clozapine-induced agranulocytosis - researchgate

Hum. Genet. 37, 183--194 (1977)

© by Springer-Verlag 1977

Clozapine-Induced Agranulocytosis A Genetic and Epidemiologic Study

Albert de la Chapelle, Christina Karl, Markku Nurminen, and Sven Hernberg

Folkh~lsan Institute of Genetics and Institute of Occupational Health, Helsinki, Finland

Summary. An epidemic of agranulocytosis and granulocytopenla occurred in 1975 in conjunction with clozapine treatment of mental patients in Finland. An attempt was made to assess the epidemiologic and genetic factors contributing to the adverse drug effect. The estimated incidence rate in Finland was 2.1,/1000 patient-months. This figure could not be compared with rates from other countries because of the inexact nature of the figures reported so far. All 16 cases occurred in seven hospitals in southwestern Finland, whereas the overall hospital net use of the drug was geographically evenly distributed. The difference between the observed and the proportional- ly expected incidence of cases amongst the hospitals where clozapine was used was statistically significant. The average consumption of the drug did not differ between the hospitals where cases occurred and those where no definite cases could be diagnosed. Six-generation pedigree analyses failed to reveal significant parental consanguinity or genetic kinship between probands. Neither did the birth places of the ancestors of the probands disclose a typical isolate pattern. In conclusion, the cases appeared to be confined to a few hospitals in southwestern Finland. Although a genetic factor is not excluded, we found no evidence in support of a genetic mechanism.

Introduction

The neuroleptic drug clozapine (Leponex®) was released for use by the Finnish National Board of Health on 29 January 1975. The first case of agranulocytosis believed to be related to the intake of that drug was reported to the Drug Adverse Reaction Register at the Finnish National Board of Health on 16 June 1975. By 28 July 1975 so many cases had been reported that the use of the drug was prohibited in Finland by decision of the Board (Id~inp~i~in-Heikkil~i et al., 1975).

Investigations into this question produced three striking findings (Ad Hoc Committee, 1975). First, the proposed incidence of clozapine-induced agranulocytosis appeared to be far higher than had been anticipated from clinical trials

184 A. de la Chapelle et al.

ca r r i ed o u t in F i n l a n d and e l sewhere . Second ly , p r e l i m i n a r y d a t a s e e m e d to

ind ica te t ha t the inc idence o f a g r a n u l o c y t o s i s in F i n l a n d was s o m e 20 t imes

h i g h e r t h a n in o t h e r coun t r i e s , e.g., Aus t r i a , Swi tze r l and , and W e s t G e r m a n y ,

where the d r u g was wide ly used (Gr i f f i t h a n d Saamel i , 1975). Th i rd ly , the F i n n i s h

cases a p p e a r e d to be u n e v e n l y d i s t r i bu t ed t h r o u g h o u t the coun t ry .

These p r e l i m i n a r y f ind ings cou ld , i f ver i f ied , give a clue to the m e c h a n i s m by

which c l o z a p i n e b r ings a b o u t ag ranu locy tos i s . Whi l e it was poss ib le to o b t a i n

d a t a on the i nc idence o f the s ide effect in d i f fe ren t hosp i t a l s (in F i n l a n d ) and in the

c o u n t r y as a w h o l e (see be low) , c o m p a r a b l e d a t a f r o m o t h e r coun t r i e s were n o t

ava i lab le . W e were t h e r e f o r e u n a b l e to c o m p a r e the inc idence in F i n l a n d wi th

tha t in o t h e r coun t r i e s . In this p a p e r we p re sen t resul ts o b t a i n e d to test t w o

hypo theses : (1) T h a t the cases o f c l o z a p i n e - i n d u c e d a g r a n u l o c y t o s i s were un -

even ly d i s t r i bu t ed a m o n g hosp i t a l s in F i n l a n d and (2) t ha t a gene t ic f ac to r

c o n t r i b u t e d s ign i f i can t ly to the side effect.

Material and Methods

Definitions. Throughout this paper the term agranulocytosis is used for convenience to cover two different states of illness, namely true agranulocytosis and granulocytopenia as defined by other authors (Pisciotta, 1973; Ad Hoc Committee, 1975). For the sole purpose of this paper we assume that the Ad Hoc Committee's conclusion about a causal relationship between the intake of clozapine and agranulocytosis is correct.

The incidence rate of agranulocytosis is defined in this investigation as the number of cases of agranulocytosis which occurred in Finnish mental hospitals from 29 January to 28 July 1975, divided by the number of patients receiving a certain minimum total amount of clozapine times the duration of treatment calculated after a special minimum latent time. The denominator of the incidence rate used here is in patient-months of experience.

Patients with Agranulocytosis. The Finnish Board of Health distributed altogether three questionnaires. The first two, sent to all mental hospitals and comparable institutions, requested information on the number of patients treated with clozapine as well as detailed data on all those clozapine-treated patients in whom agranulocytosis or leukopenia had occurred. The third questionnaire was sent to all hospitals in the country. Data on all deaths were requested and obtained. Detailed hematologic information was requested on all dead patients who had been given clozapine. The conclusions reached were briefly as follows (Ad Hoc Committee, 1975; Palva and Id/inpggn-Heikkilfi,1976; Id~inp~i/in-Heikkil~i et al., 1977). Some 2500 to 3200 patients in about 70 hospitals had been given clozapine during the period; 17 cases of agranulocytosis had occurred, eight of which were fatal. After the drug was prohibited no further cases of agranulocytosis were reported.

In addition there were 11 patients who had died or suffered from undefined illness while on clozapine treatment. In the absence of hematologic and other specific data none of these cases could be in any way linked to the drug. The present analysis is based on 16 proven cases of agranulocytosis, on which data were kindly provided by the Board. One patient (case E/F9), originally considered to have had clozapine-induced agranulocytosis (Ad Hoc Committee, 1975), was excluded because she died of acute leukemia. Two of the 16 patients mentioned were treated under ambulatory conditions. One of them was an outpatient but received his drugs from the hospital. The other had been treated as an outpatient but was referred to a general hospital when symptoms of agranulocytosis were detected. Since all or most of the drug that she had taken was from pharmacies rather than from hospitals, she was excluded from the calculations on incidence.

The drug was generally used as prescribed by the manufacturer. Overdosage in general or in the case of the 16 patients in particular could be excluded. On the basis of the information

Clozapine-Induced Agranulocytosis 185

assembled by the Board of Health, the following causes were discarded as probable explana- tions of the side effect: Overdosage, simultaneous intake of other drugs, and chemical impurities in all or part of the batches of the drug (Id~inp~ifin-HeikkiRi et al., 1977).

Consumption of Clozapine. Clozapine was sold in Finland by its sole manufacturer under the drug name Leponex®. After its use had been prohibited, hospitals and pharmacie~ i'eturned their remaining stock to the manufacturer. The figures on drug consumption give only the net amounts of clozapine used, the total being 54,600g. Since clozapine is a new drug and it is known to be effective in the treatment of psychotic patients, it was mainly used in hospitals (39,700 g), 14,900 g being sold by pharmacies to open care patients (by physician's prescription). The amounts of clozapine sold by pharmacies were evenly distributed throughout the country. With the exception of the two patients mentioned above, no side effects were reported in outpatients, but reliable data on this question are obviously difficult to obtain.

Restriction of Study Material. In this report, only the figures for hospitals (71 in all) were considered in the calculations of the denominator of the incidence rate. Those 17 hospitals for which no data were available either on the number of patients receiving medication, or on the amount of the drug consumed, or on the month the administration of the drug was started were disregarded. Thus, the number of hospitals in the study was reduced to 54. Obviously a certain latent time and a certain minimum of the drug are needed to induce the development of agranulocytosis. Data on the smallest total dose administered before the appearance of the first symptoms of infection or diagnosis of leukopenia were used to estimate the minimum dose. The latency time was estimated from the limited data on the period between the start of treatment and the exact time of case inception. When the hospitals were restricted to include only those with a minimum total medication of 10 g clozapine per patient and allowing a latent time of at least 1 month, only 27 hospitals were left for the analysis.

Statistical Methods. A Poisson probability model was assumed for the occurrence of the number of incident cases separately for each of the hospitals considered. First, we assumed that an equal probability of inception of agranulocytosis had prevailed in all the hospitals where clozapine was used (testable null hypothesis). As an alternative hypothesis we proposed an unequal distribution of the incident cases amongst the 27 hospitals in Finland, i.e., an unproportional allocation of cases, with respect to the accumulated patient-month experience, to those hospitals influenced by a risk factor, as yet unknown.

Genealogic Methods. The parish registries in Finland make it possible to trace people's genealogy back some 300 years. In the church records, each person is identified at least by name, date and place of birth, and date of death. With the aid of these records, we constructed pedigrees comprising six generations for each proband with agranulocytosis.

Results

Incidence of Clozapine-Induced Agranulocytosis in Finland. Since the a m o u n t s o f

c l o z a p i n e a d m i n i s t e r e d wh ich e x c e e d e d the ave r age to ta l o f 10g pe r pa t i en t

s h o w e d a h igh c o r r e l a t i o n wi th the a c c u m u l a t e d p a t i e n t - m o n t h s o f expe r i ence

(r = 0.96), the i nc idence ra te was c o m p u t e d e m p l o y i n g o n l y p a t i e n t - m o n t h s in the

d e n o m i n a t o r o f the inc idence ra te measu re . U t i l i z ing the l imi ted i n f o r m a t i o n

f r o m the 27 hosp i t a l s , we a r r ived at the f o l l o w i n g e s t ima te o f the f r e q u e n c y o f

a g r a n u l o c y t o s i s in c o n n e c t i o n wi th c l o z a p i n e t r e a t m e n t in F i n l a n d : I n c i d e n c e ra te

(es t imate ) = 15 cases p e r 7043 p a t i e n t - m o n t h s , t ha t is, 2 .1 /1000 p a t i e n t - m o n t h s .

Geographical and Interhospital Differences in the Incidence of Clozapine-Induced Agranulocytosis in Finland. T h e d i s t r i b u t i o n o f the a m o u n t s o f c l o z a p i n e used in all

hosp i t a l s in F i n l a n d d u r i n g the p e r i o d J a n u a r y to J u l y 1975 is s h o w n in F i g u r e 1.

~ A S T A ~ A 1.1.,,,m?s I F " ~

s-~s-

0


The geographic location of those seven hospitals in which clozapine-induced agranulocytosis occurred is shown in Figure 2. The difference in the appearance o f the two figures is striking.

The incidence rates for the individual hospitals involved ranged from 22.2/1000 patient-months to 1.1/1000 patient-months. The data used for testing the unevenness of the case allocation to the hospitals are given in Table 1. On the basis of the accrued patient-months, the deviation of the observed numbers of cases f rom those expected theoretically, when all the 27 hospitals are regarded as stochastically independent, is statistically significant at the probability level of P < 1% and, when the noninvolved hospitals are pooled, at the level of P < 0.5%, a one-sided Poisson dispersion test (Documenta Geigy, 1975).

However, the average amount of the drug taken by the individual patients did not differ in a statistically significant way between the hospitals involved (mean 17.5 g, range 13.1--21.4g) and those not involved (mean 19.1 g, range 10.0--48.0g).

Genetic and Genealogic Studies. The uneven geographic distribution of the cases of agranulocytosis as compared with the overall use of the drug in hospitals probably reflects the action of one or several local factors, either environmental or inherited.

If genetic factors contributed to the development of agranulocytosis in patients taking clozapine, a number of correlates might be expected. First, if it were a rare recessive gene, consanguinity of the parents would be likely. Secondly, if it were a rare dominant , the probands might be expected to be related. Thirdly, since Finland has until recently been a country of typical regional genetic isolates, a high incidence of the putative gene might have occurred in some regions. The geographic distribution of the birth places of the probands and their ancestors might therefore indicate their provenience from one or several isolates. These questions were tested by genealogic studies.

A. Pedigrees. A typical pedigree is shown in Figure 3. All probands but one were of Finnish origin: The exception was a proband born in Finland whose parents originated f rom the Soviet Union, and whose pedigree could not be traced. With the exception of this proband, the number and proport ion of individuals that could be identified is indicated in Table 2. It is seen that the proport ion is quite high, the total being 86%.

B. Consanguinity. In all of the remaining 15 probands a careful search was made for consanguinity; none was found between the parents of any one proband when the analysis was restricted to six-generation pedigrees. In two of the families, however, the close location of the birth places of individuals in generation I hinted at the possibility of consanguinity, and thus the ancestors in question were traced back a further two generations. In one, the search yielded no evidence of

Fig. 1. Map of Finland showing geographic distribution of the consumption of 39 700 g of clozapine in 61 Finnish hospitals and institutions during the period January to July 1975. Hospitals in which less than 8 g was used were excluded. Each circle represents one hospital. The area of the circles indicates (in arbitrary units) the amount of drug consumed

" n ~ A ~A'n~A I.I.WTS


Table 1. Distribution of agranulocytosis cases, consumption of clozapine, and patient-months among the hospitals, with estimated incidence rates and theoretical expected numbers

Hospital Clozapine Number Patient- Incidence Expected number consumption of cases months rate/1000 number" i (g) ci tl patient- ei

months

1 530 4 180 22.2 0.4 2 855 3 162 18.5 0.3 3 726 1 136 7.4 0.3 4 1 817 2 417 4.8 0.9 5 4 273 3 820 3.7 1.7 6 1 728 1 436 2.3 0.9 7 4 948 1 932 1.1 2.0

Total 1--7 14 877 15 3 083 4.9 6.6 8--27 14 251 0 3 960 0 8.4 1--27 29 128 15 7 043 2.1 15.0

28--54 7 443 0 5 200 0 1--54 36 571 b 15 12 243

a ei = (~2 c i ) ( t i / y , ti) b Estimated total net use of clozapine in 62 hospitals was 39 738 g; for 9 hospitals exact figures were not available

consanguin i ty . In the Other, two members o f genera t ion I were shown to have had a c o m m o n g randfa the r bo rn in 1690.

Hence , the pa ren t s o f on ly one o f 15 p r o b a n d s were d is tan t ly related. This is in fact fa r less than might be expected in rura l F inn i sh popu la t ions (see below). We conc lude tha t the lack o f consanguin i ty between the paren ts in the six- gene ra t ion pedigrees does no t suppor t the no t ion o f a rare recessively inher i ted m u t a t i o n being respons ib le for the side effect o f the drug.

C. Genet ic Re la t ion of P r o b a n d s to Each Other . Mos t of the pat ients with agranu locy tos i s had sch izophren ia (Ad Hoc Commit tee , 1975). Since all the cases occur red in five cities only, and since the disease has a s t rong genetic componen t , it wou ld not be unexpec ted to f ind in ter re la ted p robands , even in the absence o f a genet ic c o m p o n e n t re la ted to agranulocytos is . However , no such in te r re la t ion o f p r o b a n d s was f o u n d within the s ix-genera t ion pedigrees, a l though fur ther studies revealed re la t ions in ear l ier genera t ions . The mothers o f two o f the p r o b a n d s were d i s tan t ly re la ted, a m e m b e r o f genera t ion I o f one p r o b a n d being the niece o f a m e m b e r o f genera t ion I in the o ther p r o b a n d . The fathers o f two o ther p r o b a n d s were d i s tan t ly re la ted. They had c o m m o n ancestors (born in 1690 and 1696) eight genera t ions removed .

Fig. 2. Geographic location of hospitals in which 16 patients with clozapine-induced agranulocytosis were treated in Finland during the period January to July 1975. There were agranulocytosis patients in seven hospitals. An eighth hospital is indicated in which a patient was treated for agranulocytosis after having received ambulatory clozapine treatment from a physician at another hospital. Each circle represents one hospital. The area of the circles indicates (in arbitrary units) the number of affected patients


r~

~ " C I


Table 2. Theoretical number of persons in each generation of 15 pedigrees, and the number and proportion of persons traced in this study

Generation Theoretical Number Proportion number of persons traced of persons traced (%)

I 480 397 83 II 240 216 90

III 120 109 92 IV 60 58 97 V 30 29 97

VI 15 15 100

Total 945 824 87

Hence, in two cases probands were distantly related. This figure is far lower than might be anticipated in typical Finnish rural communities. Nevanlinna's (1972) figures for one such parish revealed, for instance, that of 60 randomly selected probands, 18 parents were third cousins and as many as 52 parents were fourth cousins. We conclude that the probands with clozapine-induced agranulocytosis were not genetically related in excess of what might be expected on a random basis. This finding does not lend support to a rare mutation (recessive or dominant) as the main factor responsible for the side effect.

D.Geographic Provenience of the Probands and Their Ancestors. The Finnish population structure is well known from a genetic point of view (Nevanlinna, 1972). Movements within the population breaking up the pronounced rural isolates and producing city populations of mixed origin have occurred only during the past 100 years, and it is only since World War I that this transition has been of major importance. Thus, the provenience of the genes of any given individual can be adequately established by determining the birth places of his ancestors two or three generations back. For this purpose we chose generation III, the proband 's great grandparents, numbering eight in each case.

Figure 4 shows the birth places of those 109 ancestors in generation I I I that could be definitely traced (Table 2). There is a concentration of birth places in Southwestern Finland, particularly in the region around the city in and near which most of the cases occurred (Tampere). This is to be expected considering that most mental patients are referred to hospitals in their own districts. Even so, it is an interesting point that many of the ancestors come from beyond the southwestern region in which all the agranulocytosis cases occurred.

There was no concentration of birth places in small parish-sized isolates. Rather, the distribution of ancestral birth places appeared fairly random in view of the geographic location of those hospitals in which agranulocytosis occurred. Hence, there was little or nothing to indicate that highly localized enrichment of a putative gene within Finland had played any major part in the causation of the side effect.

TILASTOIU~RTTA I~'TA~Y/.K.q~A t1.1|7|

D

Fig. 4. Geographic distribution of the places of birth of 109 ancestors in generation III of 15 probands with clozapine-induced agranulocytosis. Each dot represents one person


Discussion

The Ad Hoc Committee of the Finnish National Board of Health concluded (1975) that nearly 20 cases of agranulocytosis were causally related to the intake of clozapine. Even though this view may be challenged, we have assumed here that it is correct and we shall not discuss it further.

We wished to know whether the incidence in Finland is higher than that in other countries. For instance, in the light of previous data, Griffith and Saameli (1975) concluded that the incidence was approximately 20 times higher in Finland than elsewhere. In a more recent report, Gram (1976) likewise found a lower incidence in Denmark than that reported from Finland. Unfortunately, true incidence figures for many countries, e.g., those of Central Europe, are not available. There are major differences in the efficiency of reporting and in the definitions of the risk period. Further difficulties arise from the vague character of the estimations of size of the populations at risk and from the fact that precise register data like those in Finland are not available in most countries. We therefore feel that we are not yet in a position to make an adequate comparison between incidences in different countries, not even in crude relative terms.

An important assumption to be discussed is whether agranulocytosis was reported equally efficiently in all or most Finnish institutions. Serious under- reporting from some of the Finnish hospitals is unlikely but not entirely inconceivable. However, the third questionnaire, asking about all patient deaths and requesting hematologic details on all those dead patients who had taken clozapine should have revealed most cases. Therefore, the absence of cases in several large hospitals in eastern and northern Finland remains unexplained. We shall possibly never know whether the differential distribution within Finland was artifactual or not, because the drug is no longer used in Finland.

In defining the denominator of the incidence rate, i.e., in the specification of the populat ion/ t ime reference for the study population, variation in the latent time has to be considered. When the objective of an epidemiologic investigation is to determine the maximum response of human populations to an agent, e.g., when attempting to demonstrate that a causal relationship exists, it is important that the latent period be optimized. But in the statistical analysis of the incidence rates in this study, these considerations were not of pr imary concern since we refrained f rom making international comparisons and limited the study to interhospital differences in Finland. Thus we defined the risk period by the same criteria for all the hospitals.

The present evidence indicates that the geographic distribution within Finland was grossly uneven. The obvious interpretation is that a local factor must have contributed to the adverse side effect. Such a factor could be environmental or genetic, or both.

In the absence of direct tests for a putative gene for clozapine sensitivity, we tried using genealogic methods, but noted neither parental consanguinity nor kinship between probands. This fails to confirm the notion that a rare gene contributed to the side effect.

It remains to be considered whether there is a gene, another intrinsic factor, or a combination of genes or factors that are extremely common in one part of


Finland but rare in others and possibly in other countries. Some diseases are markedly prevalent in Finland, and especially so in certain parts of the country, whereas others are rare (Norio et al., 1973), a phenomenon that has been ascribed to the peculiar popula t ion structure o f this count ry (Nevanlinna, 1972). Even so, genes which are very c o m m o n in Finland and rare elsewhere are mostly fairly evenly distributed within Finland. Hence, we find no direct evidence in favor o f a gene effect. But the genealogic methods available do not, of course, permit us to rule out an individual predisposition.

If a genetic factor is not responsible, it remains to be considered whether an environmental factor might have operated. Such a factor could well have been present in some of the hospitals in par t of the country, and it could even have been of limited durat ion. Even though such environmental factors seem feasible enough, their evaluat ion is beyond the scope of this paper.

R e f e r e n c e s

Ad Hoc Committee of the Finnish National Board of Health: Harmful complications from clozapine (Leponex®) (in Finnish). Suom. L/i/ik.-L. 30, 2501--2506 (1975)

Documenta Geigy: Scientific Tables, Seventh Edition, pp. 191--192. Basel: Ciba-Geigy 1975 Gram, L. F.: Bone-marrow depression following treatment with clozapin (Leponex®). Ugeskr.

Laeger 138, 1271--1273 (1976) Griffith, R., Saameli, K.: Clozapine and agranulocytosis. Lancet 197511, 657

• Id~inp/i~in-Heikkil/i, J., Alhava, E., Olkinuora, M., Palva, I.: Clozapine and agranulocytosis. Lancet 197511, 611

Id~inp~i~in-Heikkil~i, J., Alhava, E., Olkinuora, M., Palva, I.: Agranulocytosis during treatment with clozapine. Europ. J. Clin. Pharmacol. (1977, in press)

Nevanlinna, H. R.: The Finnish population structure. A genetic and genealogical study. Hereditas 71, 195--236 (1972)

Norio, R., Nevanlinna, H. R., Perheentupa, J.: Hereditary diseases in Finland; rare flora in rare soil. Ann. Clin. Res. 5, 109--141 (1973)

Palva, I., Idanp&ian-Heikkil~i, .1.: Hematological dangers of clozapine - - A reply from the ad hoc committee (in Finnish). Suom. L/i/ik.-L. 31, 52 (1976)

Pisciotta, A. V.: Immune and toxic mechanisms in drug-induced agranulocytosis. Semln. Hematol. 10, 279--310 (1973)

Received January 17, 1977

clozapine-induced agranulocytosis - researchgate

Documents