Discoveries in Pharmacology, Volume 1: Psycho- and Neuro-pharmacology — M. J. Parnham and J. Bruinvels, eds.

© 1983, Elsevier Science Publishers B.V.

CHAPTER 4, PART 2

The discovery of the butyrophenone-type

neuroleptics

P. A. J. Janssen and J. P. Tollenaere

CONTENTS

1 TRTR TR SRt o4 e 50 R o s i (o BT 6 o7 £ oM L Y. O KNI o8 181

S U e R LT T b o R S A el e B R o RO N N 183

3. A compound with mixed morphinomimetic and neuroleptic properties .............. 184

A owardsshaloperidol '« /oot .. o 0L T4 s e SN e e i e e e 185

S ANEECRB Al Stenh). . L ile e bk s e s L e S L L e L i 188

L hRRRIODEHEQISTRWATUSIEY . . 00 o5 T L SR L vt o ey o ol oot 189

B ee B O O | N e T iovbs ogs 3 e & e 3 5 195

S T R S 0 R A o TR O L ke R e PG, | PR 195

1. INTRODUCTION

The modern pharmacotherapy of psychosis started in 1952 with the introduction

of chlorpromazine (Delay et al., 1952) into clinical practice. Chlorpromazine
(Fig. 1), a phenothiazine derivative came about from the further development
of the tricyclic antihistaminics. Shortly thereafter, reserpine (Fig. 2) an alkaloid
extracted from the Indian shrub Rauwolfia serpentina, introduced in the six-
teenth century in Europe by Leonhart Rauwolf, was used by Weber (1954) for
the treatment of psychoses. In passing it may be noted that a decoction of

181
182

s 3
/N’ CNfCHZ_CHZ- N S
CHy
cl

CcL

FIGURE 1. Perspective drawing of chlorpromazine (Tollenaere et al., 1979).

FIGURE 2. Perspective drawing of reserpine (Tollenaere et al., 1979).

Rauwolfia leaves known as ‘Pagal-Ka-Dawa’ (herb against madness) was

already in use in ancient India for the treatment of the mentally ill.
These two structurally different types of neuroleptics served as lead com-
pounds for the synthesis and subsequent pharmacological and clinical testing of
many synthetic derivatives. Except for a few, most derivatives showed a phar-
macological profile similar to the original lead compounds.
Even today one is struck by the vast structural difference between chlor-
promazine and reserpine. In 1953, when we started our research with a small
group of people and an equally small budget, it was no wonder that our efforts
were directed towards making new chemicals which could be synthesized and
purified with simple methods and equipment and which could be phar-
macologically tested at minimal expense.
 183

2. THE EARLY BEGINNINGS

With these odds against success we synthesized and tested a fairly extensive
series of diphenylpropylamines and derivatives of 4-phenylpiperidine. We were
rapidly successful with our first important drug, the anticholinergic compound
isopropamide or R 79 (each compound was referred to as ‘Research’ or R-
compound). The royalty income of R 79 increased our financial resources and
enabled us to carry out more research. One of our next hits in the series of
diphenylpropylamines was the powerful synthetic analgesic dextromoramide or
R 875 (Fig. 3). Concurrently with the diphenylpropylamine series we conducted
a thorough study of the 4-phenylpiperidines related to meperidine or pethidine
(Fig. 3) with the objective of increasing its morphinomimetic potency by replac-
ing the N-methyl group of meperidine by other simple chemical moieties. This
turned out to be rather easy: a Mannich reaction of normeperidine and
acetophenone yielded the propiophenone derivative R 951 (Fig. 3) which was
about a 100-times more potent than pethidine (Janssen et al., 1958, 1959b, c).
Our simultaneous study of diphenylpropylamine (Janssen, 1960) and
4-phenylpiperidine derivatives proved to be a lucky and rewarding choice. In
fact, in the course of previous studies it was noted that the correlation between
analgesic and constipating activity of several hundreds of analgesically active

IC-()-CMZ-CH3
t
CH;—N

FIGURE 3. (A) Perspective drawing of dextromoramide (Tollenaere et al., 1979). (B) Perspective
drawing of meperidine (Tollenaere et al., 1979). (C) Structural formula of R 951. (D) Perspective
drawing of diphenoxylate based on X-ray crystallography data and molecular model building.
184

compounds is rather poor. Some very active analgesics as for instance dex-
tromoramide, produce significant inhibition of the rate of gastrointestinal pro-
pulsion at high doses. Codeine on the other hand, is known to produce constipa-
tion in man at relatively low doses. It, therefore, seemed reasonable to assume
that analgesic type compounds devoid of analgesic activity but having con-
stipating activity might be synthesized. Shortly after the synthesis of dex-
tromoramide in the middle of 1956, the antidiarrhoeal R 1132 or diphenoxylate
(Fig. 3) chemically related to pethidine and the nitrile derivatives of R 79, was
synthesized (Janssen et al., 1959a).

3. A COMPOUND WITH MIXED MORPHINOMIMETIC AND NEUROLEPTIC PROPERTIES

Compound R 951 proved to be a pivotal element in our thinking because the

high potency of R 951 was contrary to the firm belief at that time that only small
chemical groups or moieties at the nitrogen atom of the piperidine ring were
compatible with high morphinomimetic potency (Beckett, 1952; Braenden et al.,
1955). This quite unexpected observation fueled our efforts to examine
systematically the effect of structural modification on the potency. One of the
most challenging and obvious structural modifications was of course, to make
the N-substituent even larger, among others by lengthening the chain between
the aromatic moiety and the piperidine ring. This time, however, we pushed our
luck too far as the morphinomimetic potency of R 1187 (Fig. 4), for example,
dropped quite considerably. However, when this new chemical was injected in
mice, we noted that the usual morphinomimetic effects (morphine-like excite-
ment, mydriasis and insensitivity to painful stimuli) were followed by chlor-
promazine-like effects in the sense that the animals became progressively calm,
sedated, and slightly catatonic.
As already mentioned, in the early days of our laboratory, in the middle fif-
ties, only rather simple pharmacological tests were in use for the detection of
analgesic and anticholinergic activity. One of them consisted of a procedure of
measurement of the hot-plate activity (AD,) (Janssen and Eddy, 1960),
mydriasis (MD,) (Jageneau and Janssen, 1956) and gross behavioural effects

fi
1C-0-CH-CH,
©0—0H2—CH2-CHZ—N

R1187

FIGURE 4. Structural formula of R 1187.

 185

in mice. This simple screening procedure easily distinguishes between mor-

phinomimetic and chlorpromazine-like compounds. For quite some time it was
empirically known that the AD,/MD,, ratio was not significantly different
from unity for all the classical morphinomimetics known at that time whereas
this ratio was significantly less than unity for chlorpromazine, its congeners and
reserpine (Table I).
Thus, although the morphinomimetic activity of R 1187 was markedly reduc-
ed, we observed a concomitant decrease of the mydriatic activity. The ratio
AD,/MD,; turned out to be 0.18, a value significantly less than unity and
highly unusual for a morphinomimetic.

TABLE I

MOLECULAR MODIFICATIONS LEADING TO THE SYNTHESIS OF THE

BUTYROPHENONES AND HALOPERIDOL (R 1625)

o
I
x,—@—c—tcaz),fi»‘ A
ki

R-compound X, n R X, AD,* MD** AD./MD,

R 951 H 2 COOCH,CH, H 11 1.9 0.6

R 1187 H 3 COOCH,CH; H 6.7 38 0.18
R 1246 H 2 COOCH, H 12.8 >280 <0.4
R 1338 H 3 COOCH;, H 6.2 139 0.04
R 1472 H 3 OH H 6.1 >100 <0.06
R 1589 H 3 OH H 2.9 >100 <0.03
R 1600 H 3 OH F 10 > 58 <0.17
R 1625 ¥ 3 OH Cl 1.4 >100 <0.01

Chlorpromazine 4.8 >100 <0.05

Reserpine 4.1 >100 <0.04
Pethidine 81 81 1
Morphine 30 40 0.75

* Median effective dose (s.c.) in hot-plate test in pmol kg“.

** Median effective dose (s.c.) mydriatic activity in gmol kg™

4. TOWARDS HALOPERIDOL

This was a great day in our laboratory since in 1957, only reserpine, chlor-
promazine and its congeners were known to behave in a qualitatively similar
fashion in our pharmacological screening procedure. From the structural point
186

of view, R 1187, being structurally unrelated to chlorpromazine and reserpine,

came as a great surprise and enticed us to dig deeper into this fascinating and
potentially rewarding field, especially since R 1187 was deemed to be of no prac-
tical value due to its mixed morphinomimetic and chlorpromazine-like proper-
ties. Compound R 1187, synthesized in January 1957, only proved to be a
milestone in our odyssey towards a potent useful neuroleptic. In retrospect, the
fact that we recognized right from the start the unusual biological characteristics
of R 1187 were actually far more important than the synthesis of R 1187 itself.
It was, therefore, immediately decided to synthesize related molecules. Com-
pound R 1187 being the lead compound, embedded in the general structure
represented in Fig. 5, could be structurally modified, resulting in a prohibitively
large number of molecules. Given our limited manpower and financial
resources, our only assets to tackle this problem were our faith and confidence
in the ultimate success of the laboratory and our willingness to work hard long
hours.
An analysis of the laboratory records of that period shows that of the 438
compounds we synthesized between R 1187 and R 1625 (haloperidol), 304 com-
pounds contained moiety 1 of Table II and that almost one third of the com-
pounds we synthetized contained moiety 3. Remember that we were attempting
to increase the morphine-like potency of pethidine!
Thus, in total, moiety 1 with R = COO alkyl occurred 158 times, or in other
words in about 52 per cent of all 304 4-phenylpiperidine derivatives synthesized
between R 1187 and R 1625. It is interesting to note that moiety 6 occurred only
69 times. When it did appear around R 1300, it was combined with a piperidine
nitrogen substituent which was not compatible with chlorpromazine-like activi-

! R
@Y-(CHZ’nTN
Xy !
|
|

"
|
Xy= H,F, CL,Br,Me,OMe X2+ H,F, Cl,Br, Me,OMe
¥=-CH=CH-, - CH(OH), -C=0 0
I
R:-C-0-(CH,)-CH; n=01,23
1]
=-0-C-(CHply-CHy n=01,2,3
=-OH
FIGURE 5. General structure showing the various synthetic possibilities: R 1187 (see Fig. 4) can
o
]
be seen when Xl =H;Y=C=O0O;R = 7C70—CH:CH3;n =3X,=H
 187

TABLE II

FREQUENCY OF OCCURRENCE (N) OF MOIETIES RELATED TO 4-PHENYLPIPERIDINE

BETWEEN R 1187 AND R 1625 (HALOPERIDOL)

Moiety N

R
—N

1 304
XE

1C—O0—CH,
=N 2 11

TC—0—CH,—CH,
i 3 9%

iC—O—(CH,), —CH,
it n=3,4.. 4 51

A2

i
0—C—(CH,), —CH,
=K n=01.2. 5 34

OH

) 6 69

9
(Oey n X
X n=23. 7 8
188

ty. Thus, mere coincidence prevented us from recognizing that 4-OH-4-phenyl-

piperidine was the moiety we eventually needed. Actually, our ultimate success
was delayed by merely half a year or so as only 13 months elapsed between the
synthesis of R 1187 and R 1625.
Today one can only guess as to what guided us to pick up that particular com-
bination of molecular fragments out of many thousands of possible combina-
tions we could have synthesized. A certain insight in the structure-activity rela-
tionships, luck and serendipity were definitely important factors on our side.

5. THE FINAL STEPS

At the time R 1472 was synthesized (Fig. 6) and tested we had a structure in our
hands which was devoid of morphinomimetic properties and which was almost
indistinguishable from chlorpromazine as far as its pharmacological CNS pro-
file was concerned. Subsequent substitution at both aromatic rings of R 1472
using simple substituents such as F, Cl, OMe and Me resulted in another eleven
compounds, shown in Table III, of which the last one was R 1625 or
haloperidol. Haloperidol was synthesized in February 1958 and filed at the Great

TABLE 111

AROMATIC SUBSTITUENT COMBINATIONS LEADING TO HALOPERIDOL (R 1625)

Lo %

Oy OH

xz

R-compound X, X, AD,/MD,*

1472 H H <0.06
1519 4-Cl H <0.15
1520 4-Me H <0.16
1532 H 4-Cl <0.2
1589 (peridol) 4-F H <0.05
1597 3-F H 0.34
1600 H 4-F <0.17
1605 2,4(CH,), H <0.6
1606 2,5(CH,), H <0.85
1611 4-OMe H <0.6
1616 4-F 4-F <0.08
1625 (haloperidol) 4-F 4-Cl <0.01

* Ratio of the hot-plate and mydriatic median effective doses.

 189

& OH
@c —CH,~CH;CHy-N

R1472

FIGURE 6. Structural formula of R 1472.

Britain Patent Office in April 1958 (Janssen, 1958). Haloperidol was by far the
most active neuroleptic known in 1958. It was several times more potent than
chlorpromazine, was both longer and faster acting and was almost devoid of the
antiadrenergic and other autonomic effects of chlorpromazine (Janssen and
Niemegeers, 1959; Janssen et al., 1960a, b; Delay et al., 1960; Divry et al., 1958,
1959, 1960).
Thus the course of events, from the chemical point of view, leading to
haloperidol can be summarized as follows: starting with the original objective
of increasing the analgesic potency of pethidine, R 951 came along with a struc-
ture which at that time was not thought to be compatible with high analgesic
potency. Subsequent drastic structural modification resulted in the
butyrophenone derivative R 1187 a less potent analgesic than R 951 but endow-
ed with chlorpromazine-like activity. Our smelling a rat directed us towards
more subtle and less drastic structural modification resulting in R 1472 which
brought us much closer to our goal. Fine tuning this structure by substitution
of both aromatic rings of R 1472 eventually gave us R 1625.

6. FROM HALOPERIDOL ONWARDS

The discovery of haloperidol generated an immense interest in the

butyrophenone derivatives resulting in the synthesis and pharmacological
evaluation of many thousands of compounds. From our laboratory alone, some
25 butyrophenone-type neuroleptics were evaluated in clinical trials. Ten of
these drugs are in use today in human and veterinary medicine (See Table IV).
A couple of months after the synthesis of haloperidol it was absolutely clear
to us that the p-fluorobutyrophenone moiety was an essential structural ingre-
dient of potent neuroleptics. In fact, as can be seen from Fig. 7, the frequency
of compounds containing the p-fluorobutyrophenone piperidine moiety steadily
increased with a maximum of about 25% of all compounds synthesized during
the period in which trifluperidol (R 2498) and pipamperone (R 3345) were
discovered. Over the following years the use of this moiety in our synthetic ef-
forts gradually fell off again as other interesting chemical fragments caught our
attention which eventually would lead to long acting neuroleptics of the di-
190

TABLE IV

DATE OF SYNTHESIS, YEAR OF INTRODUCTION ON THE MARKET AND NUMBER OF

PAPERS ON BUTYROPHENONE-TYPE NEUROLEPTICS

Compound Trade name Date of Year of Number of

synthesis introduction papers*

Haloperidol (R 1625) Haldol Feb. 1958 1959 4675

Fluanisone (R 2028) Haloanisone Aug. 1958 1960 214
Trifluperidol (R 2498) Triperidol Mar. 1959 1961 339
Pipamperone (R 3345) Dipiperon Jan. 1960 1961 191
Moperone (R 1658) Luvatren Feb. 1958 1963 139
Droperidol (R 4749) Dehydrobenzperidol Jun. 1961 1963 2272
Benperidol (R 4584) Frenactil Apr. 1961 1965 198
Azaperone (R 1929) Stresnil Jun. 1958 1968 215
Spiperone (R 5147) Spiropitan Dec. 1961 1969 505
Bromperidol (R 11 333) Impromen Apr. 1966 1981 65

* Number of scientific papers from the year of synthesis up to the middle of 1981. Private com-
munication: S. Van Gestel, Department of Medical Documentation, Janssen Pharmaceutica. A
search in the BIOSIS file (Bio Science and Information Science, 2100 Arch Street, Philadelphia,
PA 19103, U.S.A.) revealed that in the period 1977 to the middle of 1981, haloperidol and chlor-
promazine score 2422 and 2338 entries, respectively. The figures for the period 1969 — 1977 are
1196 for haloperidol and 3760 for chlorpromazine.

phenylbutylamine-type e.g. R 6238 or pimozide (Janssen and Tollenaere, 1981)

and the anthelmintic R 8299 or levamisole (Symoens, 1979). On the other hand
our firm belief in the p-fluorobutyrophenone moiety enabled us to focus our
synthetic efforts on the modification and the introduction of substituents at the
4-position of the piperidine and piperazine rings. Again this proved to be the
correct decision and led us to explore, among other things, the important class
of 4-anilinopiperidine derivatives which gave rise in June 1960 to R 4584 or
benperidol, R 4744 or droperidol and R 5147 or spiperone. R 3345 or
pipamperone was derived from 4-piperidino-piperidine whereas R 2028 or
fluanisone and R 1919 or azaperone were derived from 4-arylpiperazine. Fig.
8 shows the three-dimensional structure of the most important p-fluorobutyro-
phenone type neuroleptics as determined from X-ray crystallography data
(Tollenaere et al., 1979).
Careful pharmacological testing revealed that the chemical differences are
reflected in each drug having its own particular spectrum of pharmacological
properties and consequently its own special indications. Thus droperidol is
mainly used in neuroleptanalgesia whereas azaperone and fluanisone are used
in veterinary medicine. The remainder of these butyrophenone derivatives are
primarily intended for psychiatric prescription covering indications going from
various psychopathic syndromes such as destructive and belligerent behaviour,
 191

2 3 o
R1625 — Feb 1958
g : e
R 1929 Jun 1958
R 2028 Aug 1958

Mar 1959

R 3000 Oct 1959

R 3345 Jan 1960

R 4000 Sep 1960

R 4584 Apr 1961

R 4749 Jun 1961
R 5000 Oct 1961
R5147

R 6000 [ Nov 1962

R 7000 t— Jul 1963

FIGURE 7. Frequency of occurrence (N) of the p-fluorobutyrophenone piperidine moiety per hun-
dred R-compounds synthesized between R 1625 and R 7000.

o]
R1625 g )L CH{CHTCHZ-N
OH

CL
FIGURE 8. Legend on p. 194.
192

F [ OH
F -@E-CMTCH{CHZ'N

R 1658
CHy

R 1919 FOE-CHZ-CHZ-CHZ-NDQ
0

o OCHy

Rooos FObouegens

FIGURE 8 (continued). Legend on p. 194.

 193

o OH
R 2498 F@-g—CHz—CHz—CHI-N ):

CF3

1
C-NH,
R 3345 F@fi_CHfCHfCHZ-,Dg

[
F Ofi'c"z‘c“i CHzN O
o z §
R 4584

FIGURE 8 (continued). Legend on p. 194.

194

0
R 4749 F-@-fi—cuicuicnz-u >
0

[} [}
1 NH
R 5147 F-OC-CHQ-CHZ-CHiN XNJ

OH
R 11333 FQfi-cHicHiCHiN >§
o

Br

FIGURE 8. Perspective drawings based on X-ray crystallography data (Tollenaere et al., 1979) of
the most important butyrophenone-type neuroleptics.
 195

disturbing hypersexual tendencies, normalization of mood and disturbed sleep

patterns, psychomotor agitation and hallucinations to drug-resistant chronic
schizophrenia.

7. SUMMARY

By the end of 1961, we had discovered some ten valuable neuroleptics, all
derivatives of butyrophenone, while our original aim was to develop more
powerful analgesics through systematic and thorough structural modification of
pethidine. Careful evaluation of the pharmacological results followed by
clinical observations and a prepared mind enabled us in the early days of the
laboratory to recognize that the time was ripe to explore the CNS-depressing
properties of the compounds we were dealing with at the time. Hard work, an
open mind for the unexpected and serendipity eventually culminated in a struc-
turally novel compound that is called haloperidol. Novel it was, as the structure
of haloperidol was totally unrelated to the then known neuroleptics chlor-
promazine, its congeners and reserpine. The discovery of haloperidol set in mo-
tion a number of activities as varied as human activities can be and proved to
be a stepping stone for new discoveries in the field of analgesics and neuroleptics.

8. REFERENCES

Beckett, A. K. (1952) C. B. Patent No. 881893; Belgian Patent No. 577977 (1959); U.S. Patent
No. 3438991 (1960).
Braenden, O. J., Eddy, N. B. and Halbach, H. (1955) Bull. W.H.O. 13, 937 —998.
Delay, J., Deniker, P. and Harl, J. M. (1952) Ann. Méd. Psychol. (France) /10/2, 112—-117.
Delay, J., Pichot, P., Lempériére, T. and Elisalde, B. (1960) Press Méd. 68, 1353 —1355.
Divry, P., Bobon, J. and Collard, J. (1958) Acta Neurol. Psychiat. Belg. 58, 878 — 888.
Divry, P., Bobon, J., Collard, J., Pinchard, A. and Nols, E. (1959) Acta Neurol. Psychiat. Belg.
59, 337-366.
Divry, P., Bobon, J. and Collard, J. (1960) Acta Neurol. Psychiat. Belg. 60, 7—9.
Jageneau, A. H. M. and Janssen, P. A. J. (1956) Arch. Int. Pharmacodyn. Ther. 106, 199 —206.
Janssen, P. A. J. (1958) G.B. Patent No. 8811893; Belgian Patent No. 577977 (1959); U.S. Patent
No. 3438991 (1960).
Janssen, P. A. J. (1960) International Series of Monographs on Organic Chemistry, Vol. 3,
Pergamon Press, Oxford.
Janssen, P. A. J. and Eddy, N. B. (1960) J. Med. Pharm. Chem. 2, 31-45.
Janssen, P. A. J. and Niemegeers, C. J. E. (1959) Arzneim.-Forsch. 9, 765 —767.
Janssen, P. A. J. and Tollenaere, J. P. (1981) In: Chronicles of Drug Discovery, (Bindra, J. and
Lednicer, D., Eds.), Wiley-Interscience, New York, NY, in press.
Janssen, P. A. J., Jageneau, A. H. M., Van Proostdy-Hartzema, E. G. and De Jongh, D. K.
(1959a) Acta Physiol. Pharmacol. Neerl. 7, 373 —402.
Janssen, P. A. J., Jageneau, A. H. M., Demoen, P. J. A. and Van de Westeringh, C.,
Raeymaekers, A. H. M., Wouters, M. S. J., Sanczuk, S., Hermans, B. K. F. and Loomans,
J. M. (1959b) J. Med. Pharm. Chem. 7, 105-119.
196

Janssen, P. A. J., Jageneau, A. H. M., Demoen, P. J. A., Van de Westeringh, C., De Canniére,
J. H. M., Raeymakers, A. H. M., Wouters, M. S. J., Sanczuk, S. and Hermans, B. K. F.
(1959¢) J. Med. Pharm. Chem. 7, 309-317.
Janssen, P. A. J., Jageneau, A. H. M. and Huygens, J. (1959d) J. Med. Pharm. Chem. 7, 300 — 308.
Janssen, P. A. J., Niemegeers, C. J. E. and Schellekens, K. H. L. (1960a) Arzneim.-Forsch. 10,
955.
Janssen, P. A. J., Jageneau, A. H. M. and Schellekens, K. H. L. (1960b) Psychopharmacologia
1, 389-392.
Symoens, J., De Cree, J., Van Bever, W. F. M. and Janssen, P. A. J. (1979) In: Pharmacological
and Biochemical Properties of Drug Substances, (Goldberg, M. E., Ed.), Vol. 2, Am. Pharm.
Assoc.
Tollenaere, J. P., Moereels, H. and Raymaekers, L. A. (1979) Atlas of the Three-Dimensional
Structure of Drugs, pp. 100— 108, Elsevier/North-Holland Biomedical Press, Amsterdam.
Weber, E. (1954) Schweiz. Med. Wochenschr. 84, 968 —970.