Forty-three schizophrenic patients participating in this study were se
rotyped for human leukocyte antigens (HLA-A, -B, -C, -DR, -DQ antigens
). Thirty-six of them were hospitalised in two state mental hospitals
and 7 in our general hospital, psychiatric unit. The patients from our
unit were typed for HLA before commencing clozapine treatment whereas
the patients from state hospitals were typed after commencing treatme
nt. Three out of 43 patients developed agranulocytosis. One had a comb
ination of both 'high-risk' haplotypes (HLA-B 16(38,39), DR4, DQ3 and
HLA-DR2, DQ1), another had HLA-DR2, DQ1, whereas the last had a totall
y different haplotype. Between non-agranulocytic patients 1 was found
to carry the HLA-B16(38,39), DR4, DQ3 haplotype and 14 (out of 40) had
the HLA-DR2, DQ1. Taking into account other factors supposed to be in
volved (a noxious metabolite, and the presence of a humoral cytotoxic
factor) we must admit that despite the finding of a high-risk haplotyp
e in Jewish populations there are other aspects of this question await
ing clarification.