, which were treated with
5-fluorouracil.
Within a few days of the onset of 5-fluorouracil treatment, signs and symp-
toms such as weakness, slurred speech dysarthria, nystagmous, and ataxia were
noted. In one case, computed tomography was normal. Following the diagnosis of
5-fluorouracil, the drug was discontinued and in one case, symptoms were greatly
improved in a few days. In the other case following diagnosis, thiamine therapy was
instituted, and symptoms were completely reversed in 3 days.
A literature review by these authors found the neurological symptoms of ataxia,
slurred speech, and cerebellar ataxia in patients receiving 5-fluorouracil to be only
2–5%. The symptoms of 5-fluorouracil neurotoxicity are fully reversible with ces-
sation of 5-fluorouracil therapy and/or treatment with thiamine. This 5-fluorouracil
disorder has an acute onset, and is present in the absence of other possible
contributors such as brain lesions, diabetes, or fever.
The mechanism by which 5-fluorouracil produces neurological symptoms is
unclear. One metabolite produced when 5-fluorouracil breaks down is fluorocitrate,
which is in turn a TCA cycle inhibitor. It has also been shown that 5-fluorouracil
can block the conversion of thiamine to TPP, the active form of thiamine. Since
nutritional deficiencies may be borderline in cancer patients, the added burden
of 5-fluorouracil may serve to push vulnerable patients into frank thiamine defi-
ciency. Further 5-fluorouracil toxicity studies should shed light on the mechanisms
of neurological deterioration associated with its use.
In another study, the diuretic drug furosemide used to treat congestive heart fail-
ure was examined for its ability to produce thiamine deficiency (Seligmann, H. et al.,
1991). Patients were divided into two groups: group A with 23 patients with chronic
heart failure receiving furosemide and group B with 16 control patients without con-
gestive heart failure and not receiving any diuretics. All patients received the same
diet, and thiamine status was evaluated by measuring urinary thiamine output, and
measuring the blood transketolase activity, and the TPP effect.
Results showed an increased TPP effect in 21 of the group A patients and only
2 of the group B patients. Urinary excretion of thiamine was greater in the group
A patients as compared to the control group. Transketolase activity measurements
were similar between group A and group B.
Thiamine deficiency can cause cardiovascular and neurological signs and symp-
toms. These signs and symptoms are rapidly reversed by thiamine administration.
Mild thiamine deficiency can exist unnoticed for a long time and be measured by
such tests as increased thiamine in the urine, high blood lactate and pyruvate levels,
and decreased transketolase activity and increased TPP effect. It has been found that
the diuretic furosemide produces an increase in urine thiamine and an increased TPP
effect, both indicators of thiamine deficiency.
The mechanism of furosemide toxicity is not clear. The diuretic does
cause anorexia, hyponatremia, and hypomagnesemia. These three alterations in
metabolism could change normal thiamine metabolism. The initiation of thiamine
therapy resulted in the rapid return to normal of the increased TPP effect. Thiamine
treatment also served to improve cardiac contractility. Thus, thiamine deficiency
should be diagnosed and treated since a failing heart can be adversely affected by
thiamine deficiency, as seen in beriberi heart disease.
The drug omeprazole (prilosec) has widespread use as a drug for treating acid-
induced inflammation and ulcers in the stomach and duodenum. Data show that
omeprazole is detrimental to the normal function of thiamine metabolism (Nixon, P.
et al., 1992). In this study, the effect of omeprazole was evaluated in vitro on the
activity of the two thiamine-dependent enzymes transketolase and pyruvate decar-
boxylase.
Transketolase was prepared from human red blood cells. Possible omeprazole
inhibition was evaluated by adding the drug to media containing apotransketo-
lase and measuring activity spectrophotometrically. Pyruvate decarboxylase was
prepared from the bacteria Z. mobilis and as above, measuring activity spectropho-
tometrically.
Results showed that both thiamine-dependent enzymes transketolase and pyru-
vate decarboxylase were inhibited by omeprazole. The results indicate that omepra-
zole is an analog of thiamine, and inhibits two thiamine (TPP) requiring enzymes,
transketolase and pyruvate decarboxylase. This drug therefore may have a detri-
mental affect on patients who may at ready have threatened thiamine reserve due to
anorexia.
While thiamine deficiency can occur in cancer patients, and can be deleterious,
in another view, the over administration of therapeutic thiamine may be conducive
to tumor growth (Boros, L. et al., 1998). It is well known that the pentose phosphate
pathway is a key source of ribose production, and therefore RNA and DNA synthesis
depend on a well-functioning pentose phosphate pathway. The rate-limiting step in
this pathway is the enzyme transketolase, a TPP-requiring enzyme.
There are clearly several mechanisms whereby cancer patients may develop
thiamine deficiency (see above). This thiamine deficiency may be severe in that
Wernicke’s, beriberi, lactic acidosis, etc., may develop, further complicating the
cancer treatment. It has been shown that ribose production in pancreatic adenocar-
cinoma cells proceeds through the pentose phosphate pathway and that the adminis-
tration of the antithiamine metabolite oxythiamine inhibits tumor cell proliferation
significantly (Boros, L. et al., 1997). The tumor cell inhibition is demonstrable both
in vitro and in vivo.
Since the treatment of even mild thiamine deficiency in cancer patients is usually
achieved by an administration of several times the minimal daily requirement for
thiamine, this treatment regime might be conducive to tumor growth. This excess
treatment of cancer patients with thiamine could increase ribose production, thereby
rendering treatment of cancer patients with anticancer drugs less affective.
The finding of thiamine deficiency in patients with gastrectomy has been
described (Sekiyama, S. et al., 2005). This paper describes a patient who had a total
gastrectomy for a gastric signet ring cell carcinoma. The patient felt that the can-
cer might have been related to diet, so stopped eating meat, chicken, fish and milk,
and ate only rice, noodles, cabbage, onions, radish, etc. A couple of years later, the
patient had lower extremity edema, plus tingling and pain in the feet.
A neurological exam showed the patient had edema, lower extremity weakness,
and absence of deep tendon reflexes. Lower extremity hyperesthesia and paresthe-
sias were present. A diagnosis of beriberi was confirmed based on serum concentra-
tions of vitamin B1 and B12, which were decreased from normal. Thiamine therapy
was initiated (75 mg/day), and a proper diet was started. The patient’s signs and
symptoms improved, and she made a full recovery.
This patient had a gastrectomy which acted to decrease the length of duodenum
available for thiamine absorption. In addition, the diet she adopted following surgery
was inadequate in many ways, including thiamine content. Other cases of thiamine
deficiency have been associated with gastric resection, and this case emphasizes
other problems which can coexist, such as poor diet and alcoholism. The poor
diet is frequently seen in cases of gastrectomy because patients erroneously think
that a protein-rich diet may have been responsible for their disease. This patient
also had a vitamin B12 deficiency, which may also have contributed to her clinical
picture.
Emphasizing the risks for certain types of gastric bypass surgery such as
the “Roux-en-Y” type, a recent study examines neurological complications
(Juhasz-Pocsine, K. et al., 2007). In this study, nearly one half of patients showed
ataxia/walking difficulty and loss of sensation in the lower extremities. Some
patients may have had brain involvement. In other patients, the neurological com-
plications appeared to be limited to the peripheral nerves. These neurological signs
often develop from 1 to 2 months after surgery and when diagnosed can be effec-
tively treated with thiamine in a proper dose.
Wernicke’s Disease and Gastrectomy
Wernicke’s encephalopathy can develop many years after gastrectomy

(Karapanayiotides, T. et al., 2006). This brief article describes a patient who
developed unsteady gait, vomiting, and dizziness. Three years earlier he had
experienced polyneuropathy. He had been drinking 1–2 drinks per day, but was
not an alcoholic. Twenty-eight years earlier, he had a partial gastrectomy due to
a peptic ulcer. He developed nystagmus, abducens palsy, confusion, stupor, and
continued to vomit following surgery.
MRI showed lesions in the floor of the fourth ventricle, periaqueductal gray, and
hypo and medial thalami. These images were suggestive of Wernicke’s disease, and
thiamine treatment was initiated. Three hours after thiamine therapy was begun,
symptoms (abducens palsy) began to abate. One month later, there was minimal
nystagmus and amnesia. The authors state that the poor diet, alcohol consumption,
and vomiting precipitated this case of Wernicke’s disease many years after the gas-
trectomy. Thiamine therapy resolved most signs and symptoms.
In a brief case presentation (Boniol, S. et al., 2007), a patient was described
with malaise, large weight loss (50 lbs), and diffuse abdominal pain. He was jaun-
diced, had lymphadenopathy, and vomiting. He was ultimately diagnosed with non-
Hodgkin’s lymphoma. Nineteen days after initial chemotherapy, he complained of
difficulty in walking and blurred vision. Exam showed nystagmus, but MRI was
unremarkable. Thiamine therapy was initiated, and within 48 hours all neurologic
symptoms were gone. Ten weeks later he was still free from neurological symptoms.
This represents a case of Wernicke’s disease in a non-alcoholic patient who devel-
oped thiamine deficiency through decreased oral intake and increased cell turnover.
Since the diagnosis of Wernicke’s disease can be largely clinically based, this possi-
bility, although rare, must be considered since prompt thiamine therapy can quickly
reverse symptoms.
In the mid to late 1980s, several clinical papers were published describing thi-
amine deficiency like lesions (Wernicke’s disease) shown by autopsy findings in
patients dying of AIDS. The first descriptions seem to have been published by
Foresti, V. et al. (1986), and by Rosenberg (Rosenberg, S. et al., 1986). A diagnosis
of Wernicke’s disease was only occasionally suspected before death; most cases
were diagnosed by autopsy findings. Lesions were found in the usual sites for
Wernicke’s disease: the pons, thalamic nuclei, and mammillary bodies. Occasional
lesions were noted in other brain areas. The lesions were hemorrhagic in nature, and
microscopically, there was neuronyl cell loss and demyelination.
HIV/AIDS
In an attempt to determine if AIDS might be associated with thiamine deficiency, 39
patients were studied for the presence of thiamine deficiency (Butterworth, R. et al.,
1991). In this study, the AIDS patients were studied as regards their “TTP effect”
status. This test involves measuring the thiamine-requiring enzyme transketolase
both with and without added TPP. An increase of 15% or more of transketolase
activity after TPP addition is taken as indication of thiamine deficiency.
In an initial screening of the 39 AIDS patients, six were found to be thiamine
deficient, and in another screening 6–9 months later found three additional patients
with a TPP effect greater than 15%. The total number of thiamine deficient patients
was nine, nearly 25%. In none of these positive patients was there a history of alco-
hol abuse, and none had any overt Wernicke’s disease symptoms.
The association of thiamine deficiency and Wernicke-like lesions in AIDS
patients is most likely due to the extreme wasting and poor nutritional status of
AIDS patients. Indeed, thiamine deficiency had been documented in several can-
cers and in gastric bypass surgeries (see above, this chapter). In cases of thiamine
deficiency in cancers, as in AIDS patients, the degree of deficiency may be mild and
subclinical. However, the actual effect of thiamine deficiency on the course of AIDS
remains to be assessed.
Since diagnosis of thiamine deficiency in AIDS patients during life is difficult,
it behooves the health care worker to be cognizant of the possibility that thiamine
deficiency could be present (25% of cases), and if suspected, initiate treatment. The
paper of Butterworth et al. is important in that it shows beyond doubt that thiamine
deficiency is present in a significant number of AIDS patients, and actually calls
for thiamine treatment in all newly diagnosed AIDS cases. These data have been
replicated by other investigators, who found a very similar incidence for thiamine
deficiency as did Butterworth (Muri et al., 1999).
Other problems associated with thiamine deficiency in AIDS patients may
occur. One such problem is a severe lactic acidosis which has a low frequency of
occurrence in AIDS patients treated with nucleoside analogue reverse-transcriptase
inhibitors (NARTIs). The use of NARTIs in AIDS patients leads to a mild lactic
acidemia in about 20% of patients. However, in a much smaller number of AIDS
patients (1%), there is a complication of an abrupt severe lactic acidemia. Symptoms
from this may include malaise, vomiting, dyspnea, asthenia, etc. Lactate levels may
elevate to over 6 mmoles/L. The mortality rate for this development may approach
50% (Calza L. et al., 2005; Walker, V., 2004).
MRI studies of AIDS patients who develop Wernicke-like symptoms show
abnormal symmetric hyperintensities in the mammillary bodies, periaqueductal
gray matter, and the inferior colliculus, so characteristic of Wernicke’s disease
(Tattevin, P. et al., 2006).
Treatment is supportive, and an immediate cessation of NARTI treatment. Many
cases have in common the inclusion of stavudine in the treatment regimen. The
mode of action may involve the fact that these compounds are inhibitors of gamma
polymerase, an enzyme involved in the replication of mitochondrial DNA. Other
potential mitochondrial toxins such as valproate and acetylsalicylic acid must also
be withdrawn. Vitamin therapy can be started, and one case was successfully treated
with thiamine (Schramm, C. et al., 1999). This treatment of NARTI-induced severe
lactic acidosis with thiamine reinforces the concept that vitamin deficiency is a vital
feature of this relatively rare complication of NARTI therapy in AIDS patients.
The treatment of this patient with thiamine reversed the hyperlactic acidosis within
hours. Possibly, high doses of vitamins should be given to all patients showing this
life-threatening development of hyperlactic acidemia.
Malaria
Another area of serious illnesses in which thiamine deficiency has been shown to
have a possible detrimental effect is malaria (Krishna, S. et al., 1999; Mayxay, M.
et al., 2007). Malaria is, of course, a serious health problem in many parts of the
world such as sub-Saharan Africa, parts of South and Central America, and in the
Western Pacific, where there are as many as 2 million cases and 10,000 deaths per
year.
Perhaps the first to make a connection between malaria and thiamine deficiency
was Krishna et al. (1999). In the Lancet they described studies in which the thiamine
status was assessed in patients with malaria, and in controls. Results showed that
50% of patients with severe malaria and 20% with mild uncomplicated malaria were
also thiamine deficient as compared to control non-malaria patients.
Whether thiamine deficiency contributes to severe malaria or whether malaria
exacerbates thiamine deficiency is unclear. It is certain that the combination may
speed up the malaria condition, contributing to the morbidity and mortality. The
authors suggest that thiamine treatment should be combined with anti-malaria ther-
apy in order to increase treatment outcomes in a positive manor.
A recent paper (Mayxay, M. et al., 2007) examined the occurrence of malaria and
thiamine deficiency in the same patient. Severe falciparam malaria can be associated
with metabolic acidosis (Dondorp, A. et al., 2004). Malaria is one of the several
conditions which may increase glucose flux, thereby raising the need for thiamine
in its coenzyme form. Based on these observations, 310 patients were studied as to
their thiamine status. A follow-up assessment of the patients’ status was performed
42 days later, and the results compared. Thiamine status was evaluated by measuring
the activity of the thiamine-requiring enzyme transketolase.
Results showed that 12% of the patients initially tested showed a severe level of
thiamine deficiency based on the transketolase assay. Forty-two days later, only 3%
of patients were thiamine deficient. On hospital admission, patients were treated
with anti-malarial drugs (chloroquine, sulphadoxine-pyrimethamine, and methlo-
quine). Patients were also supplemented with a multivitamin tablet daily containing
1 mg of thiamine. The decrease from 12% at admission to 3% forty-two days later
in thiamine status suggests that the thiamine treatment did in fact decrease what was
a true thiamine deficiency.
Since younger red blood cells may have higher transketolase levels, and malaria
tends to destroy young red blood cells, the estimates of transketolase activity may be
underestimated. This study was done in Laos, where polished rice is consumed. Also
betel nuts are chewed, and betel nuts contain a thiaminase. So it is highly possible
that these people might have compromised thiamine metabolism. Nevertheless none
of the patients had any overt clinical evidence of beriberi, or any Wernicke-like
symptoms.
The fact that malaria elevates body temperature may also act to increase thiamine
deficiency. The authors call for further studies to examine the incidence and mor-
bidity/mortality of thiamine deficiency superimposed on malaria and other acute
infections. Otherwise, it seems simple enough to institute thiamine therapy in all
cases of malaria. Further education is essential regarding consumption of polished
rice, chewing of betel nuts, etc., in order to irradicate once and for all these out-
breaks of thiamine deficiency by themselves, and in association with other serious
illnesses.
Amyotrophic Lateral Sclerosis
Because of some similarities between inherited ataxias and amyotrophic lateral

sclerosos (ALS), thiamine and thiamine monophosphate levels were measured in
plasma and CSF in ALS patients, alcoholics, and controls. The ALS patients’
diagnosis was confirmed clinically, by electromyography, and by muscle biopsy
(Poloni, M. et al., 1982).
Results showed that thiamine and thiamine monophosphate were decreased in
plasma to the same extent, rendering the ratio of thiamine/thiamine monophosphate
unchanged. In CSF, the decreases were not equal, such that the thiamine/thiamine
monophosphate ratio was nearly doubled.
The authors speculate that this finding in CSF is unique in that other neurologi-
cal conditions show an equal decrease in thiamine and in thiamine monophosphate
yielding an unchanged ratio. A selective decrease in thiamine monophosphate raises
the ratio. The possibility of a decrease in numbers of neurons in ALS patients is dis-
cussed but dismissed. The authors suggest that the diminished thiamine monophos-
phate could be due to a decrease in thiamine pyrophosphotase activity, leading to
reduced enzymatic (protein) synthesis in nerve cells seen in ALS.
A separate paper by the same group (Poloni, M. et al., 1986) reconfirms the
previous study in a larger number of patients. In this study, 50 patients with ALS
and 37 controls had CSF drawn, and thiamine and thiamine monophosphate and the
ratio were measured.
The findings were as in the previous paper. A selective decrease in thiamine
monophosphate resulted in an elevated thiamine/thiamine monophosphate ratio.
This finding is almost unique to ALS. The mechanism of the selective decrease
is not clear. The decrease is not due to thiamine deficiency per se, nor is it seen
in most other neurological disorders such as dementia or Wernicke’s disease. The
authors speculate that there is a defect in the production of thiamine monophosphate
in patients with ALS.
Alzheimer’s Disease
There is some evidence to suggest changes in thiamine and thiamine-dependent
enzymes in Alzheimer’s disease (Gibson, G. et al., 1988; Rao, V. et al., 1993). In
the second of the above studies, thiamine diphosphatase was measured in frontal
and temporal cortex in eight confirmed Alzheimer’s patients, and controls. Results
showed a 62% reduction of enzyme in the temporal cortex, and a 28% decrease in
the frontal cortex, as compared to controls.
These results are in agreement with similar results from other laboratories show-
ing decreased thiamine-dependent enzyme activities in erythrocytes and fibroblasts
in Alzheimer’s patients. Collectively, these results suggest that altered thiamine
metabolism is a key feature of Alzheimer’s disease and that altered metabolism
is not limited to the CSF. How the findings of changes in thiamine diphosphatase
and alterations of TPP-dependent enzymes are related is not clear. Electron micro-
scopic studies have shown that Alzheimer’s patients have decreased synaptic densi-
ties, which could relate to the observed neurochemical results.
The above studies have been confirmed (Gold, M. et al., 1998) in a study which
looked at plasma and erythrocyte thiamine levels in a group of Alzheimer’s dis-
ease, and Parkinson’s disease patients. Results showed that the Alzheimer’s disease
patients had lower plasma thiamine concentration than did the Parkinson’s patients.
These data confirm previous findings that thiamine deficiency occurs in the plasma
of Alzheimer’s patients. The Parkinson’s group served as a control group.
A study directly measuring the thiamine-dependent enzymes pyruvate dehydro-
genase, alpha ketoglutarate dehydrogenase, and transketolase activities in temporal
cortex showed effects in Alzheimer’s disease. The results were obtained from six
Alzheimers patients and eight controls. Results showed a 70% decrease in pyruvate
dehydrogenase activity and a 70% decrease in alpha-ketoglutarate dehydrogenase
activity. Transketolase was decreased in temporal cortex by 52%. These results are
in keeping with the concept that alteration in key enzymes in energy metabolism
may be playing an important role in the pathogenesis of Alzheimer’s disease. Fur-
ther studies might examine energy metabolite levels in this disorder.
An interesting animal model of Alzheimer’s disease has been developed in which
thiamine deficiency is produced, and the pathology of plaque formation is studied
(Karuppagounder, S. et al., 2008). A mouse model of Alzheimer’s disease developed
in Tg 19959 transgenic mice was utilized. These mice over express a double mutant
form of the amyloid precursor protein. Results showed that thiamine deficiency aug-
mented amyloid plaque development. The areas affected in the cortex, hippocampus,
and thalamus increased by up to two times. Thiamine deficiency increased AB1-42
levels, B-CTF levels, and B-secretase levels significantly.
These results demonstrate, in a mouse model, that thiamine deficiency increases
plaque development and increases key protein levels. This finding is important in
that it is direct evidence of a deleterious effect of thiamine deficiency. Some form
of treatment of this defect of thiamine metabolism in Alzheimer’s disease should be
explored.
World Health Concerns
In the United States thiamine deficiency is relatively rare, and when present is usu-
ally associated with chronic alcoholism and Wernicke’s disease. In the world, how-
ever, the picture of thiamine deficiency is much different. For example, the reader
is directed to a chapter in this monograph on an outbreak of thiamine deficiency
in Nigeria called African Seasonal Ataxia. In these cases, thiamine deficiency was
caused by eating beetle larvae which contained a thiaminase. Many hundreds of
cases were reported. Other outbreaks of thiamine deficiency have occurred from
eating white cereals or from eating polished rice. Outbreaks of thiamine deficiency
have recently been associated with food ration distributions. These food aid distri-
butions usually contain food stuffs not well balanced. When consumed by people
already semi-deficient, overt thiamine deficiency can appear quickly.
In fact, a “subclinical” thiamine deficiency afflicts large numbers of the world’s
population who subsist on marginal or submarginal thiamine intake. These people
are then prime targets for developing overt thiamine deficiency given the right cir-
cumstances, and major epidemics may develop. It is sometimes difficult to predict
these outbreaks due to the lack of a simple reliable test for subclinical thiamine
deficiency.
There is a low capacity for the human body to store thiamine, the liver being the
main site. Overt thiamine deficiency can develop in as little as 2–3 weeks in subclin-
ical patients when an incomplete thiamine diet is administered. High-carbohydrate
diets and low-thiamine intake precipitate thiamine deficiency and can lead to the
easily detectable thiamine deficiency. Anorexia, a common symptom in thiamine
deficiency, can be considered beneficial in that it reduces the deleterious effect of
high carbohydrate meals. Another at-risk group is people who are at the margin
of thiamine deficiency and suddenly have increased needs. Examples include preg-
nancy and lactation, people who have heavy exertion, and those with other illnesses
such as liver disease, cancer patients, and in situations in which absorption is com-
promised such as in GI disease.
The symptoms of mild thiamine deficiency are hard to recognize and are often
attributed to other disorders. These early symptoms are quickly and easily reversed
by thiamine treatment given proper nutrients and the knowledge to administer them.
Deficient biochemical thiamine states have been associated with a variety of other


disorders, including reduced growth in children, lactic acidosis, renal dysfunction,

arrythmias, SIDS, endocarditis, cancers, and HIV/AIDS.
There were large outbreaks of thiamine deficiency in Southeast Asia around the
early 1900 s which were related to large-scale production of milled rice and large-
scale distribution. The milled rice was cheap and popular. Milled rice is a significant
risk factor, which is increased in pregnancy, fever, and muscular activity. Other risk
factors reported (Rolfe et al., 1993) in Gambia included alcohol consumption, fever,
pregnancy, diabetes, and dysentery. Milled rice constituted nearly one half of all
caloric intake. Intense activity during the agriculture/rainy season was also a risk
factor for developing overt thiamine deficiency. In some cultures, eating thiaminase
containing larvae can quickly lead to thiamine deficiency.
Breastfeeding women with signs of thiamine deficiency usually were not eat-
ing as well as those without symptoms. These women were eating well under the
minimum daily requirement of thiamine. In addition many women chewed betel
nuts, which contain a thiaminase. Thiamine-deficient mothers have thiamine levels
in their breast milk insufficient to provide enough thiamine to fulfill the minimum
daily requirement for infants (Valyasevi, A. and Vimokesant, S., 1968). Symptoms
in infants include dyspnea, GI dysfunction, vomiting, crying, cardiac signs, etc. In
infants, these signs usually progress quickly to stupor and death.
The minimum daily requirement for thiamine in adults is about 1.2 mg per day.
This number varies and depends in part on the percent of intake of carbohydrates
plus risk factors. Other factors also influence thiamine requirements such as ambient
temperatures, increasing age, body weight, and overall health.
Thiamine is widely distributed in food stuffs, both plant and animal, but most
contain low quantities. In addition, preparation (polishing) of rice, consumed in
large quantities world-wide, removes nearly all thiamine. Pasteurization of cow’s
milk removes some thiamine. The consumption of refined cereals as well as prod-
ucts made from these white cereals (bread) does not contain enough thiamine to
prevent illness and death. Other sources such as intestinal bacteria, which synthe-
size thiamine, and intestinal fermentation add to thiamine available to humans.
Strategies need to be developed and implemented in order to supplement thi-
amine availability in third world countries. Possible approaches include providing
food rations that contain a variety of food stuffs, including fruits and vegetables.
Another possible solution is to reduce the milling of rice. The idea of fortification of
food stuffs – adding thiamine to the milled rice, for example – has been around for
many years, but has not been widely adopted by manufacturers of rice or wheat flour.
The fortification process is not easily achieved, and what is available is a compound
called vitarice, which is added in the home. Vitamin supplement can also be taken
in popular foods such as biscuits or cookies enriched in vitamins. Supplementation
using vitamin tablets is logistically difficult and expensive.
Loss of thiamine can be lowered if people can be educated and decrease the
milling and polishing of rice prior to consumption. Cooking rice in lower volumes
serves to preserve thiamine. Timed cooking of both rice and vegetables serves to
lower loss of vitamins. Also, eating a prepared meal as soon as it is ready lowers
loss of vitamin. Finally, as mentioned earlier, many substances contain a thiaminase,
World Health Concerns 147
which can breakdown thiamine and produce symptoms in only a few hours. An
earlier chapter in this book details the thiaminase in larvae responsible for African
Seasonal Ataxia. Much world wide education is required.
Lest these comments be thought of as trivial or confined to just a few cases,
major outbreaks of thiamine deficiency have occurred. There are many examples of
large outbreaks. In northern and eastern areas of Thailand where rice is consumed
in large quantities, the prevalence of thiamine deficiency may always be from 5 to
8% in adults. This translates to thousands of cases. In Nepal in a 1-year period of
1993–1994, 12,000 cases were suspected. In one refugee camp in Nepal, as many as
two cases were reported per 10,000 people per day. Many more cases may go undi-
agnosed because of cultural reasons for not seeking medical advice. It is possible
that subclinical cases around the world, and in the US, are frequently undiagnosed,
yet contribute to overall morbidity.
In a recent study (Fattal-Valevski, A. et al., 2005), an outbreak of thiamine
deficiency was described in Israel. In this study, an initial case was described in a
5-month-old infant who presented with nystagmus, ophthalmoplegia, and vomiting.
Some form of Wernicke’s disease was suspected, and thiamine was administered.
Improvement was noted within several hours. Eight other infants were eventu-
ally admitted with similar symptoms. Findings included elevated blood transke-
tolase (TTP effect) to “severe” levels. Lactate and pyruvate levels were elevated
in blood and CSF in some of the patients. Three patients had neurological symp-
toms, and eight of nine had evidence of GI distress, including vomiting, diarrhea,
and in one case, intussusception. An MRI of one patient showed symmetric sig-
nals in the basal ganglia, mammillary bodies, and periaqueductal gray matter. This
patient had a rapid downhill course and became comatose. Thiamine was admin-
istered and improvement noted 24 hours later. Many patients had infectious pro-
cesses, including pneumonia, Salmonella infection, upper respiratory infections,
etc. It was finally determined that all cases had consumed a soy-based formula
(Remedia Super Soya 1) which had essentially no thiamine content. This for-
mula was quickly removed from the market and warnings distributed. Estimations
were that only about 20 cases developed probably because the formula was only
marketed a few months before outbreaks developed. This outbreak took place in
Israel, and it was suspected that between 1500 and 5000 infants had consumed the
formula.
The early clinical symptoms were at first attributed to the infections found in
these infants. The neurological symptoms of nystagmus and ophthalmoplegia were
indicative of brain stem involvement and suggested a severe condition. There are
similarities between these cases of pure thiamine deficiency, which is treatable, and
Leigh’s disease, which is currently not treatable. Once again it is noted that adminis-
tration of dextrose and other carbohydrates may cause the rapid depletion of remain-
ing thiamine, neurological symptoms. Early diagnosis and treatment with thiamine
can produce precipitating dramatic reversal of symptoms and once again points out
the presence of a biochemical lesion. In one patient in this study, thiamine treatment
did not start until the third day, and some neurological sequelae remained long after
treatment was started.
One problem with diagnosis is that the transketolase TTP effect assay is not
available in many hospitals. The lactate/pyruvate assays are not always specific for
thiamine deficiency. These factors plus the relative rarity of thiamine deficiency in
developed countries may slow down diagnosis. In addition, all patients had infec-
tious processes, which contributed to slower diagnosis.
A retrospective study of an outbreak of beriberi in a prison in the Ivory Coast
was recently published (Ahoua, L. et al., 2007). In this study, symptoms such as
bilateral leg edema, dyspnea, ataxia, and paresthesias were taken as evidence for
possible beriberi. Seven hundred and twelve cases were identified, 115 were labeled
“probable”, and 597 “definite.” The overall prison rate was 14%, and among beriberi
patients, the fatality rate was 1%.
Outbreaks of thiamine deficiency (beriberi) in prisons are not uncommon, and
as in this case, the main diet was composed of rice. Additional food sources were
a prison garden and food brought in by families. These extra activities around the
time of the outbreak had been suspended because of political crises both inside and
outside the prison.
The outbreak in this prison was recognized early and thiamine treatment initi-
ated. This resulted in 99% of the patients benefiting from the treatment. Only seven
patients died, yielding a fatality rate of 1%. The outbreak also resulted in a more
equitable distribution of food from the kitchen such that each prisoner received
more food and each received the same portions. Malnourished prisoners received
two times the normal ration.
This paper reported a large outbreak of 712 cases in a prison population of over
5000. The outbreak was higher in some buildings than others due to unequal food
distribution. It was also noted that prisoners with a history of cholera had a higher
chance of developing beriberi. The author concludes that it is important to assess
the quality and quantity of food rations in prison settings. Mild cases of beriberi
may well go undiagnosed by healthcare providers. They should be alert for signs
and symptoms in their population of patients.
Another outbreak of beriberi has recently been described in commercial fisher-
men in Thailand (Doung-ngern, P. et al., 2007). In this outbreak, the crew of a large
fishing vessel (28 members) were those who had symptoms such as lower extrem-
ity edema, dyspnea, chest pain and discomfort, and extremity numbness. Hospital
laboratory tests included transketolase and the TTP effect test.
Results from this outbreak study were that 15 out of 28 developed symptoms
of beriberi, and of these, 2 died. All affected patients were hypertensive, nine had
pitting edema, five had ascites, etc. Three who had laboratory tests were all positive
for thiamine deficiency. Four of the patients had X-ray evidence of cardiomegaly
and pulmonary edema. The two patients who died were the first two who developed
symptoms, and each died 2 days after onset of symptoms. The next group of patients
was treated with thiamine and dramatically improved. Diets of affected sailors con-
sisted mainly of fish and rice.
The authors point out that follow-up to this outbreak was difficult because crew
members dispersed when the ship docked. The authors tried through newspapers to
remind people that the risk for developing beriberi was still persistent. It is important
World Health Concerns 149
for healthcare providers to be aware of the signs and symptoms of beriberi and to
develop plans for the nutritional well-being of sailors and crew of vessels carrying
their own limited food supplies.
Another place where mild thiamine deficiency surely is associated with serious
illness would be in Africa among the millions with HIV/AIDS. Even the toll of
HIV/AIDS alone is difficult to estimate because of under-reporting. The recogni-
tion of the possibility that more than 25% might also be thiamine deficient probably
has not been considered. But as treatment develops, the co-presence of other dis-
eases such as malaria and thiamine deficiency must be considered. Remember that
millions of indigenous people live at substandard dietary intake.
Epilogue: Future Prospects
This final chapter is a collection of thoughts, questions, prospects, and speculations

regarding some aspects of the previous chapters. Some issues raised in this chap-
ter may be trivial or have already been answered. Others will hopefully stimulate
discussion and perhaps lead to further investigations into the pathogenesis of thi-
amine deficiency and associated disorders. These comments are arranged in the
same sequence as the chapter headings.
Animal Models of Thiamine Deficiency
One seemingly unanswered question remains the mechanism by which thiamine

deficiency produces such striking focal bilateral symmetric lesions. This question is
unanswered both in experimental animal models and in human examples of thiamine
deficiency such as Wernicke’s disease.
In experimental animal models, the model most likely to generate useful data is
that of the pure thiamine-deficient diet. The combined use of diet and antimetabo-
lite (pyrithiamine) to produce overt thiamine deficiency, although faster than diet
alone, clearly produces results which are somewhat at variance with the diet model
alone. These differences include lesions in different areas, and neurochemical results
which are at variance with the more chronic model produced by thiamine-deficient
diet alone. Certainly Wernicke’s disease is a chronic outcome, and any animal model
should be as close to the human counterpart as is possible. The diet alone model
takes 4–4.5 weeks, a greater percentage of the lifespan of rats than anti-thiamine
models.
Several investigators have attempted to discern the regional mechanisms in
thiamine-deficient animals responsible for the highly selective lesions. There also
have been studies centered on whole brain assays. These studies, by definition,
include in the assay material large quantities of tissue which may be non-affected –
at least by lesion placement. Inclusion in the sample of non-affected tissue dilutes
out the affected tissue, rendering results much less interpretable. If a brain region
such as a portion of the pons weighing 30 mg in a rat is homogenized with the rest of
the brain weighing 1.2 g, any meaningful change in the pons will certainly be lost in


data from whole brain. It will be lost through dilution, through animal (biological)
variation, and through methodology resolution.
One other frequently omitted feature of a good thiamine deficiency study is the
inclusion of controls called “pair-fed controls.” These animals are fed the exact
amount of food consumed by the thiamine-deficient animals. Each day a weighed
amount, for example, 20 g, is given to each individually caged thiamine-deficient
animal. Next day at the same time, the leftover food, both in a cup and dropped,
is weighed. This determines the amount consumed by the deficient animal. This
amount, for example, 3 g, is then offered to the pair-fed control in a one-to-one pair-
ing. This provides a semi-starvation control for each thiamine-deficient animal. This
accounts for the effects of starvation. Although a 30-day period for the development
of symptoms and pair-feeding is more labor intensive, the data derived from such a
study is much more meaningful.
Studies directed at the mechanisms underlying such highly specific localized
changes in brain regions such as the thalamic nuclei or cerebellar layers need to
be performed. Assays that can be coupled to NAD/NADP have been developed over
the last 50 years by Passonneau and Lowry (1993). These methods allow measure-
ment of metabolites directly in single cells. Use of these methods will yield clear
answers as to the mechanisms producing highly regional changes in neurochemistry
in thiamine deficiency. Remember, studies can be done in experimental animals
which cannot be done in man. Why not use the best assay techniques and the animal
models most applicable to man?
Remember that animal models represent a unique opportunity to control and
manipulate the conditions of the experiment. Carefully controlled experiments are
possible, unlike gathering data from human subjects. Animal experiments can be
designed very precisely, and controls made nearly perfect. Why compromise this
opportunity by not using the best possible model? Or by not having the appropriate
controls, or not obtaining results from the most selective brain region? Use direct
biochemical methods of analysis designed to answer valid questions, not methods
which happen to be available, then try to link results through many steps to possible
mechanisms of neuropathology.
Anesthetics prior to sacrifice, or actual method of sacrifice, sometimes mandated
by animal-care committees, often produce their own modification of neurochemi-
cals. It is sometimes difficult to determine these possible effects due to the nature
of the regulation. Scientists should take a more active role in the sense that arbitrary
creation of these animal-care regulations, many of which have acted to significantly
increase the cost of medical research, also may compromise data validity.
Another advantage of animal experimentation is to control the killing method
such that compounds to be measured are fixed in as close to in vivo as is possi-
ble. For example, when labile metabolites are to be measured, rapid inactivation of
metabolic processes is essential. An example of how this might lead to spurious data
is shown in Fig. 1. Imagine that at 0 seconds levels of a labile metabolite are differ-
ent than control levels and that their decay rates are different. Note that if freezing
does not occur until 30 seconds, then levels appear the same. Some years ago, much
time and effort was expended developing rapid killing methods. Devices such as
Epilogue: Future Prospects 153
Fig. 1 Figure showing decay of labile metabolites over time unless frozen
freeze blowers, brain choppers, and high intensity microwave ovens were created to
minimize killing artifact. These devices yielded data as close to true in vivo values
as have ever been seen.
Beriberi
In the past, beriberi was a serious health risk for millions of people. This was largely
linked to the consumption of polished rice. Vitamin B1 exists in the outer shells of
the rice, which are removed in the process of producing polished rice. Many people
were by circumstance forced to eat most meals consisting of almost exclusively pol-
ished rice. As overall world diet improved and knowledge increased, the incidence
of beriberi dropped.
Nevertheless, outbreaks of beriberi continued to occur. These were often on a
smaller scale than before; still hundreds might be affected in these outbreaks. For
example, in prison populations a change in rice venders or source could quickly trig-
ger an outbreak in people whose nutritional status was already threatened. If such
an outbreak was recognized and treated the outcome could be relatively positive.
If medical care was spotty, and patients not treated, then significant sequelae could
occur. These outbreaks are frequently under-reported or not reported due to possi-
ble political ramifications. This is currently (2009) being witnessed in the cholera
epidemic in Zimbabwe.
A solution to the problem of isolated outbreaks of beriberi is actually simple
and inexpensive. The first part of a solution must be education. Education about
beriberi, how it occurs, how it is treated, etc., should be directed to both healthcare
workers and the general population. This can be accomplished in schools, healthcare
training, and brochures and fliers widely distributed.
A second aspect of dealing with an outbreak of suspected beriberi is to have
some simple test or clinical criteria with which to base a diagnosis. Hospitals in
the western world currently use the measure of the thiamine-dependent enzyme
transketolase and the effect of added thiamine pyrophosphate on enzyme activ-
ity as a measure of thiamine deficiency. This assay in its present form is not a
realistic possibility for third-world countries. Some simpler test needs to be avail-
able so that healthcare workers can be assured they are pursuing correct treatment
procedures.
The third aspect of delivering proper healthcare is to have on hand large numbers
of vitamin supplementation including thiamine. The cost of this treatment is almost
insignificant relative to the health benefits achieved. The shelf life of such vitamins
is measured in years. It should not be a major issue to develop a strategy for distribu-
tion to even the most remote areas. It would be easy to develop a distribution plan in
which geographic areas with a high risk could be supplied first and regularly. These
needs are, of course, not as pressing as fighting AIDS or malaria, but the cost/health
benefit is highly positive. World health organizations or grants to local governments
(with oversight) should proceed with such plans.
Wernicke’s Disease
The impact of alcohol abuse in the US is staggering. It has been estimated that there
are at least 15 million alcoholics, and many cases go unreported. Wernicke’s disease
is thought to occur in about 12% of alcoholics, meaning almost 2 million cases. The
cost of this in terms of lost wages, care, etc., is overwhelming. Adding thiamine to
alcoholic beverages would go a long way toward reducing the impact of alcoholism
and Wernicke’s disease. This has been resisted by distillers (for cost reasons), yet
folates have been added to food (bread) for decades. Legislation to affect this change
should be passed as soon as possible (Figs. 2 and 3).
The lesions – development, location, etc. – of Wernicke’s disease are an enigma.
All who view them are stricken by the bilateral symmetry and highly localized
nature of the lesions. Microscopically, the tissue transforms from normal to diseased
in a linear space of one or two cells. The reason for this highly localized transfor-
mation is totally speculative. Various reasons which have been advanced include
the following: some unique embryological reason, blood flow differences from sur-
rounding tissue, a different energy metabolism in affected tissue, different enzyme
reserve capacity, etc.
John Blass and coworkers 30+ years ago showed that various areas of the cere-
bellum had unique biochemical features. One of these was that the enzymatic abil-
ities as regards pyruvate dehydrogenase, for example, were only marginally greater
than flux required, as compared to other brain regions where there was a several
fold reserve capacity. These kind of sophisticated assays have not been applied to
Fig. 2 Brain slice from a Leigh’s patient showing basal ganglial lesions, but complete sparing of
the mammillary bodies. Figure courtesy of Dr. Michael D. Norenberg
Fig. 3 Brain slice from a patient with Wernicke’s disease showing impressive bilaterally sym-
metrical lesions in the mammillary bodies, with relative sparing above. This figure together with
Fig. 2 emphasize the highly selective yet different nature of the lesions in these two related dis-
eases. How is it that such a small anatomic location as the mammillary bodies are severely affected
in one disorder, whereas they are not affected in another associated disease? Figure courtesy of
Dr. Michael D. Norenberg
small areas such as the mammillary bodies. The mammillary bodies are important
in memory/limbic system, and serve as synaptic sites. It is possible that the energy
metabolism (ATP) requirements are such that small depletions in metabolic sub-
strates could lead to cell damage and death.
The next comment also relates to the chapter on CPM/ACD/MBD. How is it
that chronic alcoholism can lead to the above three plus Wernicke’s in patients with
largely indiscernible histories? Why is it that some people develop Wernicke’s dis-
ease and another patient with identical alcohol exposure develops CPM or ACD?
Are there genetic reasons? Are there metabolic reasons?
Why do some patients with Wernicke’s respond to early and vigorous treatment
while others do not? One hypothesis was that if treatment was initiated before struc-
tural changes started, then successful treatment was possible. Now, it is recognized
that some structural changes can be reversed, especially in patients admitted to the
hospital for the first time, implying early stages. This makes early diagnosis and
treatment imperative. Some report that many – over half – of all cases of Wer-
nicke’s disease go undiagnosed and therefore untreated. This seems a highly unac-
ceptable statistic. Diagnostic and imaging strategies need to be mobilized to assure
quick and accurate diagnosis and treatment. The avoidance of great morbidity by
missing treatment opportunities far outweighs the added effort to make correct and
early diagnoses. Methods for ensuring compliance in diagnosed patients must be
stressed.
CPM/ACD/MBD
MBD is a rare consequence to chronic alcoholism, which has the usual highly spe-
cific neuropathological lesions. At first it was thought to be a disorder that was found
solely in Italian wine drinking males – a highly specific population. Soon however
it became apparent that there was no localized population affected by MBD, but
that it could affect anyone. Interestingly, there are many reported cases in which
early diagnosis and treatment can yield favorable results. The key hope for success-
ful treatment seems to lie in initiating a treatment regime at the earliest time. This
would fit the concept that generally speaking, if thiamine therapy can start as early
as possible, structural changes can be minimized.
The structural changes that one would hope might not occur are those of neuronyl
cell death. Other structural changes such as capillary alterations with extravasation
of blood into the lesion area and glial responses may be reversible and may predate
unalterable changes in neuron structure and function.
The cerebellar vermis change seen in ACD can neatly be explained, as stated
above, by the unique cerebellar metabolism described by Blass and coworkers.
The finding of a low reserve capacity of the cerebellum for pyruvate dehydroge-
nase activity should at least imply that similar circumstances might also underlie
changes seen in other very small areas. Thus, lesions in mammillary bodies and
thalamic nuclei may have the same underlying biochemical limits as does the cere-
bellar vermis. Confirmation awaits the properly designed experiments.
One problem with this hypothesis is that of timing and location of lesions. If
these lesioned areas of the brain reflect a narrow range of enzyme capacity for
catalyzing critical reactions, then how is it that some chronic alcoholics develop
CPM while others develop MBD or ACD? Maybe it is “chance” or that there are
minor enough differences in enzyme expression to account for the different disease
processes. Regarding timing, why do decades pass before biologically significant
changes occur in cerebral neurons? Why is there not a “down and out” sequence of
events as seen in beriberi outbreaks, the duration of which may occur over a few
weeks only? That is certainly the case in ASA in which patients may have been on
the edge as regards thiamine-deficient enzymes, then one meal of Anaphe venata
larvae pushes them into a neurological crisis in just a few hours. It seems unlikely
that process exists in Wernicke’s cases. Thiamine treatment in cases of ASA pro-
duces a rapid response. Thiamine treatment of MBD and Wernicke’s disease may
also begin reversal of neurological symptoms in just a few hours.
Leigh’s Disease
Leigh’s disease has had an interesting history. For a number of years, the disease was
viewed from a neuropathological viewpoint that regardless of treatment trials which
might have resulted in transient improvements, the final outcome was inevitable.
Much effort was expended on the urinary inhibitor concept, which was thought to
be an “agent” or toxin which if eliminated or rendered inactive might ameliorate the
relentless progress of the disease. Unfortunately that was not the case, and it now
appears that Leigh’s disease consists of many different phenotypic expressions of an
equal number of genetic point mutations. Each of these has, in common, alteration of
thiamine-deficient enzymatic processes, which in turn produces many, yet, similar
case presentations. The likelihood that Leigh’s disease ultimately falls under the
umbrella of the inherited ataxias is strong.
Thus all Leigh’s disease cases may represent mitochondrial defects, for example,
in pyruvate dehydrogenase, which lead to very similar phenotypes. But what does
this matter if the outcome remains the same?
The correlation is always drawn between lesions in all these cases, especially
in the comparison between Wernicke’s disease and Leigh’s disease. Except for the
mammillary body lesions in Wernicke’s disease, the lesions are very similar. Thi-
amine treatment can be beneficial in Wernicke’s disease, but except for possible
transient improvement in Leigh’s disease, thiamine does not work. The transient
improvement reported in some cases may have been subjective, or stimulated an
increased “blip” in enzyme activity. In the cases of point mutations in genes con-
trolling enzyme function a large amount of thiamine will not change the outcome,
but the appearance of the patient will appear as one with thiamine deficiency.
One area which appears not to have been tried is that of enzyme replacement
therapy. Dr. William Sly, for one, has successfully treated cases of mucopolysac-
charidoses with large doses of enzyme which have crossed the blood brain barrier
and achieved improvement in patients with this disorder. This should be tried in
Leigh’s patients.
Another hypothesis regarding similarities between Leigh’s disease and
Wernicke’s disease relates to the blood brain barrier. Alcohol has a deleterious effect
on astrocytes, decreasing their functional capacity. If the blood brain barrier is com-
promised in alcoholics, then this could actually work in favor of thiamine treatment
in that a functionally decreased blood brain barrier would be conducive to the pas-
sage of thiamine into the brain. Thiamine in serum is associated with high molecular
weight albumen and ordinarily is low in concentration in brain as compared to other
tissues. Symptoms of Leigh’s disease characteristically start around the second year
of life, when the blood brain barrier is functional and could prevent entry of thi-
amine. Studies of transport of labeled thiamine into brain with increasing age and
in the presence and absence of alcohol would answer many questions.
Another area mentioned earlier is that of metabolic rate in lesioned brain
sites. Using techniques described by Passonneau and Lowry, metabolite levels and
turnover could be measured in vivo in lesioned areas in mice or rats. This would
directly answer metabolic questions regarding maturation of energy metabolism as
well as effects of thiamine deficiency and alcohol in lesioned areas. Wernicke’s dis-
ease is a thiamine deficiency problem, and Leigh’s disease has strikingly similar
brain lesions suggesting similar biochemical similarities.
Inherited Ataxias
The above comments pertain as well to the inherited ataxias. These cases have in
common ataxia as a first symptom, and usually at an early childhood age. These all
appear to be mitochondrial defects involving several sites, but all having a very sim-
ilar phenotype. Lactic acidosis, altered enzyme levels (pyruvate decarboxylase) var-
ious neurological deficits such as nystagmus, etc. are common features. The ataxia
is frequently described as cerebellar ataxia.
Once again, the group of maladies called inherited ataxias represents a collec-
tion of defects associated largely with pyruvate dehydrogenase, and sometimes
other enzymatic deficits closely connected to thiamine-requiring enzymes. Thi-
amine treatment has been tried with little success. Ubiquinone treatment is unsuc-
cessful. Friedriech’s ataxia, another inherited ataxia has responded to a ketogenic
diet, and this treatment regime should be tried in other inherited ataxias using con-
trolled clinical trials.
It seems that there are multiple variations of the inherited ataxias which are
similar to those of Leigh’s disease. It is important to garner information regard-
ing the exact underlying mechanism of defect in these disorders. Early diagnosis
and treatment are essential for ultimate successful treatment regimes. Again, as in
Leigh’s disease, enzyme replacement therapy might prove beneficial in some of

these disorders.
Serious Illnesses and Thiamine Deficiency
The category of thiamine deficiencies associated with serious human illnesses such
as gastric bypass surgery, HIV/AIDS, ALS, and liver failure is one of considerable
interest. In some of these cases, such as gastric bypass surgery, the mechanism cre-
ating thiamine deficiency is relatively clear. In gastric bypass surgery for tumors,
decreased eating might precede surgery. Following surgery, vomiting is a frequent
problem in that it further decreases food intake. Studies show a significant occur-
rence of thiamine deficiency in these patients. This responds rapidly to thiamine
treatment. There should be no reason not to routinely administer thiamine to gastric
bypass patients.
The anticancer agent 5-fluorouracil is a drug commonly given to cancer patients
suffering from a variety of tumors. Many patients receiving 5-fluorouracil show
mild signs of thiamine deficiency, or show no signs, but test positive for thi-
amine deficiency using the transketolase assay and TPP effect. 5-fluorouracil has
been shown to diminish the formation of thiamine pyrophosphate from thiamine.
A 5-fluorouracil metabolite, fluorocitrate, may act as an inhibitor of the Krebs
cycle. How 5-fluorouracil causes impairment of normal thiamine related functions
is not clear, but this low level of thiamine deficiency induced by 5-fluorouracil is
rapidly reversible with thiamine treatment. That a low level of thiamine deficiency
might add to the metabolic burden in cancer patients is clear, so this concomi-
tant problem should be watched for, and treated quickly and aggressively when
recognized.
Another related view is that there is an increase in the activity of the pentose
phosphate pathway in some cancers in order to produce RNA. Thiamine deficiency
in cancer patients might act to decrease flux through the pathway, thereby slowing
cancer growth. Until good controlled studies are performed, these questions will
remain unanswered.
Thiamine deficiency measured by the TPP effect/transketolase assay in non-
alcoholic AIDS patients has shown a nearly 25% occurrence. In autopsy material
from these patients, cerebral lesions were similar in location and type to non-AIDS
patients dying of Wernicke’s disease. While AIDS is a wasting disease, which
might be associated with vitamin disease, the exact relation between AIDS and
thiamine deficiency remains to be elucidated. Whether thiamine deficiency signif-
icantly alters the course of AIDS, treated or untreated, awaits controlled clinical
evaluation.
The occurrence of thiamine deficiency in these and other serious illnesses is rec-
ognized. Many unanswered questions, and obvious clinical evaluation needs to be
done before these questions can be answered. The synergistic interactions of more
than one possible life-threatening illness are unknown, but are probably not good. In
the very least case, healthcare workers should be aware of this potential for thiamine
deficiency, watch patients for signs and symptoms, and initiate therapy as soon as
thiamine deficiency is suspected.
Conclusions
It is hoped that this chapter will have highlighted the many unanswered questions
related to thiamine deficiency and associated clinical disorders. And perhaps sug-
gested many more. Controlled clinical studies are not always easy to perform, for
example, in a pediatric setting, but are essential to provide useful information. In the
absence of this type of information, false leads may send investigators down barren
pathways. More use of animal models – carefully controlled – should prove benefi-
cial. Experiments can be designed to answer specific questions if there is attention
to detail. I am not trying to belabor this point, but what can be gained by asking
questions about circumstances in the mammillary bodies, then grinding up whole
brain looking for answers? All too frequently investigators use techniques that are
“up and running” to look at a question/problem rather than designing methodol-
ogy to directly answer questions. These circuitous approaches to research are non-
productive. Time is better spent doing it right.
Thiamine deficiency and associated clinical disorders represent an exciting field
of study. The earliest observations regarding the reversibility of symptoms in these
conditions led to the concept of a biochemical lesion. The reversibility “feature” of
thiamine deficiency has proved to be highly beneficial. Symptoms such as opistho-
tonus and ataxia can be reversed in only hours by thiamine treatment. This can be
used to test the importance of biochemical changes in specific areas. Thus, if symp-
toms revert to normal in 3 hours, while decreases in various metabolites remain
unchanged, an experimental biochemical dissection has been performed.
More importantly, the reversal of symptoms carries over to patients suffering
from disorders such as beriberi and Wernicke’s disease. This type translation is
essential. Frequently when recognized early, symptoms of beriberi heart disease,
and Wernicke’s symptoms such as nystagmus revert toward normal in hours. This
feature of biochemical lesions permits direct evaluation of mechanisms in these dis-
orders. This provides hope for enigmatic problems such as seen in Leigh’s disease.
If an as yet unknown treatment becomes available, results in early diagnosed cases
might be apparent in a remarkably short time. This is possible only if attention is
paid to detail, and if carefully controlled, and carefully conducted experiments are
designed and executed.
A key paper is that of Reynolds and Blass (1976), showing a minimal excess
of pyruvate dehydrogenase in the cerebellum. Just as there is a normal range of
blood glucose levels, there is likely a normal range for brain enzymes. This may
vary from region to region, and from person to person. This could explain why one
individual drinking for 25 years develops Wernicke’s disease, and another with the
same history develops CPM. The different results could be attributable to different
ranges of focal brain enzyme capacities. This could awaits experimental verification.
Index
Note: Appendix pages have been indexed with appendix number in brackets followed by “app”
A consumption, 104
Aalsmeer, W., 21 nutrients in, 103
Abarbanel, J., 135 Andrews, V., 35
Absalon, M., 96 Anemia, 87
Acetylcholine, 21 Anesthetics, 152
decreased synthesis of, 25 Anorexia, 145
Acetyl moiety of acetyl-CoA, 6 Aphonia, 36
Active form of thiamine, see Thiamine Arbelaez, A., 76
pyrophosphate/cocarboxylase Ashiru, M., 103
Adamolekun, B., 104, 105, 106, 107, 108, 109 Aspartate, 26–27
Adams, R., 48, 49, 59, 65, 71 Astrocytes, 158
Adenine nucleotides, thiamine-deficient rats, Astrocytes, lesions in
22 central pontine myelinosos, 66
African seasonal ataxia (ASA), 103–112 in mammals, 23
lesions, 110–111 Astrocytic proliferation, 73, 83
symptoms, 105 Ataxia, 13–14, 21, 24, 35, 55, 56, 72, 76, 105,
thiamine, 111 106, 113, 122, 124, 157
A 8344 G mutation, Leigh’s disease, 97–98 and beriberi, 35
Ahoua, L., 148 recovery, 58–59
AIDS, 139, 146, 149, 159 rhesus monkeys with thiamine-deficient
Aikawa, H., 24 diet, 25
Akert, K., 55 TCA cycle enzymes, decreases in, 117
Alanine, 27, 113 and Wernicke’s disease, 55
Alcohol, 158 See also African seasonal ataxia (ASA);
Alcoholic cerebellar degeneration, 72–75 Inherited ataxias
etiology, 74–75 ATP, 3, 5, 6, 21, 22, 38, 39, 42, 88, 119, 156
neuropathological examination, 73 cardiac, 41
symptoms, 72 pyrithiamine treatment and, 28
Alpha-ketoglutarate dehydrogenase complex, 4 ATP/ADP ratio, 6, 116
Alpha-ketoglutarate oxidation, 18 Attas, M., 39, 44
Alzheimer’s disease, thiamine deficiency in, Axonyl degeneration, 43, 45
143
Alzheimer’s glia, 52 B
Ambrose, M., 62 Basal ganglia, 27, 69, 82, 87, 118, 147
Amprolium, 17, 30 Basu, T., 132
Amyloid plaques, 143 Beetle larvae, 104, 105, 106, 109
Amyotrophic lateral sclerosis (ALS), 142 Benit, P., 98
Anaphe venata, 103, 104, 107, 108, 109, 145, Berendzen, K., 123
157 Bergmann’s astrocytes, 73
185
186 Index
Beriberi, 31–46 and thiamine diet, 29

blood findings, 34 Central nervous system dysfunction, 118–119
chest X-rays, 39 Central neuritis, 69
clinical account, 33 Central pontine myelinosos (CPM), 65–71
clinical manifestations, 32 cases of, 69–70
countries reporting outbreaks of, 35 clinical symptoms, 77
death due to, 34 hyperosmolality, 71
Eijkman’s speech on, 163–169 (app C) lesions, 67–68
first medical description of, 1, 161 (app A) massive demyelination, 66
infantile form of, 35, 36 cortical spinal tracts, 67
in male adults, categories, 31–32 microscopic lesion, 71
thiamine deficiency in, 131–132 neuropathological features, 65
thiamine treatment, 44–45 Cerebellar ataxia
vasculature aspect of, 44 abnormalities of PDH, 115
wet and dry, 45–46 ASA and, 106–107
Berkovic, S., 97 attacks, case study and treatment, 113–114
Bessey, O., 14 treatment, 115
Betel nuts, 146 Cerebellar cells, thiamine deficiency on, 28
Betrosian, A., 45 Cerebellar granule cells (CGCs), 28
Bignami, A., 75 Cerebellar layers, Purkinje cells and Bergmann
“Biochemical lesion,” 34, 42 glia, 57
Blass, J., 25, 75, 90, 113, 114, 115, 116, 117, Cerebellar metabolism, 156
154, 156 Cerebellar vermis, 49, 57, 61, 73, 74, 156
Blood brain barrier, 158 Cerebellum, 21, 23, 27, 29, 49, 56, 61, 70, 73,
Boniol, S., 139 110, 152, 156
Bontius, J., 31 Cerebral blood flow, 24–25, 154
description of Beriberi, 161 (app A) Cerebral cortex, 52, 69, 70, 73, 76, 77, 78, 79,
Boros, L., 138 80, 85, 92
Bradley, W., 71 Cerebral spinal fluid, 60
Brain chopper, 153 Chang, L., 7
Brain stem, 21, 82, 83, 89, 147 Charcot-Marie-Tooth disease, 97
effect of Wernicke’s disease, 52 Chastek Paralysis, 109–110
of thiamine deficient rats, 21 lesions, 110
Brin, M., 18, 20, 33, 37, 168 Cheney, D., 21
Brunette, M., 114 Cholera, 153
Burgess, J. A., 35 Chou, S., 89
Butterworth, R., 19, 26, 27, 61, 140 Clayton, B., 85, 86
Cline, J., 2
C Clinical account, beriberi, 33
CABC1 gene, defect in, 120–121 Coakham, H., 105
Calza, L., 140 Cole, M., 69, 71
Cancer, thiamine deficiency in, 132–133 Collins, G., 48, 49
Carbohydrates Coma, 55, 84, 100, 105
metabolism of Complex 1, 100
thiamine pyrophosphate in, 3–4 defect in, 118
and vitamin requirement, 9–10 deficiency, 101
Cardiac metabolism, 41 Cooper, J., 19, 86
Cardiopathology, 34 Corpus callosum, 68, 73, 75, 76, 115
Cassidy, C., 22, 25 Coxon, R., 15
Caudate nucleus, 118 CPM, see Central pontine myelinosos (CPM)
Celik, Y., 54 Craig, L., 90
Cell death Cranial neuropathy, 58
pyrithiamine treatment and, 28 Crome, L., 84
Index 187
Crosby, T., 89 brain at level of superior cerebellar

Cruickshank, A., 133 peduncle, 120
Cytochrome c oxidase deficiency, 95 cerebellum, 119
“dying back of neurons” in, 122
D lumbar spinal cord, 124
Debray, F., 124, 125 olive in, 121
De Langen, C., 32, 33 thoracic spinal cord, 123
Demaerel, P., 77 Fukumoto, J., 79
Demyelination Funk, C., 2, 10
in ataxia, diffuse and symmetrical, 116–117
Leigh’s disease, 85, 92 G
in peripheral nerves, 14 Gamper, E., 47, 52
Denny-Brown, D., 32, 34, 131 Gastrectomy
DeVivo, D., 91 thiamine deficiency in, 134–136
Dickerson, J., 132 Wernicke’s disease and, 139
DiMauro, S., 95 Gavrilescu, N., 11, 12
DNA, 95, 96, 97 Geel, S., 25
Doherty, M., 53 Gibson, G., 19, 116, 143
Dolman, C., 117 Girard, P., 48
Donath, W., 2, 169 Gliosis, 50
Doung-ngern, P., 148 “Global confusional state,” 55
Dow, R., 74 Glucose, 15
Dreyfus, P., 18, 25 Glucose absorption, thiamine defeciency, 15
Drummond, J., 10 Glutamate, effect on, 26–27
Dunn, H., 117 Glycolysis and TCA cycle, 19
Durck, H., 34 Gold, M., 143
Goswami, P., 78
E Grove, I., 23
Edema, 32, 34 Grover, W. D., 84
African seasonal ataxia, 111 Growth curve, thiamine-deficient rats, 17
due to beriberi, 34, 35, 37–38 Grunnet, M., 60
Eijkman, C., 1, 2, 163–169 Gubler, C., 4, 17, 37
speech on antineuritic vitamin and beriberi, Gubler study, 17
163–169 (app C) Gudden, H., 47
Electron microscopy, 5
Endoplasmic reticulum stress (ER stress), H
29–30 Haas, L., 78
Entomophagy, 103, 145 Hakim, A., 24, 25
“Epilepsia partialis continua,” 120, 121 Harper, C., 55, 60, 61
Harper, H., 15
F Harrington, D., 23
Fattal-Valevski, A., 147 Harris, C., 79
Feigin, I., 54, 83 Hawes, R., 33
Ferracci, F., 77 Hayashi, T., 76
Fibrose gliosis, 74 Heart
5-fluorouracil, 132 isolated perfused, 42
and decreased transketolase activity, 132 thiamine-deficient, 40
thiamine deficiency in, 136–139 Heart disease, beriberi, 37
Follis, R., 34 Heart failure, 32, 37, 40, 41, 43,
Foresti, V., 139 45, 46
Fournier, H., 27 Hematologic tumors, thiamine deficiency in,
Freeze blower, 153 134
Friedereich’s ataxia, 117 Hemodynamics in alcoholic patients with
biochemical lesion, 117 beriberi heart disease, 43
188 Index
Hemorrhages Karuppagounder, S., 143

Marchiafava–Bignami disease, 79 Kashi, M., 58
in Wernicke’s disease, 49–50, 61 Katz, J., 41
Henneman, E., 74 Kaya, M., 54
Henwood, M., 125 Ke, Z., 28
Heroux, M., 26 Kerr, D., 97, 126
Hess, W., 10, 55 Kim, M., 79
Hexose monophosphate shunt, 22–23 Kinnersley, H., 10, 11
Hirayama, K., 80 Koike, H., 134
HIV/AIDS, thiamine deficiency in, 140–141 Koike, M., 4
Ho, L., 95, 119 Kornreich, L., 48, 64
Hodgkin, W., 91 Korsakoff, S., 41, 47
Hommes, F., 85, 88 Korsakoff’s psychosis
Hyperemesis gravidarum, 45 lesions and monoamine containing
Hyperosmolality, CPM, 71 neurons, 60
norepinephrine, dopamine, and serotonin
I levels in, 60
Ibikunle, F., 107 thiamine administration, 59–60
Iiizumi, M., 121 vs. Wernicke’s disease, 47–48, 50
Inanition, 13 Krampitz, L., 111
Inborn errors of metabolism (IEM), 122 Krebs cycle (TCA) cycle, 3, 4, 91, 97, 115
See also Ataxia Krill, J., 62, 63
Indraccolo, U., 45 Krishna, S., 141
Infantile beriberi, 35, 36 Kunz, W., 125
treatment of, 36 Kuo, C., 77
Inherited ataxias, 113–129
Leigh’s disease, overlap of, 128
L
symptoms, 113
Labile glycoprotein, 86
thiamine treatment, 115, 129
See also Friedereich’s ataxia Labile metabolites, 153
Intensive care units, thiamine deficiency in, Lacasse, L., 60
133–134 Lactate
Interstitial proliferation in beriberi, 38 elevated, 21
Intracerebral hemorrhage, Wernicke’s disease, myocardial utilization of, 38
52 role in symptoms generation, 9
Itokawa, Y., 16 Lactate/pyruvate ratio, 14, 91, 118, 119, 122
IV Parentrovite, 106 Lactic acid
Iwalewa, E., 108 and avitaminosis pigeons brains, 10–11
increased in opisthotonic stage, 11
J Leigh, D., 81, 82
Jackson, F., 108 Leigh’s disease, 81–101
Jagadha, V., 48, 49, 61 brain, 93
Jane, J., 84, 85 findings on, 81, 85
Jansen, B., 2, 169 lesions in, 98, 99
Johnson, L., 4 cases of death, 92
Jope, R., 116 cortex and subcortex, acute deterioration,
Jubb, K., 111 127
Juhasz-Pocsine, K., 138 effects on liver, 85–86
findings, 81, 85
K hypointense putamina, 99
Kamoshita, S., 84 inhibitor, 90
Kang, H., 125 Krebs cycle intermediates, 90
Kant, F., 47 lactate levels, 91
Karapanayiotides, T., 139 and MELAS, 101
Index 189
mitochondrial DNA depletion, 96–97 lesions, 75

myelination delayed, 100 MRI as diagnostic technique, 78
neuronal damage, 81 symptoms, 76
PDC deficiency, 94–96 thiamine treatment, 77
periventricular white matter, 126 types of, 79
putamina/thalamus at acute deterioration, Marrian, F., 10
127 Mayxay, M., 141
swollen caudate nuclei, 127 McCandless, D., 19, 21, 22, 23, 24, 25, 28, 41,
symptoms, 84 45, 91, 109, 116
treatment McEntee, W., 60
L-glutamine, 91 McKee, A., 71
thiamine, 86, 88 Mediodorsal thalamic nucleus, Wernicke’s
Lesions, 13–14 disease, 53
in brainstem and cerebellum, 21 lesioned areas, 63
central pontine myelinosos, 67–68 Mediodorsal thalamus
due to pyrithiamine-induced thiamine lesioned mammillary bodies and, 62
deficiency, 24 in Wernicke’s disease, 54
Korsakoff’s psychosis/Wernicke’s disease, Meneegon, P., 77
47, 52–53 Meng, J., 29
Marchiafava–Bignami disease, 75 Merritt, H., 60
mechanisms of distribution, 29 Messerle, H., 10
nerve and vascular, 23 Mesulam, M., 25
neuropathological findings, 23 Metabolite turnover, 22
in Wernicke’s disease, 49, 61 Mianserin hydrochloride, 79
Lewis, C., 116 Microglia
Liang, C., 15 effects of treatment after thiamine
Lichtenstein, A., 32, 33 deficiency, 29
Lim, E., 32 Leigh’s disease, 81, 85
Lipoamide dehydrogenase, 117 Midbrain, 25, 85, 99
Liu, Y., 48, 53 Miranda, A., 95
Liver, 7 Mitochondria, 6, 89, 101, 125
Localized lesions, 154 Mitochondrial disorders, 125
Lofland, H., 16 Mitochondrial DNA depletion, 96–97
Lohmann, K., 3 Mollet, J., 120
Lowry, O., 22, 38, 152, 158 Moore, D., 136
Lowry “closed system,” 22 Morcos, Z., 60
Lum, C., 60 Moreadith, R., 118
Morris, A., 97
M Moruzzi, G., 74
Magnesium, 3 MRI, 62, 63, 64, 77, 78, 80, 97, 98, 100, 101,
Magnesium deficiency, relationship between 126, 127, 128, 140
thiamine and, 20 Muri, R., 140
Malamud, N., 48, 49, 50 Murine pyruvate dehydrogenase deficiency, 27
Malaria, thiamine deficiency in, 141–142 Murphy, J., 88, 90
Mammals Myelin, 24, 43, 65, 67, 69, 71, 111
methods of producing thiamine deficiency Myelinated fibers
in, 4 beriberi, 43
thiamine deficiency in, 17–30 Leigh’s disease, 83
Mammillary bodies, 23, 49, 51, 52, 53, 69, 84, Myocardial isometric tension, 39–40
87, 140, 156, 157
Mancall, E., 71 N
Marchiafava, E., 75 NADH:ubiquinone oxidoreductase, see
Marchiafava–Bignami disease (MBD), 75–80 Complex 1
190 Index
Ndububa, D., 107 due to Wernicke’s disease, 50

Necrotic muscle fibers, in beriberi, 38 Peripheral neuritis, 44
Nerve conduction Peters, R., 10, 11, 12, 13, 15
thiamine in, 19–20 Phillips, G., 59
TTP in, 19–20 Phosphocreatine, thiamine-deficient rats, 22
Neural degeneration Piatnek, D., 44
Leigh’s disease, 85 Pigeons, 1, 12
Neuronal damage, Leigh’s disease, 81 Pigeons, early studies of thiamine deficiency
Neurons on, 9–10
African seasonal ataxia, 111 Pincus, J., 19, 23, 86, 88, 89
replaced by glial fibers and astrocytes, 51 Plaitakis, A., 26
Neuropathological component, 42–43 Platt, B., 32
Neurotransmitter serotonin, effect on, 26 Pliss, L., 27
Nishimune, T., 108 Polished rice, 1, 2, 10, 12, 14, 37, 40, 146, 153
Nixon, P., 137 Poloni, M., 142
Nose, Y., 5 Polyneuritis, 2
Novak, D., 58 Polyneuropathy
Nucleoside analogue reverse-transcriptase and beriberi, 45
inhibitors (NARTI), 140 MBD and, 45
Nystagmus Wernicke’s disease, 58
Leigh’s disease, 101 Pons, 49, 51, 65, 66, 70, 82, 151
symptoms of thiamine deficiency, 26 central pontine myelinolysis, 65
Wernicke’s disease, 56 Leigh’s disease, 85
lesions due to, 81
O
Pure dietary thiamine-deficient diet, Gubler
Okeda, R., 29
study, 17
Oligodendrocytes, 81
Purify and isolate vitamin, 2
Olson, R., 44
Purkinje cell layer, effect of Wernicke’s
Omeprazole, 137
disease, 52
Opisthotonus, 10, 11, 12, 13, 14–15, 24
Pyrithiamine, 4, 17, 18, 19, 20, 23, 26, 27, 28,
Osuntokun, B., 105
29, 37
Oxygen consumption/uptake
in normal and beriberi tissues, 10 Pyrithiamine-induced thiamine deficiency, 23
in tissues and vitamin B, 10, 13 advantages of, 27
vitamin B1 addition to brain areas, 12 neuropathological and biochemical effects,
during vitamin B deficiency induced 24
symptoms, 11 on rat brain myelination, 24
Oxygen uptake, 11, 12 Pyruvate, 13, 15, 18, 21, 24, 38, 39, 42, 44,
Oxythiamine, 4, 17, 18, 19, 37 116, 147
Oxythiamine pyrophosphate (OTPP), 37 Pyruvate carboxylase, 85
developmental pattern of, 7
P Pyruvate decarboxylase, 21
Pair-fed controls, 152 reserve capacity, 22
Pannunzio, P., 28 reversal of enzyme activity/symptoms, 22
Pappius, H., 25 Pyruvate decarboxylation and contraction of
Parkinson’s disease, 60 myofibrils, 44
Passonneau, J., 152, 158 Pyruvate dehydrogenase (PDH), 3, 4–5, 114
Patel, M., 6 activity, 18, 28
Pentose phosphate pathway and thiamine in fibroblasts, 117–118
deficiency, 41 increased, dichloroacetate, 123
Periaqueductal gray, 47, 48, 53, 60, 63, 67, 87, biosynthesis of, 5
139, 140, 147 deficiency, 119
Peripheral nerves changes treatment, 125
due to beriberi, 34 factors modulating activity, 116
Index 191
low activity of, 115 Structure for thiamine, 2

Pyruvate metabolism, ataxia and, 115 Stupor, 79, 84
Pyruvic acid, elevation and, 34 Subacute necrotizing encephalomyelopathy
(SNE), see Leigh’s disease
R Sub-acute neurotoxic behaviors, 109
“Ragged-red fibers,” 89, 114 Suzuki, T., 40, 79, 174
Rahman, S., 96 Swank, R., 13
Read, D., 23 Symptom reversal, 9, 11, 21, 22, 23, 40, 43,
Red blood cell transketolase activation, 27 58, 59, 80, 107, 147
Reed, L., 3, 4 Synergistic interactions, 159
Regional studies, 154, 160 Synthesis in laboratory, 2
Rembrandt, 31
Reynolds, S., 25, 75, 116
T
Richardson, E., 71
Tattevin, P., 140
Richter, R., 82, 83
Taylor, S., 6
Rolfe, M., 146
TCA cycle, 19, 117
Rosenbaum, M., 60
Thalamus, 18, 26, 29, 30, 47, 49, 52, 53, 54,
Rosenberg, S., 139
61, 63, 67, 81, 85, 87, 157
Rosenblum, W., 54
Leigh’s disease, 85
Rydin, H., 12
Thiaminase, 108, 111, 147
S Thiamine, base of, 2
Santorelli, F., 96 Thiamine deficiency, 137
Sarcoplasmic reticulum, thiamine-deficient, 40 mild, symptoms, 145
Sarcosomes’ oxidative phosphorylation, effects outbreaks
on, 38 food ration distributions, 145
Saudubray, J., 116 in prisons, 148
Schenker, S., 19, 21, 24, 28, 109, 116 reversal feature, 1
Schiff, M., 126, 127, 128 Thiamine pyrophosphate/cocarboxylase, 3,
Schramm, C., 141 15–16
Schuster, P., 3 role in enzymes of carbohydrate
Schwartzenburg, F., 22 metabolism, 3–4
Sciatic nerve, degeneration, 13–14 Thiamine-requiring enzymes, 16
Sedel, F., 122, 128 and disease process, 61–62
Seizures, 21, 82, 109, 114, 120, 121, 123, 125 Thiamine synonyms, 162 (app B)
Seizures, thiamine deficiency, 24 Thiamine transport, 158
Sekiyama, S., 138 Thiamine treatment
Selective vulnerability, 28 beriberi, 44–45
Seligmann, H., 135, 137 Chastek paralysis, 85
Shanske, S., 101 gastrectomy, 135
Shen, L., 5 HIV/AIDS and, 140
Sheu, K., 117 inherited ataxias, 114–115
Shimazono, J., 33, 44 Leigh’s Disease, 87–88
Shimomura, T., 135 MBD, 76–77
Sijens, P., 101 symptoms reversed, 9, 14
Silkworm larvae, see Anaphe venata Wernicke’s disease, 62
Simopoulos, A., 87 Thiamine triphosphate, 88, 89, 90, 132, 133,
Skillicorn, S., 48, 49, 50 136, 148
Spinal cords, 87, 89, 123, 124 Tilghman, S., 6
degeneration due to beriberi, 34 Tobita, M., 76
effect of Wernicke’s disease, 52 Tomita, I., 6
effects of thiamine deficiency, 20 Tomlinson, B., 70, 71
Steyn-Parve, E., 19, 37 Toshima, K., 92
Stoltz, E., 14 Tran, H., 46
192 Index
Transketolase, 16, 18, 22, 24, 25, 27, 41, 132, mental confusion, 55
135 thiamine therapy and its importance, 59
decreased activity, 18, 28 Wernicke’s disease, 21, 47–64
red blood cell, 27 and alcoholism, 61
Tsao, C., 97 brain slices of mammillary bodies, 51
Tulp, 31 cerebellum in, 56
clinical findings in, 64
V decreases in thiamine-requiring enzymes,
Valanne, L., 99, 100 61–62
Valyasevi, A., 146 and gastrectomy, thiamine deficiency in,
Van der Lelie, J., 134 139
Van Zaanen, H., 134 hemorrhagic component of, 54–55
Vasculature aspect of beriberi, 44 incidence, distribution, and neuropathology
Vazquez, C., 80 in, 60
Victor, M., 48, 49, 55, 58, 59, 60, 72 lesions, 48, 49, 64
Vimokesant, S., 146 of mammillary bodies, close-up, 52
Vitamin, etymology, 9 mediodorsal thalamic nucleus, 53
Vitamin-deficient (beriberi) tissue, 10 spongy degeneration, 54
“Vitamine,” 2 symptoms, 49–50, 56
Vitamin supplimentation, 154 thiamine therapy, 59
Vitarice, 146 vs. Korsakoff’s psychosis, 47–48, 50
Vorderman, A., 2, 168 Williams, R. R., 2
Vorhees, C., 25 Wilson’s disease, 68, 70
Wolf, A., 83
W Wooley, D., 111
Walker, U., 140 Wright, J., 105
Wallis, W., 55
Wang, X., 29 Y
Watanabe, I., 23 Yashon, D., 84, 85
Weil-Malherbe, H., 3 Yoshinori, I., 9
Wenckebach, K., 32
Wernicke, C., 47, 105, 110 Z
Wernicke–Korsakoff syndrome, 48 Zieve, L., 20
clinical and pathological changes in, 52, 55 Zuccoli, G., 62

Early Thiamine Deficiency

Uploaded by

Copyright:

Available Formats

Early Thiamine Deficiency

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Early Thiamine Deficiency

Uploaded by

Copyright:

Available Formats

Contemporary Clinical Neuroscience

Helen A Baghdoyan, Ph.D.

For other titles published in this series, go to

ISBN 978-1-60761-310-7 e-ISBN 978-1-60761-311-4

© Humana Press, a part of Springer Science+Business Media, LLC 2010

Printed on acid-free paper

As stated above, thiamine plays an important role as coenzyme in reactions in

month 6–12 following normal development. There appears lethargy, nystagmus,

Animal Models of Thiamine Deficiency . . . . . . . . . . . . . . . . . . . 151

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 1

The active form of thiamine is thiamine pyrophosphate or cocarboxylase. This

One of the first to synthesize thiamine pyrophosphate was Weil-Malherbe (1940),

The alpha-ketoglutarate dehydrogenase complex is large (2.7 million M.W.). It

dihydrolipoyl dehydrogenase. This composition is very similar to that of alpha-

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 9

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 17

alpha-ketoglutarate was decreased by oxythiamine, but both were decreased with

(Steyn-Parve, 1967). When antimetabolites are administered, many symptoms noted

increase in amplitude of the compound action potential and reversal of post-tetanic

than pyruvate decarboxylase, duplicating previous studies. Regional results showed

The presence of two different neuropathological patterns was seen as a different

dehydrogenase was decreased regionally, whereas activities of pyruvate dehydro-

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 31

Beriberi Clinical Symptoms

Source Breathlessness Chest Palpations Edema Enlarged Heart Tachycardia Neuropathy

Frequently, the onset of symptoms is linked to a febrile episode. This is asso-

Beriberi: Frequent blood findings

Most knowledge of the pathological changes seen in patients dying of beriberi

too is decreasing as knowledge of the dangers of highly milled rice increases.

pyruvate–fumarate substrates. Decreased inhibition of swelling shows a decreased

on this physiological phenomenon (Aldinger, E., 1965). Thiamine deficiency was

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 47

Wernicke’s Disease—Gross Brain Lesions

Source Mammillary Thalamus Cerebellum Cerebral Periaqueductal Hypothalamus

Wernicke’s Disease—Microscopic Lesions

Source Thalamus Hypothalamus Cerebellum Mammillary Bodies Pons Medulla Fornix

Specific neuropathological changes have been described in Wernicke’s dis-

Wernicke”s Disease—Type of Microscopic Lesions

Source Cell Necrosis Decreased Increased Hemorrhage Loss of

examination of 62 cases of Wernicke–Korsakoff syndrome over several years. In

Fig. 5 Schematic cartoon at the level of the mammillary bodies

Ocular symptoms occurred in 96% of patients. These symptoms consisted of nys-

Fig. 12 Cerebellum in Wernicke’s disease. Figure courtesy of Dr. Roland Auer

27% of patients. The remaining group made an incomplete recovery, plateauing at

In a retrospective study of clinical signs (Harper et al., 1986), careful analysis

significant decreases in thiamine-requiring enzymes may be essential in the disease

to produce altered physiology, and onset of lesions. Alteration of structure in the

Classical Clinical Findings in Wernicke’s Disease

Central Pontine Myelinolysis

Central pontine myelinolysis (CPM) is a distinct disorder with characteristic CNS

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 65

Fig. 1 Photomicrograph of central pontine myelinosos showing massive demyelination. Figure

In another case, described by the above author, a malnourished alcoholic woman

Fig. 2 Photograph of central

Neuropathological findings were normal regarding superficial appearance, and

Central Pontine Myelinolysis—Microscopic Lesion

Alcoholic Cerebellar Degeneration

The disease subacute necrotizing encephalomyelopathy (SNE) was first described

D.W. McCandless, Thiamine Deficiency and Associated Clinical Disorders, 81

for myelin showed widespread demyelination in affected areas. Microscopic dam-