HLA-B38, DR4, DQw3 and clozapine-induced agranulocytosis in Jewish patients with schizophrenia.

Agranulocytosis develops in approximately 1% of patients with chronic schizophrenia treated with the atypical neuroleptic drug clozapine. Previous studies have not identified the mechanism or risk factors for this adverse reaction. Because of an observed association between Jewish ethnic background and the development of agranulocytosis in our patient sample treated with clozapine for refractory symptoms, HLA typing was performed in 31 patients (19.4% of whom had developed agranulocytosis). The HLA-B38 phenotype was found in 83% of patients who developed agranulocytosis and in 20% of clozapine-treated patients who did not develop agranulocytosis. Because B38 is part of a haplotype known to occur frequently in the Ashkenazi Jewish population, the frequencies of the combined alleles HLA-B38, DR4, and DQw3 were examined. The incidence of HLA-B38, DR4, DQw3 was significantly increased in patients with agranulocytosis (five of five patients) compared with control patients of Ashkenazi Jewish ancestry (two of 17 patients). These findings indicate that genetic factors marked by major histocompatibility complex haplotypes may be associated with the susceptibility of Jewish schizophrenic patients treated with clozapine to develop agranulocytosis. We postulate that gene products contained in the haplotype may be involved in mediating drug toxicity.

Neuroleptic-associated autoantibodies. A prevalence study.

Antinuclear antibodies (ANA), antiphospholipid antibodies (APA), rheumatoid factor (RF), and immunoglobulin (Ig) M levels were determined in 184 male chronic psychiatric patients on long-term therapy with neuroleptics, and in 35 age-matched normal male controls. The prevalence of one or more of these autoantibodies was 70% in the neuroleptic-treated patients and 9% in the normal controls. Polyclonal IgM elevation was frequently seen among patients treated with phenothiazines. There was a significant correlation between the presence of ANA, APA, and RF; ANA and APA were more frequently associated with CPZ therapy, but the prevalence of RF was high in all treatment groups. These findings suggest that antibodies against the Fc fragment of IgG are the most common autoantibody associated with neuroleptics. Alternatively, the presence of RF could be a common finding in patients with neuropsychiatric disorders.

Immunogenetic markers in chlorpromazine-induced tardive dyskinesia.

The presence and severity of tardive dyskinesia were determined in 66 patients with chronic psychiatric disorders treated with chlorpromazine. The patients were classified according to the presence of antinuclear antibodies, the lupus anticoagulant, and the HLA antigen Bw44. The severity of orofacial dyskinesia was estimated using the Rockland Research Institute Scale. Patients with autoantibodies and the Bw44 antigen had higher tardive dyskinesia scores than those with AAB without the Bw44 antigen and also patients without autoantibodies regardless of their HLA phenotype (P less than 0.01). These studies suggest that the presence of autoantibodies in association with the HLA Bw44 antigen is related to, and can be a predictor of, neurological complications of long-term chlorpromazine therapy.

Clinical trial of a new bleeding-time device.

The authors report a clinical trial comparing a new bleeding time device (Simplate II) with the Mielke Template. Bleeding times were determined at the same time on the same arm using the two devices. Subjects of the study were 24 healthy volunteers, before and two hours after ingestion of 975 mg aspirin, and 28 patients. For the normal subjects the mean pre-aspirin bleeding times were 4.75 +/- 1.42 minutes (1 SD) with the Simplate II and 3.65 +/- 1.22 minutes with the Mielke Template. The mean bleeding times two hours after ingestion of aspirin were 7.86 +/- 2.76 minutes with the Simplate II and 7.84 +/- 2.94 minutes with the Template. The pre- and the post-aspirin values with the two devices were not significantly different from each other, nor were the bleeding times obtained in the patients with the two devices. The extents of scarring were similar with the two devices. The results were highly reproducible by both methods. The new device was simpler and more rapid to use.

Chlorpromazine-induced lupus anticoagulant and associated immunologic abnormalities.

Chronic administration of chlorpromazine is associated with the development of a lupus-like circulating anticoagulant and a variety of immunological abnormalities. The prevalence of these findings was studied in 123 psychiatric patients. The anticoagulant was present in 11 of 30 patients receiving chlorpromazine (CPZ), in none of 17 patients who had been off phenothiazine therapy for over a year and in none of 53 controls. It was also seen in 5 of 13 patients who had been switched from CPZ to another phenothiazine even after several years being off CPZ. The anticoagulant was characterized by prolongation of the partial thromboplastin time, thromboplastin dilution test, and Russell's viper venom time. Washed frozen platelets partially corrected the abnormality induced by the anticoagulant. In all but one case the anticoagulant was associated with positive antinuclear antibody test and/or increased serum IgM. Six of 16 patients also had decreased complement levels, and two had a positive direct Coombs' test. None of these patients manifested bleeding, hemolysis, splenomegaly, or other clinical features of systemic lupus erythematosus.

Anticardiolipin antibodies associated with HTLV-III infection.

Anticardiolipin antibody levels were determined in 73 homosexual men. Thirty of these patients had acquired immunodeficiency syndrome (AIDS), 16 patients had the AIDS-related complex (ARC) and 27 were healthy. Antibodies to human T-lymphotropic virus type III (HTLV-III) were detected in all patients with AIDS and ARC and in 11 of the healthy homosexuals. Eight patients with positive fluorescent treponemal antibody absorption test were excluded from the study. High levels of IgG-anticardiolipin antibodies were present in 23 of the 28 patients with AIDS; 12 of the 14 with ARC; five of the 10 HTLV-III positive healthy homosexuals; and none of the 13 HTLV-III negative healthy homosexuals. High levels of IgM-ACA were detected only in four patients with AIDS. The IgG-anticardiolipin levels were higher in the HTLV-III positive patients than in the HTLV negative group (P = 0.013). None of the patients with anticardiolipin antibodies exhibited venous thrombosis.

Chlorpromazine-induced anticardiolipin antibodies and lupus anticoagulant: absence of thrombosis.

Anticardiolipin antibodies were determined in 96 psychiatric patients treated chronically with chlorpromazine by an enzyme-linked immunosorbent assay using anti-IgM and anti-IgG (fab'2 fragment) as the second antibody. Fifty-four of these patients had an IgM-lupus anticoagulant, and the remaining 42 were followed as controls. Elevated IgM-anticardiolipin antibodies (ACA) levels were detected in 31 patients with the lupus anticoagulant and in 5 controls (p less than 0.001). During a median followup of 5 years, single episodes of deep vein thrombosis or pulmonary embolism occurred in three patients; one had the lupus anticoagulant and the other two had low-level ACA. Contrary to the reported experience in systemic lupus erythematosus and related autoimmune disorders, chlorpromazine-induced lupus anticoagulant and anticardiolipin antibodies levels appear not to be associated with an increased incidence of thrombosis.

Anticardiolipin antibodies in a sample of chronic schizophrenics receiving neuroleptic therapy.

Drug-induced antiphospholipid antibodies have been considered insignificant, a belief that has recently been questioned. Previous studies noted an association between chlorpromazine treatment and the presence of anticardiolipin antibodies (ACA), especially IgM ACA, and have suggested that thrombosis might be more likely in the presence of IgG ACA. We studied the cases of 27 patients receiving long-term neuroleptic therapy who had no history of cerebrovascular disease. IgG ACA was present in 7 (26%) of the 27, and IgM ACA was present in 5 (19%), yet none had shown any manifestations of cerebrovascular disease since initiation of neuroleptic therapy. These data suggest that drug-induced ACA do not predispose to cerebrovascular disease, even when IgG ACA is predominant.

A chlorpromazine-induced inhibitor of blood coagulation.

An inhibitor of blood coagulation was detected in the plasma of four schizophrenic patients receiving long-term chlorpromazine therapy. The partial thromboplastin time (PTT) was prolonged in all four patients and the prothrombin time (tpt) was prolonged in one. The inhibitor which resembled in many respects that seen in patients with systemic lupus erythematosus (SLE), was shown to be associated with the tigM-rich fractions of the serum. None of the patients manifested a clinical bleeding tendency and in two, discontinuation of the chlorpromazine was followed by a decrease in the activity of the anticoagulant. While three of the patients had a positive antinuclear antibody test (ANA) and the fourth a false positive serology for syphilis, no further evidence for SLE was found.

The hemostatic defect of chronic liver disease. Kinetic studies using 75Se-selenomethionine.

With the use of cohort labeling with 75Se-selenomethionine, simultaneous platelet, fibrinogen, and plasminogen survival studies were carried out in 8 patients with chronic alcoholic liver disease and in 5 normal subjects. Clinical features, liver function tests, coagulation and fibrinolytic system activities, and platelet function were also assessed. On the basis of platelet survival, the patients could be divided into two groups. Three patients had shortened platelet survival; they were all thrombocytopenic and had greater prolongation of the prothrombin time (PT) and activated partial thromboplastin time (PTT) than the other 5 patients. However, platelet turnover was decreased in all the patients, and there was no difference between the two groups with regard to fibrinogen or plasminogen survival nor in the in vitro evidence of disseminated intravascular coagulation (DIC). Fibrinogen survival was increased in 5 of the 8 patients. Plasminogen survival was normal in 6 patients and prolonged in 2 patients with very low plasminogen levels. The absence of increased fibrinogen turnover in the patients studied indicates that the abnormalities in coagulation tests were not due to consumption coagulopathy. The authors' studies suggest that, at least for patients with chronic stable alcoholic liver disease, the concept that the coagulopathy of liver disease is due to increased utilization of clotting factors should be revised with caution.

Production of anticardiolipin antibodies by cultured human lymphocytes.

Antibodies against negatively charged phospholipids, such as those to cardiolipin, can often be detected in the serum of patients with autoimmune related conditions, chronic infections and in patients treated with phenothiazines. In the present study, peripheral blood lymphocytes from nine healthy controls and eight patients with phenothiazine-induced IgM anticardiolipin antibodies (ACA) and the lupus anticoagulant were placed in vitro. Culture supernatants were assayed for ACA by measurement of optical densities using an ELISA. A significant difference (p less than 0.05) was demonstrated between the mean concentration of culture supernatant ACA from the patients as compared to healthy controls. The concentration of ACA in culture supernatants strongly correlated (r = 0.85) with that from the serum. There was a weak correlation between serum and culture supernatant ACA concentration and the lupus anticoagulant activity as measured by prolongation of the partial thromboplastin time. This technique uses readily accessible peripheral blood lymphocytes and should permit dissection of cytokine and cellular immune pathways regulating APA production.

Characterization and antigenic specificity of chlorpromazine-induced antinuclear antibodies.

Prolonged treatment with chlorpromazine is often associated with the development of antinuclear antibodies, an immunoglobulin M lupus anticoagulant, and polyclonal serum IgM elevation, but not with clinical features of systemic lupus erythematosus (SLE). Sera from 62 long-term psychiatric patients given treatment daily with 100 mg or more of chlorpromazine for at least 1 year were screened for antinuclear antibodies by indirect immunoperoxidase assay using HEp-2 cells. In 26 samples, antinuclear antibody titers greater than or equal to 1:40 with a homogeneous pattern were seen when anti-human IgM was used as the second antibody, three sera samples reacted with IgG, and four samples reacted with both IgG and IgM antisera. The antinuclear antibody antigenic reactivity was investigated by using histone and nonhistone nuclear antigens by enzyme-linked immunosorbent assay and passive hemagglutination techniques. Forty serum samples reacted with histone. Twenty-five samples reacted with deoxyribonucleoprotein (DNP), 28 with single-stranded DNA, and two with double-stranded DNA. No reaction was obtained with the extractable nuclear antigens RNP or Sm. These results indicate that chlorpromazine-induced antinuclear antibodies, like the antinuclear antibodies induced by hydralazine and procainamide, react mainly with histone nuclear antigens. Unlike the hydralazine and procainamide response, in which both IgG and IgM antibodies are demonstrated, the chlorpromazine-induced autoantibodies are predominantly of the IgM class.

Restriction fragment length polymorphism analysis of the V kappa locus in human lupus.

To evaluate the degree of genetic polymorphism of the V kappa repertoire in systemic lupus erythematosus (SLE), we performed Southern blot hybridizations with human gene probes corresponding to the four human V kappa gene families. In a comparative analysis, non-lymphoid cell DNA samples from three patients with idiopathic SLE, eight subjects with susceptibility to drug-induced lupus and seven control individuals were digested with the restriction endonucleases Bam HI, Bg 1 II, Eco RI and Hind III, and hybridized sequentially to the four V kappa family-specific probes. The restriction patterns on Southern blots revealed a low degree of polymorphism of the human V kappa gene repertoires of SLE patients and control individuals. This analysis, together with previous parallel studies of the V kappa locus in lupus-prone mice, implies that autoantibody hyperproduction in lupus is not associated with major modifications in the structure or genomic organization of immunoglobulin light chain genes.

Autoimmune MRL-1 pr/1pr mice are an animal model for the secondary antiphospholipid syndrome.

Patients with systemic lupus erythematosus often develop thrombosis, recurrent abortion, thrombocytopenia and neurological disease, in association with serum antibodies against phospholipids, termed the secondary antiphospholipid syndrome. In our search for an animal model for the human antiphospholipid syndrome, we examined the MRL-1pr/pr mouse, which develops a lupus-like illness. These mice were found to have significantly elevated levels of serum antiphospholipid antibodies (aPL) and thrombocytopenia compared to normal mice. By immunodiffusion, these antibodies are similar to human aPL. We observed histologic evidence of central nervous system thromboses as well as perivascular infiltrates of the choroid plexus. The high titers of serum aPL, thrombocytopenia, and thrombosis establish this strain as an animal model for the secondary antiphospholipid syndrome.