LETTERS 1389 azathioprine. Arthritis Rheum 24: 1453-1454, 1981). Hypotension and oliguria developed in 2 of their patients suffering from severe inflammatory arthritis, when the drug was administered to them a second time. Leukocytosis was present, and serum creatinine rose. Both patients had previously experienced a hypersensitivity reaction with fever and leukocytosis in response to azathioprine. We have observed a patient with quite similar clinical and laboratory features. In 1979 a 27-year-old man was diagnosed as having systemic lupus erythematosus. Symptoms were well controlled by corticosteroids (methylprednisolone 60 mgiday). A trial of azathioprine treatment (1.5 mgl kg/day) caused gastrointestinal upset (nausea and vomiting) that disappeared shortly after the patient stopped taking this drug. During a 2-year followup, the dose of methylprednisolone was maintained at 8-12 mg/day. In July 1981 because of osteonecrosis of the right femoral head, we again tried azathioprine (50 mg/day) for the patient. After 4 days, general malaise and gastrointestinal trouble abruptly reappeared; within a few hours, angioneurotic edema, hypotension (80155 mmHg), and oliguria were evident. The serum creatinine rose from 1.2 to 5 . 2 mg/dl; uric acid rose to 9.1 mg/ dl; and blood urea nitrogen rose to 47 mg/dl. The central venous pressure (CVP) was 2.5 cm H20. Hemodynamic acute renal failure was diagnosed. With the prompt withdrawal of azathioprine and intravenous administration of hydrocortisone and fluids, the symptoms disappeared. Renal function returned to normal during the next 6 days. CVP increased to 8 cm HzO. In this case and those reported by Keystone and Schabas, hypotension and oliguria are shown to be rare but serious side efects of azathioprine therapy. Although the cause of this reaction is unknown, it is probably a manifestation of hypersensitivity to the drug or one of its metabolites. In a recent review azathioprine is included among the drugs responsible for immunologically mediated toxic acute interstitial nephritis (Kleinknecht D, Kanfer A, Morel-Maroger L, Mery JPh: Immunologically mediated drug-induced acute renal failure. Contr Nephrol 10:42-52, 1978). From our experience, we suggest that symptoms such as fever, chills, skin rash, and minor gastrointestinal disorders (vomiting, nausea, or diarrhea) can be indicative of hypersensitivity to azathioprine. For patients who experience these symptoms when taking azathioprine, the physician should stop the treatment and should not attempt further azathioprine therapy. Francesco Trotta German0 Menegale Orazio Fiocchi Division of Rheumatology Division of Nephrology St. Anne Hospital 44100 Ferrara, Italy HLA specificities and psoriatic arthritis To the Editor: The existence of an association between HLA specificities and psoriatic arthritis has been widely investigated (1,2), but most studies have been limited to patients and controls. The few family studies have been limited to relatives already known to have these diseases. In our recent personal survey of 113 families of psoriatic probands (3), we attempted to determine the clinical and genetic relationship between psoriasis and psoriatic arthropathy. We found that 1) history alone identifies affected relatives in 20% of the families, but direct examination reveals affected relatives in 68% of the families and 2) psoriasis, psoriatic arthropathy, or both are found in the families of probands with either psoriasis or psoriatic arthropathy, but the distribution is different in each generation. Family studies including HLA typing are certainly needed to provide more information on these relationships. Claudio Cervini, MD Walter Grassi, MD School of Medicine University of Ancona, Italy 1 . Lambert JR, Wright V , Rajah SM, Moll JMH: Histocompatibility antigens in psoriatic arthritis. Ann Rheum Dis 35:526, 1976 2. Buchanan R, Kraag G, Rosenthal D, Siangal DP: HLA antigens and psoriatic arthritis, Transplant Proc 9: 1873, 1977 3. Cervini C, Grassi W, Milani-Comparetti M: Psoriasi ed artropatie: osservazioni genealogiche. Ann Reumatol 12: 18, 1979 Methodology errors in fibrositis study To the Editor: In a recent issue of Arthritis and Rheumatism, Payne et a1 ( I ) reported a study of psychologic disturbances in patients with fibrositis. I congratulate you for your wisdom in publishing studies of such a prevailing problem as soft tissue rheumatism, and the authors of this article for having ventured into this dark quarter of rheumatic diseases. In a spirit of the search for truth, I would point out a methodological error in the study. A control group for this condition cannot be considered without considering the issue of pain per se. To control fibrositis in the typical cases of inflammatory rheumatism is to ignore what little we have learned about pain as a quantifiable clinical manifestation. The authors might see this if they were to do further parametric analyses of their controls versus fibrositis patients. For example, they may find that the mean MMPI scores of their rheumatoid arthritis patients who have equal suffering scores on the McGill questionnaire will be the same as the mean MMPI scores of those with fibrositis.