Abstract

In a study of the clinical and immunogenetic profiles of 17 patients with rheumatoid arthritis and thrombocytopenia (platelet count less than 150 000/mm3 (150 x 10(9)/l)) due to gold therapy two clinical patterns were distinguished without knowledge of HLA type: group I, an early precipitous thrombocytopenia (10 patients), and group II, a less dramatic fluctuant fall (seven patients). In group I patients the clinical and laboratory features suggested an immune mediated, peripheral destruction of platelets, and all patients in this group were found to be HLA-DR3 positive. Two patients subsequently received penicillamine without toxicity. In group II the basis of thrombocytopenia appeared to be different, and only two patients in this group were HLA-DR3 positive. All group II patients had received penicillamine; four developed a thrombocytopenia. Mechanisms of toxicity in both groups are discussed. It would appear that HLA typing in unlikely to help in predicting all those patients at risk of toxicity during chrysotherapy.