Genetic variation of 5-HT2A receptor and response to clozapine

information that does not rely on people’s memories, and is available for research. Retrieving and checking the above information took an inordinate amount of time. In 19842Ipointed out the total lack of statistical information and the difficulty in obtaining information. Further, I have lately pointed out3 that Macleod and Sharp had come to identical conclusions about the difficulty in obtaining information. I wrote, "Only when an acceptable registration system has been established will progress be made in answering these questions", to establish whether injuries are increasing or decreasing, and the incidence of injuries in Rugby Union. I hope that such a register is formed and all injuries are recorded, but this involves work, time, energy, and money, and unless all these are found, such a registration scheme will not take place. With thanks to the Technical Officers of the English, Irish, and Welsh Rugby Football Union.

J R Silver National

Spinal Injuries Centre, Stoke Buckinghamshire HP20 8AL, UK

Mandeville

Hospital, Aylesbury,

Rugby Union Handbook. Aylesbury:

Rothmans

1

Rothmans

2

Publications, 1981 and 1984. Silver JR. Injuries of the spine sustained in rugby. BMJ 1984; 288: 37-43.

3

Silver JR. Methods of recording injuries statistics in rugby football: the experience in the United Kingdom. Sports Exer Injury 1994; 1: 46-51.

Genetic variation of 5-HT2A response to clozapine

receptor and

SiR-Arranz and colleagues (July 29, p 281) show that, for the thymine to cytosine (T to C) nucleotide polymorphism at position 102 of the 5-hydroxytryptamine (serotonin) type 2 receptor (5-HT2A) gene, the genotypes T102/C102 and T102/T102 are more common in schizophrenia patients who respond to clozapine compared with those who do not respond. Clozapine has shown efficacy in 30-60% of patients who have not responded to traditional antipsychotic drugs, and it may induce fewer undesirable extra-pyramidal effects.’ This dibenzodiazepine derivative has high affinity at 5-HT2A receptors,’ and thus may be important for predicting response to the drug. Such prediction is clinically important: responders can be spared delay resulting from failure to respond to other medications and non-responders may be protected from the potential adverse effects of clozapine (eg, agranulocytosis). The mechanism of antipsychotic action may also be elucidated through investigation of the genetic variants associated with response. Data from patients with DSM-111-R diagnoses of schizophrenia were collected at the following research clinics: Long Island (JAL), Cleveland (HYM), the Bronx (SS), and Ottawa (PC). After informed consent was obtained, the patients underwent a washout period of 2-4 weeks during which, unless clinically necessary, they received no medication prior to starting clozapine. Treatment continued for a minimum of 6 months before their response to clozapine was determined. This response was defined as a reduction of at least 20% in the Brief Psychiatric Rating Scale (BPRS) from baseline score at enrolment into the study. Clozapine blood levels were monitored during treatment to assess compliance. All genotyping of the patients’ DNA, from participating clinics, was performed at the Clarke Institute of Psychiatry, Toronto. 126 patients were evaluated: 37 women, 89 men (25 African-Americans, 101 whites); mean age was 32-6 (SD 8-4) years. 72 patients responded to clozapine, and 54 did not. Allele frequencies did not differ significantly

1108

Table: Genotype distributions of the treated patients

5HT2A gene in clozapine-

between patients from different centres or between AfricanAmericans and whites. Age, sex, and ethnicity factors were simislar between responders and non-responders. Our study yielded no evidence that the genotype or allele frequency of the T—C 102 polymorphism of 5-HT2A is associated with response to clozapine (table, chi-square test, p=05). Our genotypes did not deviate significantly from Hardy-

Weinberg equilibrium. Our failure to replicate Arranz and colleagues’ finding may be attributed to sample differences, or to differences in psychiatric rating scales. Our scale (BPRS) is more focused on the measurement of psychopathology than the GAS scale used by Arranz et al. The two studies did not differ in the percentile proportion of responders. There was a difference ratio. We found no differences in the allele frequencies between the two studies; we therefore combined both data sets (n=275) and found that, overall, responders were more likely to have the genotype T102/C102 than nonresponders, and non-responders were more likely to be C102 homozygotes (p=0’014). The allele frequency differences for responders versus non-responders did not reach statistical significance (p=0055). The T-4C 102 polymorphism of 5-HT2A does not alter the predicted aminoacid sequence of the receptor, but may be in linkage disequilibrium with a nearby polymorphism which is involved in clozapine response. Further studies are required to assess whether this polymorphism may have a predictive role for response to clozapine that is clinically relevant. in the

sex

The contributions of Phil Cola, Alfreda Howard,

Barry Jones, and

Serge Sevy are gratefully acknowledged. Mario Masellis, Andrew D Paterson, Farideh Badri, Jeffrey A Lieberman, Herbert Y Meltzer, Patricia Cavazzoni, *James L Kennedy *Neurogenetics Section, Clarke Institute of Psychiatry, Toronto, Ontario, M5T 1R8, Canada: Hillside Hospital, Glen Oaks, New York, USA; Case Western Reserve University, Cleveland, OH USA; Bronx VA, New York; Department of Psychiatry, McMaster University, Hamilton, Ontatrio; and Department of Psychiatry, Royal Ottawa Hospital, Ottawa, Ontario

1

Meltzer HY. An overview of the mechanism of action of clozapine.

J Clin Psychiatry 1994; 55 (S9): B47-52.

Simple

measure

of insulin resistance

SiR-We welcome the opportunity to examine the utility of the simple fasting insulin resistance index (FIRI-fasting glucoseXfasting insulin/25) proposed by Duncan et al (July 8, p 120) using the Caerphilly prospective study of ischaemic heart disease. We have previously reported absence of an independent contribution of fasting insulin levels to the prediction of new, major ischaemic heart events (validated myocardial infarction-fatal and non-fatal-and new major Q waves on ECG) in 5 years of follow-up’ and also the extent of the contribution of fasting glucose levels to non-diabetic mortality.2 The following variables were entered into logistic regression analyses in which the dependent variables were death from any cause and major ischaemic heart event (as defined above): age, smoking habit (never, ex, or current), pre-existing ischaemic heart disease, diastolic blood pressure, body mass index, total cholesterol, high-density