JOURNAL OF PHYSIOLOGY AND PHARMACOLOGY 2015, 66, 4, 549-556

www.jpp.krakow.pl

A.F. HAGEL, T. DE ROSSI, P.C. KONTUREK, H. ALBRECHT, S. WALKER, E.G. HAHN, M. RAITHEL

PLASMA HISTAMINE AND TUMOUR NECROSIS FACTOR-ALPHA LEVELS

IN CROHN’S DISEASE AND ULCERATIVE COLITIS AT VARIOUS STAGES OF DISEASE

Department of Gastroenterology, University of Erlangen, Germany

Mast cells secrete numerous mediators and this study investigated plasma levels of histamine, and tumor necrosis factor
alpha (TNF-α) in chronic inflammatory bowel disease (IBD). Plasma levels of histamine were determined in 68 patients
with Crohn’s disease (CD), 22 with ulcerative colitis (UC) and 13 controls. TNF-α levels were assessed in 29 CD
patients, 11 UC patients, and in 11 controls. Plasma histamine levels in the control group were 0.25 ng (0.14 – 0.33) and
showed no difference to CD (0.19 ng, 0.09 – 0.35) or UC (0.23 ng, 0.11 – 0.60). Significantly lower histamine levels
were only found in CD patients on 5-aminosalicylic acid treatment (P ≤ 0.04). Plasma TNF-α levels in the control group
were significantly lower 0.44 ml/m2 (0 – 1.15) than in CD patients (4.62 ml/m2, 1.82 – 9.22, P = 0.005) or UC (3.14
ml/m2; 0.08 – 11.34, P = 0.01). In CD disease activity, fistula, and extraintestinal manifestations (EM) were associated
with significantly higher plasma TNF-α values, but not the type of treatment. We concluded that in contrast to TNF-α,
histamine levels were normal in CD and UC. There is no correlation with histamine and thus the proportion of TNF-α
secreted from mast cells in the plasma in patients with IBD is less important.

K e y w o r d s : plasma histamine, tumor necrosis factor-a, inflammatory bowel disease, Crohn’s disease, ulcerative colitis

INTRODUCTION main mediators, within minutes (5, 6). Interestingly, TNF-α -

similarly to histamine - can alter vascular and intestinal
Today, idiopathic chronic inflammatory bowel diseases permeability and TNF-α can also stimulate the proliferation of
(IBD), Crohn’s disease (CD), and ulcerative colitis (UC) are smooth intestinal muscle cells and fibroblasts, which can cause
thought to be caused by defects in the interplay between genetic typical IBD lesions such as bowel wall thickening, stricture, and
predisposition, exogenous factors (intestinal flora, nutrition, stenosis (2, 6-11). The local accumulation of mast cells has been
physical activity and smoking), and a dysregulated immune reported in particular for strictures and inflammatory lesions in
response with chronic activation of immune and inflammatory IBD and thus a pathogenetic link could exist between TNF-α and
cells (1-3). CD involves discontinuous transmural inflammation histamine in CD and/or UC (6-8). Indeed, although up to now no
that can affect the entire gastrointestinal tract. The ileocecal causative or curative medical treatment exists for IBD, inhibitory
region is affected most frequently. By definition UC starts in the properties to suppress mast cell activity have been reported for
rectum and spreads continuously through the mucosa and conventional therapies such as aminosalicylates, corticosteroids,
submucosa in a proximal direction, at worst developing into dietary treatment, and immunosuppressants and for more rarely
pancolitis with or without backwash ileitis. Both diseases can be used drugs such as antihistamines and mast cell stabilizers (5-8).
associated with diverse extraintestinal manifestations (EM), Considerably higher numbers of mast cells have been found in
which in CD often also involve formation of fistulas (1, 2). IBD affected areas in both CD and UC patients than in nonaffected areas
is usually diagnosed serologically, clinically, endoscopically, in the same patient or in healthy bowels (5-10). In addition, mast
pathohistologically, and radiologically with respect to the cells in CD patients contain much more TNF-α than do those in
presence of certain pathophysiological inflammatory signs. Its UC patients or controls; mast cells secrete more histamine in IBD;
extent of disease is often quantified by endoscopy and the and the histamine content is significantly increased in jejunal
clinical disease activity documented by using scientific scores. perfusion fluid in CD patients (6-10). Since histamine has a very
According to more recent investigations, local and systemic short half-life of 1 – 2 min in blood, rapid and precise processing
inflammatory activity is essentially caused by various interleukins is needed to determine plasma levels of this compound. Several
(IL-1, IL-6, IL-12, IL-17, etc.), leptin and tumor necrosis factor studies have postulated an important role for mast cells and
alpha (TNF-α). TNF-α can be synthesized and secreted by many histamine in IBD (6-10); however, except for one treatment study
kinds of cells after cell activation via NFκB (e.g., monocytes, no systematic investigations using precise methods have been
neutrophilic granulocytes, and T lymphocytes) but mast cells and conducted to determine the magnitude of plasma histamine levels
basophilic granulocytes also are also capable of producing TNF-α in IBD (8, 13). Therefore, the goal of the present study was to
(1, 2, 4-6). Mast cells actually contain TNF-α in preformed investigate whether 1) there is a correlate for functional mast cell
vesicles and can secrete TNF-α together with histamine, one of its activity at the systemic plasma level and 2) whether by comparison
550

of histamine with the known proinflammatory cytokine TNF-α, plasma samples using ELISA to detect histamine were 3.2 and
any indications for a central pathogenetic relevance of mast cells 4.1%, respectively. For TNF-α this intra- and interassay
for IBD can be found in the plasma compartment. To this end, by variation was 7.1 and 13.6%, respectively.
using a cross-sectional study design, nonselected IBD patients
were investigated as regards inflammatory activity, anatomical Statistical tests
extent of inflammation, presence of EM, and medications.
For statistical testing, the software Graph Pad Instat 3 and
graphics program Graph Pad Prism Version 4.00, April 3, 2003,
MATERIAL AND METHODS were used. The t-test was employed to analyze normal
distribution for age of controls and patients in two independent
Patients samples. To test for significance in comparing the median values
in the various patient groups nonparametric U-tests for
This study was approved by the local ethics committee (No. independent samples (Mann-Whitney test) were applied. A test
925) and all patients gave their informed consent for the result with an error probability of P < 0.05 was considered
mediator diagnostic studies. We included a total of 91 out- and statistically significant. Pearson correlation analyses for
in-patients and IBD inpatients who had already received a independent samples were used to calculate the correlation
diagnosis of CD (n = 69) or UC (n = 22) according to between histamine and TNF-α in the plasma and for activity
serological, clinical, endoscopic, radiological, and histological indices with histamine and TNF-α. Here, the correlation
examinations before entering the study (Table 1). The study coefficient r and coefficient of determination r2 were calculated.
patients were referred to the University Erlangen, Department of
Medicine 1, for follow-up examinations, during acute episodes,
to exclude neoplasias, owing to new complaints, or to adjust RESULTS
their medication. Patients in the IBD group were classified
according to their current medication (no treatment, 5- The age of patients with CD (inactive, active, according to
aminosalicylic acid, steroids, or immune suppressants), disease the Crohn’s Disease Activity Index < 150 points or > 150
activity (CDAI, CAI), and for the presence of EM or fistula. points), UC (inactive, active, according to Rachmilewitz Colitis
The control group comprised 13 healthy individuals and Activity Index < 4 points or > 4 points), and controls showed a
clinic staff who were not on any medication and who were not normal distribution and at P > 0.05 no statistically significant
suffering from IBD, allergy, or an infection (Table 1). differences were found between the individual groups (Table 1).
Exclusion criteria for the study were drugs that could induce However, for both histamine and TNF-α, the mediator
a nonspecific release of histamine such as antibiotics, NSARs, plasma levels mostly did not show a normal distribution in the
opioids, muscle relaxants, or blood products. Furthermore, disease groups as compared to controls. Therefore, we worked
pregnant women (false-negative plasma histamine with mean values and 25 – 75% percentiles.
concentrations (10)), patients on anticoagulants, and those being
treated with anti-TNF-α antagonists were excluded from the Plasma histamine in Crohn’s disease and ulcerative colitis
present study.
An overview of the median plasma concentrations of
Determining mediators of histamine and tumor necrosis factor- histamine (ng histamine/ml × m2 body surface area) can be
α in plasma found in Table 2. By standardizing the plasma levels of
histamine to ng histamine/ml × m2 body surface area, a normal
Testing for plasma mediators in the blood was performed in range of 0.14 – 0.33 for histamine was calculated in the control
fasting patients between 7 and 9 a.m. All subjects had eaten group with a variance of approx. 47.7%.
normally the previous day (no diet) but had fasted from 6 p.m. Compared to the median of the controls (0.25) no significant
onwards. Fasting conditions for drawing blood were particularly changes in plasma levels of histamine were found for CD (0.19)
important because certain foods (cheese, wine, sauerkraut, etc.) and UC (0.23) even after differentiating between inactive and
contain histamine, and exogenous histamine can produce false- active disease state. However, the plasma histamine levels varied
positive results for histamine levels in plasma (6-8, 12). to a considerably greater degree in all disease groups than in
The EDTA blood (S-MonovetteR, Sarstedt, Numbrecht, healthy controls (variance 63.6 – 164.7%, Table 2).
Germany) was cooled on ice (4°C) immediately after being drawn By using our strict, standardized methods to calculate the
from the vein in order to avoid any possible histamine degradation normal range of laboratory values for plasma histamine from the
(12, 13). Then, the blood samples were centrifuged immediately for mean value in the control group and two standard deviations
10 min at 4000 U/min and cooled (4°C) so that the plasma could be (mean + 2 S.D., 0.24 + 2 × 0.11 = 0.46), we found a relatively
separated from the erythrocytes. The plasma samples were frozen at narrow normal range at P ≤ 0.46.
–20°C in separate batches for histamine and TNF-α testing until the The distribution for individual histamine concentrations in
ELISA assay could be performed to quantiatively determine levels the normal controls showed that none of the 13 healthy
of those compounds (12-14). To avoid false-positive results for individuals (0%) had values above the normal range of 0.46. For
histamine, we took care that no hemolytic, icteric, or lipemic CD, 11 of 68 patients (16.1%) showed values above the
samples were frozen: when certain concentrations are exceeded calculated normal range of 0.46 (inactive CD, 6/31 patients
(hemoglobin 5 mg/mL, bilirubin 1 mg/mL, triglycerides 30 19.3%; active CD, 5/37 13.5%). Among the UC patients, 6 of 22
mg/mL), cross-reactions with methylhistamine and acetylhistamine subjects (27.2%) showed values higher than 0.46 (inactive UC
can occur during the ELISA assay (13, 14). 6/12 patients 50%; active UC 0/10 0.0%).
Plasma histamine and TNF-α levels were each determined
photometrically by carrying out special ELISA tests according to Effect of inflammatory bowel disease medications on plasma
the manufacturer’s instructions and controlled according to levels of histamine
standard samples and curves (12-14). The test sensitivity for
histamine is 0.05 ng/ml and that for TNF 2.3 pg/ml (13, 14). Analysis of plasma histamine levels in IBD patients with
Coefficients of intra- and interassay variation for more than 20 respect to medication showed that patients not on treatment
 551

Table 1. Patients characteristics.

Sex Age
Disease Group n = pts. Range
f/m (median)
CG 13 7/6 38.0 19 – 68

CD 68 39/29 43.8 23 – 78
CD inactive 31 16/15 42.7 26 – 68
CD active 37 21/16 44.8 23 – 78

UC 22 11/11 38.3 19 – 60
UC inactive 12 7/5 37.6 19 – 51
UC active 10 5/5 38.8 22 – 60

n - number of patients (pts.), f - female, m - male, CG - control group, CD - Crohn’s disease, inactive CDAI < 150 points, active CDAI
>150 points, UC - ulcerative colitis, inactive CAI < 4 points, active CAI > 4 points.

Table 2. Plasma histamine levels (ng/ml × m2 body surface area) in inflammatory bowel disease and controls.

25% 75% coefficient

Histamine n = pts Median mean ± S.D.
percentile percentile variation %
CG 13 0.25 0.14 0.33 0.24 ± 0.11 47.7

CD 68 0.19 0.09 0.35 0.31 ± 0.5 146.3

CD inactive 31 0.15 0.08 0.35 0.30 ± 0.5 164.7
CD active 37 0.19 0.09 0.38 0.32 ± 0.4 133.0

UC 22 0.23 0.11 0.60 0.34 ± 0.3 93.4

UC inactive 12 0.43 0.22 0.82 0.50 ± 0.4 71.7
UC active 10 0.17 0.05 0.24 0.15 ± 0.1 63.6

Table 3. Plasma histamine levels (ng/ml × m2 body surface area) in inflammatory bowel disease in relation to medication.

Histamine immuno-
No medication 5-ASA steroids
Group suppressants
n = 10 n = 15 n = 31 n = 10
CD 0.31 0.121,2 0.23 0.17
(0.12 – 0.65) (0.07 – 0.19) (0.10 – 0.40) (0.07 – 0.37)
n=2 n=9 n=8 n=3
UC 0.72 0.23 0.18 0.25
(0.62 – 0.83) (0.09 – 0.29) (0.08 – 0.52) (0.22 – 0.83)
1 P = 0.04 versus controls and 2 P = 0.03 versus CD without medication; 5-ASA 5-aminosalicylic acid.

tended to have higher plasma levels of histamine (Table 3) than affected region of the bowel was precisely indicated, the median
those being treated with 5-aminosalicylic acid, steroids, or plasma histamine levels for proctitis, colitis, ileitic, ileocolitis,
immune suppressant drugs. Significantly lower levels of plasma or left-sided colitis and pancolitis were in the normal range
histamine as compared to the controls (P = 0.04) and to untreated under 0.46.
patients with CD were only observed, however, in patients with
CD who were taking 5-aminosalicylic acid. Effect of extraintestinal manifestations and fistula on plasma
The UC group tended to have higher plasma levels of levels of histamine
histamine; however, these values were not statistically
significantly different from those of controls or CD patients Similar, nonsignificant results were found for the plasma
(Table 3). concentrations of histamine in patients with CD and EM (0.20;
0.09 – 0.43, n = 27) or fistula (0.18; 0.08 – 0.38, n = 14). This
Effect of disease localization on plasma levels of histamine lack of significance also held true after differentiating between
active and inactive disease (P > 0.05 in all subgroups).
Disease localization of IBD did not have a statistically In the four patients with UC and EM, the plasma levels of
significant effect on the concentration of histamine in plasma histamine tended to be higher at 0.39 (0.18 – 0.70), but this was
(Table 4). Indeed, for both CD and UC patients in whom the not significantly different from the control group values.
552

Table 4. Plasma histamine levels (ng/ml × m2 body surface area) in inflammatory bowel disease in relation to disease location and
extent.
Histamine
Proctitis Colitis Ileitis Ileocolitis
Group
n=4 n = 17 n = 36 n=7
CD 0.19 0.20 0.14 0.16
(0.17 – 0.73) (0.13 – 0.28) (0.07 – 0.23) (0.08 – 0.38)
Left-sided colitis Subtotal or total colitis
n=4 n = 10
UC
0.40 0.21 – –
(0.13 – 0.72) (0.08 – 0.26)
Į
Table 5. Plasma TNF-α levels (ng/ml × m2 body surface area) in inflammatory bowel disease and controls.
TNF-Į 25% 75% coefficient
n = pts median mean±S.D.
Group percentile percentile variation %
CG 11 0.44 0.00 1.15 1.14 ± 1.9 166.1

CD1 29 4.62 1.82 9.22 5.91 ± 4.6 78.7

CD inactive2 13 4.09 1.20 10.13 5.99 ± 5.2 87.5
CD active3 16 4.86 2.67 8.31 5.84 ± 4.3 73.4

UC2 11 3.44 0.08 14.64 6.19 ± 5.6 127.4

UC inactive 4 3.03 0.06 7.91 3.71 ± 3.1 100.6
UC active 7 4.12 0.60 14.46 4.88 ± 4.9 113.3
Significance: 1 P = 0.005, 2 P = 0.01, 3 P = 0.008 versus control group.

Table 6. Plasma TNF-α levels (ng/ml × m2 body surface area) in inflammatory bowel disease in relation to medication.
TNF-α immuno-
No medication 5-ASA Steroids
Group suppressants
n=7 n = 10 n = 17 n=8
CD 4.091 8.222 6.262 8.362
(1.10 – 7.50) (2.28 – 13.54) (0.92 – 9.22) (5.68 – 11.40)
n=3 n=6 n=3
UC 1.67 4.01 4.78 –
(0.08 – 3.21) (0.96 – 14.64) (0.89 – 5.25)
1 P = 0.05 versus controls and 2 P < 0.007 versus controls; 5-ASA 5-aminosalicylic acid. There was no significant difference between
the groups with and without medication either in CD or UC.

Plasma levels of tumor necrosis factor-α in Crohn’s disease value in the control group and two standard deviations (mean + 2
and ulcerative colitis SD, 1.14 + 2 × 1.9 = 4.94), we found a normal range at P ≤ 4.94.
The distribution of the individual plasma levels of TNF-α in
Table 5 presents an overview of the median plasma levels of healthy controls shows that only in one of the 11 healthy
TNF-α (pg TNF-α/ml × m2 body surface area). Standardizing the individuals (9.0%) were values above the normal range of 4.94. For
TNF-α levels to pg /ml × m2 body surface area, the control group 14 of 29 CD patients, corresponding to 48%, values were above the
showed a normal range of TNF-α from 0.0 – 1.15, with a wide calculated normal range of 4.94 for plasma TNF-α (inactive CD
variance in these values of approx. 166.1%. In comparison to the 6/13 patients, 46.1%; active CD 8/16, 50.0%). For 5 of 11 UC
controls (0.44), TNF-α concentrations were significantly higher patients (45.4%), values were above the normal range of 4.94
at, respectively, 10-fold and eightfold in the CD (4.62) and UC (inactive UC 2/4 patients, 50.0%; active UC 3/7 patients, 42.8%).
(3.44) patients, who also showed a lower variance than did the
control group (78.7%). After differentiating between inactive and Effect of inflammatory bowel disease medication on plasma
active disease states, the difference between the CD patients and levels of tumor necrosis factor-α
controls was still statistically significant. In the UC patients, no
significant changes could be observed owing to the low number With respect to medication, plasma TNF-α levels in CD
of samples; however, patients with active disease tended to have patients were lower in patients not on treatment (Table 6) than in
higher levels of TNF-α, similarly to the CD patients. those being treated with 5-aminosalicylic acid, steroids, or
By using our strict, standardized methods to calculate the immunosuppressants. Indeed, plasma TNF-α levels were
normal range of laboratory values for TNF-α from the mean significantly higher in all treated groups than in the controls (P <
 553
Į
Table 7. Plasma TNF-α levels (ng/ml × m2 body surface area) in inflammatory bowel disease in relation to disease location and extent.

TNF-Į
Proctitis Colitis Ileitis Ileocolitis
Group
n=3 n=7 n=4 n = 12
CD1 7.59 3.96 4.74 5.31
(4.50 – 12.90) (1.21 – 9.03) (1.22 – 8.0) (2.88 – 10.33)
Left-sided colitis Subtotal or total colitis1
n=4 n=8
UC
2.98 4.04
n.a. n.a.
(0.08 – 9.44) (0.89 – 14.64)
1 P < 0.03 compared with control group.

Table 8. Plasma TNF-α levels (ng/ml × m2 bodyĮ surface area) in inflammatory bowel disease in relation to extraintestinal
manifestations and fistula.

TNF-α
Without EM With EM Fistula
Group
n = 17 n =11 n=7
CD1 4.46 6.26 6.26
(0.86 – 9.22) (2.30 – 13.06) (1.27 – 13.00)
n = 91 n=3
UC 4.08 4.78
(0.89 – 12.6) (0.42 – 8.89)
1 P < 0.03 compared with control group.

Table 9. Correlation of plasma mediators and Crohn’s disease activity (CDAI) or colitis activity index (CAI), respectively (ng/ml × m2
body surface area or pg/ml × m2 body surface area).

MC CU MC UC
histamine and CDAI histamine and CAI TNF and CDAI TNF and CAI
N 68 22 29 11
Pearson r –0.06967 –0.47900 –0.00065 –0.47120
95% confidence
–0.30310 – 0.17170 –0.74930 – 0.07180 –0.36720 – 0.36600 –0.95610 – 0.70370
Interval
r2 0.00485 0.22940 0.03016 0.22200
P > 0.05 0.02410 > 0.05 > 0.05

0.007), but not in the group of CD patients not on medication. Effect of extraintestinal manifestations and fistulas on plasma
Comparison of the nontreated CD group and the treated subgroups levels of tumor necrosis factor-α
did not show any statistically significant differences, however.
A similar result was observed for plasma TNF-α levels in the Plasma TNF-α levels were again significantly higher in CD
UC group, which, however, did not reach statistical significance patients both with (P = 0.005) and without (P = 0.01) EM and
owing to the low number of patients. Nontreated patients also in patients with fistulas than in the controls (P = 0.01, Table 8).
presented lower TNF concentrations than those on treatment Between the patient groups presenting with and without EM,
with 5-aminosalicylic acid or steroids. however, no statistically significant differences were observed
in CD.
Effect of disease localization on plasma tumor necrosis factor- In the three patients with UC and EM, a mean plasma TNF-α
α levels level of 4.78 was found, which, however, did not differ from that
of controls in a statistically significant manner.
The localization of chronic inflammation in the bowel did
not have any significant effect on plasma TNF-α levels in CD Correlation between plasma levels of histamine and tumor
patients (Table 7) as the concentrations of TNF-α did not necrosis factor-α with disease activity in Crohn’s disease and
increase statistically significantly with an increase in affected ulcerative colitis patients
areas in the bowel in these patients. For the individual affected
areas in CD patients, consistently higher values than in controls No correlation was found between plasma histamine levels
were observed again, however (P < 0.03). and disease activity in CD patients. For UC, a low-grade
For UC a statistically significant increase in plasma TNF-α negative correlation with significance was established (P = 0.02,
levels was found for both subtotal and total colitis (P < 0.03). Table 9).
554

For TNF-α, a significant correlation with CDAI could not be present with systemically elevated plasma levels of histamine that
found for CD patients either. For UC, a low-grade correlation are well outside of the normal range represent people with
with CAI was observed, which, however, did not reach allergies, atopic disease, or with a reduced capacity to
significance (Table 9). enzymatically degrade histamine and are among those patients
Furthermore, neither of the mediators investigated, plasma with CD or UC, further investigations need to be conducted (6-8,
levels of histamine and TNF-α, correlated significantly with each 10, 17-22). Indeed, higher plasma levels of histamine tended to be
other for CD (r2 = 0.01, P > 0.05) or for UC (r2 = 0.14, P > 0.05). found among the UC patients. This is in line with literature reports
of a higher frequency of atopic diseases, denser eosinophilic
infiltration in the bowel tissue, Th2 immune regulation, and higher
DISCUSSION serum levels of IgE in UC than in CD patients (7, 8, 23).
Numerous subgroup analyses could not establish any
The role of mast cells in IBD has still not been clearly difference in plasma histamine levels with respect to disease
elucidated (5, 11). Since about 1975 reports have been appearing localization, extent of inflammation, presence of EM or fistula;
about increased degranulation of increasing numbers of mast with respect to medications; however, plasma histamine levels
cells in IBD tissue. Recently even mast cells loaded with TNF-α tended to be higher in untreated IBD patients than in those being
IL-16-positive mast cells, substance P-positive mast cells, and treated. Indeed, 5-aminosalicylic acid significantly lowered
potent protease-secreting mast cells have been discovered in plasma histamine levels (Table 3) whereas steroids and
bowel tissue from IBD patients by immunohistochemistry (8-11, immunosuppressants did not help to lower these levels
16-19). Although these findings speak for a local contribution of significantly in either CD or UC patients. The inhibitory effect of
mast cells to the inflammatory process with the secretion of 5-aminosalicylic acid on mast cells and basophils was previously
histamine, tryptase, cytokines, and other mediators, no experimentally demonstrated (24). It remains unknown, however,
systematic studies have been conducted to measure the plasma why 5-aminosalicylic acid treatment only reaches significance in
levels of histamine in the blood of patients with CD or UC. CD and not UC patients. Possibly the aforementioned,
By using a precise technique with a rapid, cooled processing tendentially higher plasma histamine levels in UC and the
of samples so as to avoid histamine degradation to determine predominating immunological mechanisms of increased Th2
plasma levels of histamine and standardizing these values with immune regulation and stronger mast cell activation in UC
respect to body surface area (ng histamine/ml × m2 body surface patients may play a decisive role (7, 9, 10, 17, 25).
area), we found a relatively narrow normal range for plasma It was previously demonstrated that, in CD, disease activity
histamine at below 0.46 in a healthy control group. The median correlates significantly with urine excretion of methylhistamine
histamine levels in the patient groups of CD and UC did not differ in contrast to the findings here dealing with plasma histamine
from those of the controls. Thus, for our study patients with levels (6); therefore, this may indicate, at least for CD, that the
chronic IBD the plasma levels of histamine were not increased at histamine released locally in the bowel is rapidly methylated in
a systemic level, although contrasting findings of increased mast bowel tissue, endothelium and in the liver and thus that high
cell counts, degranulation, or high concentrations of mediators systemic plasma histamine levels cannot be expected (6, 12, 21,
locally in the bowel have been reported elsewhere (8-10, 16-19). 22). Therefore, the present results with plasma histamine can
Considering individual plasma levels of histamine, too, only about only be considered in terms of their potential systemic effect.
16% of CD patients and about 27% of UC patients presented with Indeed, the local rate of intestinal histamine release, the tissue
systemically elevated plasma histamine levels that were outside metabolic rate, and the proinflammatory effect of histamine as a
the standard normal range (Fig. 1). Thus, systemic effects of local tissue hormone cannot be estimated by plasma histamine
histamine are only to be expected in a small proportion of patients determinations, owing to the different compartments of bowel
with IBD. In order to determine whether the few individuals who and blood plasma (6-8, 19,20).

% patients with plasma mediator Plasma histamine

levels > mean + 2SD Plasma TNFalpha
100

75

50
48.2%
40.0%

27.2%
25
16.1%

Fig. 1. Frequency of IBD patients

0 with significantly increased plasma
Crohn‘s Disease ulcerative colitis mediator levels above the normal
range.
 555

In direct comparison with histamine, this study the different compartments involved, bowel-blood plasma, these
demonstrated a much greater systemic significance for TNF-α investigations at the blood plasma level could not demonstrate a
than for histamine in IBD. In contrast to the histamine findings, predominat role of mast cells or of histamine for IBD (1, 2, 5,
a significantly increased, systemically effective plasma TNF-α 32). The previous reports of heterogeneous treatment effects of
concentration was found for both CD and UC (Table 5). Previous antihistamines, mast cell stabilizers, or anti-allergenic treatments
studies also reported an increased production of TNF-α in bowel could perhaps be explained by the low rate of IBD patients
tissue and subsequent detection of increased TNF-α levels in presenting with plasma histamine levels outside of the normal
stool and also systemically in serum. The greater systemic role range (16 – 27%) or an associated atopic or allergic disease (5-
of TNF-α is also highlighted by the association of IBD with 8, 13, 23). However, histamine and TNF-α seem to be secreted
certain polymorphisms of the TNF gene and the known independently of each other in IBD, which means that varying
correlation between soluble TNF receptors in urine and disease types of active immune cells are involved in the expression of
activity (1, 11, 26-30). In agreement with numerous other inflammation, with a clearer, but not exclusive predominance of
studies, which have reported a better therapeutic response to TNF-α (1, 11, 34-38). Since only about 50% of all patients with
anti-TNF antibodies in CD than in UC, we found higher TNF-α chronic bowel inflammation demonstrate systemically increased
levels and considerably more pronounced, statistically plasma levels of TNF-α, future long-term clinical studies should
significant differences in the CD group than in the UC group. determine in which patients expressed TNF-α mechanisms
This applies both to the CD group as a whole as well as to active predominately contribute to disease manifestations, which
and inactive disease states of CD 1, 29-32), for patients with and measures actually can normalize TNF-α levels, and most of all,
without EM, and fistulas. what other cytokine mechanisms exist in those individuals who
In spite of the significantly increased median plasma TNF-α do not show increased TNF-α production either locally or
concentrations in the IBD cohort, the individual TNF-α levels systemically (11, 30-32, 37, 38).
were only far outside the normal range for healthy individuals in
about 48% of CD and 40% of UC patients. In direct comparison Conflict of interests: None declared.
with histamine, the clearly more important role of TNF-α emerges
here, too (Fig. 1). These data also show that inflammation in IBD
does not develop uniformly via the proinflammatory cytokine REFERENCES
TNF in all patients, but rather that additional heterogeneous
mechanisms exist that induce and promote the inflammation (e.g., 1. Henderson P, van Limbergen JE, Schwarze J, Wilson DC.
IL-6, IL-17, IL-18, IL-23, etc.) (1, 2, 5, 11, 33). Function of intestinal epithelium and its dysregulation in
No significant differences in TNF-α plasma concentrations inflammatory bowel disease. Inflamm Bowel Dis 2011; 17:
were observed either with respect to the different disease 382-395.
localizations or the extent of inflammation in CD patients. 2. Karlinger K, Gyorke T, Mako E, Mester A, Tarjan Z. The
However, for TNF-α a significant contrast to histamine was epidemiology and the pathogenesis of inklammatory bowel
found concerning medication. Indeed, IBD patients currently not disease. Eur J Radiol 2000; 35: 154-167.
on treatment had the lowest TNF-α levels and the treated patient 3. Bilski J, Mazur-Bialy AI, Wierdak M, Brzozowski T. The
groups presented with significantly higher levels. Neither for impact of physical activity and nutrition on inflammatory
patients on 5-aminosalicylic acid nor for those on steroids or bowel disease: the potential role of cross talk between
immunosuppressants could a reduction in plasma TNF-α levels adipose tissue and skeletal muscle. J Physiol Pharmacol
to the normal range be found (Table 6). However, this was a 2013; 64: 143-55.
cross-sectional study of IBD patients in various stages of disease, 4. Biesiada G, Czepiel J, Ptak-Belowska A, et al. Expression
and thus we cannot clarify whether patients with active disease and release of leptin and proinflammatory cytokines in
had possibly had even higher TNF-α levels before treatment was patients with ulcerative colitis and infectious diarrhea.
initiated than those measured while they were receiving J Physiol Pharmacol 2012; 63: 471-481.
medication. On the other hand, the TNF-α results with respect to 5. Gurish M, Castells MC. Mast Cells. UpToDate ®, Version
medication could also be interpreted to mean that IBD patients 16.1, 2008, 1-33.
who were not receiving any medication had gone into remission 6. Winterkamp S, Weidenhiller M, Otte P, et al. Urinary
and could stop taking the medication because the pronounced excretion of N-methylhistamine as a marker of disease
inflammatory activity and hence TNF-α production had dropped activity in inflammatory bowel disease. Am J Gastroenterol
so decidedly. Similar to the results of other investigations, we did 2002; 97: 3071-3077
not find a correlation between plasma TNF-α levels and disease 7. Raithel M, Winterkamp S, Weidenhiller M, Muller S, Hahne
activity either for CD or UC (30, 33, 34). EG. Combination therapy using fexofenadine, disodium
This may in part be due to the limitations of the present cromoglycate, and a hypoallergenic amino acid-based
study: heterogeneous patients in various stages of disease and formula induced remission in a patient with steroid-
receiving different medications were analyzed using a cross- dependent, chronically active ulcerative colitis. Int J
sectional study design. In addition, systemically effective Colorectal Dis 2007; 22: 833-839.
mediator levels and the disease activity indices used measure 8. Xie H, He SH. Roles of histamine and its receptors in
different aspects of chronic bowel inflammation, different allergic and inflammatory bowel diseases. World J
compartments were compared with each other, and the Gastroenterol 2005; 11: 2851-2857.
significance of the various types of immune cells secreting TNF- 9. Raithel M, Winterkamp S, Pacurar A, Ulrich P, Hochberger
α may be different for CD and UC (fibroblasts, macrophages, J, Hahn EG. Release of mast cell tryptase from human
epithelial cells, paneth cells, mast cells, neutrophils, and colorectal mucosa in inflammatory bowel disease. Scand J
eosinophilic granulocytes). Furthermore, numerous other Gastroenterol 2001; 36: 174-179.
immune mechanisms may need to be taken into consideration (1, 10. He SH. Key role of mast cells and major secretory products
2, 5, 7, 26, 30, 32-35). in inflammatory bowel disease. World J Gastroenterol 2004;
In our study, we did not find a correlation between plasma 10: 309-318.
levels of histamine and TNF-α; nor were quantitatively similarly 11. Fiocchi C. Inflammatory bowel disease: etiology and
pronounced changes in IBD patients observed. Bearing in mind pathogenesis. Gastroenterology 1998; 115: 182-205.
556

12. Keyzer Y, Breukelman H, Wolthers B, van den Heuvel M, 25. Szkudlapski D, Labuzek K, Pokora Z, et al. The emering
Kromme N, Berg WC. Urinary excretion of histamine and role of helminths in treatment of the inflammatory bowel
some of ist metabolites in man: influence of the diet. Agents disorders. J Physiol Pharmacol 2014; 65: 741-751.
Actions 1984; 15: 189-194. 26. Beil WJ, Weller PF, Peppercorn MA, Galli SJ, Dvorak AM..
13. Raithel M, Nagel A, Zopf Y, et al. Plasma histamine levels Ultrastructural immunogold localization of subcellular sites
(H) during adjunctive H1-receptor antagonist treatment with of TNF-α in colonic Crohn´s disease. J Leukoc Biol 1995;
loratadine in patients with active inflammatory bowel 58: 284-298.
disease (IBD). Inflamm Res 2010; 59 (Suppl. 2): S257-S258. 27. Murch SH, Braegger CP, Walker-Smith JA, MacDonald TT.
14. TNF-α ELISA, Instruction for Use. Hamburg, IBL Locations of tumour necrosis factor-α by
Gesellschaft fur Immunchemie und Immunbiologie MBH immunohistochemistry in chronic inflammatory bowel
2006, pp. 1-33. disease. Gut 1993; 34: 1705-1709.
15. Histamine ELISA (RE 59221), Hamburg, IBL Gesellschaft 28. Bouma G, Xia B, Crusius JBA, et al. Distribution of four
fur Immunchemie und Immunbiologie MBH (Version 2) polymorphism in the tumor necrosis factor (TNF) genes in
2006, pp. S1-S7. patients with inflammatory bowel disease (IBD). Clin Exp
16. Fox CC, Lichtenstein LM, Roche JK. Intestinal mast cell Immunol 1996; 103: 391-396.
responses in idiopathic inflammatory bowel disease. 29. Van Deventer SJ. Tumor necrosis factor and Crohn´s
Histamine release from human intestinal mast cells in disease. Gut 1997; 40: 443-448.
response to gut epithelial proteins. Dig Dis Sci 1993; 38: 30. MacDonald TT, Hutchings P, Choy MY, Murch S, Cooke A.
1105-1112. Tumor necrosis factor-alpha and interferon-gamma production
17. Fox CC, Lazenby AJ, Moore WC, Yardley JH, Bayless TM, measured at the single cell level in normal and inflamed human
Lichtenstein LM. Enhancement of human intestinal mast intestine. Clin Exp Immunol 1990; 81: 301-305.
cell mediator release in active ulcerative colitis. 31. Sands BE, Tremaine WJ, Sandborn WJ, et al. Infliximab in
Gastroenterology 1990; 99: 119-124. the treatment of severe, steroid-refractory ulcerative colitis:
18. Sanderson IR, Leung KBP, Pearce FL, Walker-Smith JA. a pilot study. Inflamm Bowel Dis 2001; 7: 83-88.
Lamina propria mast cells in biopsies from children with 32. Travis SP, Stange EF, Lemann M, et al. European evidence-
Crohn’s disease. J Clin Pathol 1986; 39: 279-283. based Consensus on the management of ulcerative colitis:
19. Raithel M, Horauf AM, Matek M, Baenkler HW. Kinetics of current management. J Crohns Colitis 2008; 2: 24-42.
histamine released from rectal mucosa. Agents Actions 1989; 33. Danese S, Fiocchi C. Etiopathogenesis of inflammatory
28: 164-167. bowel diseases. World J Gastroenterol 2006; 12: 4807-4812.
20. Bachert C. The role of histamine in allergic disease: re- 34. Murch SH, Lamkin VA, Savage MO, Walker-Smith JA,
appraisal of its inflammatory potential. Allergy 2002; 57: MacDonald TT. Serum concentrations of tumor necrosis
287-296. factor α in childhood chronic inflammatory bowel disease.
21. Schmidt WU, Sattler J, Hesterberg R, et al. Human intestinal Gut 1991; 32: 913-917.
diamine oxidase (DAO) activity in Crohn’s disease: A new 35. Westendorf AM, Templin M, Geffers R, et al. CD4+ T cell
marker for disease assessment? Agents Actions 1990; 30: mediated intestinal immunity: chronic inflammation versus
267-270. immune regulation. Gut 2005; 54: 60-69.
22. Petersen J, Raithel M, Schwelberger HG. Analysis of 36. Pierik M, Rutgeerts P, Vlietinck R, Vermeire S.
diamine oxidase gene polymorphisms in patients with Pharmacogenetics in inflammatory bowel disease. World J
inflammatory bowel disease. Inflamm Res 2001; 50: (Suppl. Gastroenterol 2006; 12: 3657-3667.
2): S68-S69. 37. Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab for
23. Bartunkova J, Kolarova I, Sediva A, Holzelova E. induction and maintenance therapy for ulcerative colitis.
Antineutrophil cytoplasmatic antibodies, anti- N Engl J Med 2005; 353: 2462-2476.
Saccharomyces cerevisiae antibodies, and specific IgE to
food allergens in children with inflammatory bowel disease. R e c e i v e d : December 16, 2014
Clin Immunol 2002; 102: 162-168. A c c e p t e d : March 23, 2015
24. Fox CC, Moore WC, Lichtenstein LM. Modulation of
mediator release from human intestinal mast cells by Author’s address: Dr. Alexander Hagel, Department of
sulfasalazine and 5-aminosalicylic acid. Dig Dis Sci 1991; Medicine I, Gastroenterology, University of Erlangen-
36: 179-184. Nuremberg, 18 Ulmenweg Street, 91054 Erlangen, Germany.
E-mail: alexander.hagel@uk-erlangen.de