Ketoacidosis associated with protease inhibitor therapy

We describe a case of a ketoacidosis related to protease inhibitor therapy, which might be a result of the inhibition of the protease converting proinsulin to insulin. A 57-year-old African woman was hospitalized in November 1997 for ketoacidosis. She had been infected by HIV in 1983 by transfusion. She had no history of AIDS-related infection except oral candidiasis. Indinavir was introduced in February 1997. She had also been treated with a combination of lamivudine and stavudine since December 1996. Her viral load had been below the cut-off value (200 copies/ml) since March 1997 and her CD4 cell count was 273 × 106/l (11%). She was also treated with co-trimoxazole for primary Pneumocystis carinii prophylaxis and with acebutolol for arterial hypertension for several years. In November 1997, she was hospitalized for weight loss (7 kg in 1 month), polyuria, polydipsia, blurred vision and fatigue. On physical examination, her blood pressure was 10/6 mmHg, her temperature was 38°C, and she was dehydrated and weak. No bacterial infection was noted. Serum glucose level was 900 mg/dl, serum pH was 7.17, serum potassium was 5.9 mmol/l, and serum creatinine was 210 µmol/l. Urine ketone measurement was highly positive. Parenteral fluid therapy and intravenous insulin therapy were administered. Protease inhibitor and other antiretroviral therapy were stopped. Under treatment for ketoacidosis, her clinical and biological status improved promptly. The patient was initially insulin-resistant requiring 85 U insulin daily (weight, 55 kg) during the first week. Serum glucose was subsequently controlled with 52 U intermediate-acting insulin daily. Indinavir was reintroduced after 3 weeks of disease evolution. In this case, treatment with a protease inhibitor was followed after 8 months by new-onset severe diabetes. No predisposing condition has been found. She had no familial or personal history of diabetes. Her plasma glu-cose level was 1.01 g/l in July 1997. Birth weights of her six children were below 3 kg. There was no sign of neurological, nephrologic or ophthalmic complication of diabetes, and no other aetiological factor of this new-onset diabetes was found. There was no current opportunistic infection, and no other treatment she received was described to be responsible for this event. No pancreatitis was present (no abdominal pain, normal abdominal computed tomography scan, normal serum lipase level, normal serum cholesterol and triglycerides), and islet cell antibodies and insulin anti-bodies were negative. Diabetes mellitus in HIV infection [1] has been found relative to pentamidine use [2] and to didanosine [3]. Few cases of diabetes related to protease inhibitor use have been reported to the US Food and Drug Administration [4]. One case of severe diabetes occurred after 2 weeks of treatment [5], although this patient had a familial predisposition to diabetes, and protease inhibitor therapy was continued. Five cases of moderate hyperglycaemia under indinavir have also been reported [6]. These patients were treated by oral hypoglycaemic agents or diet modification. In one case, discontinuation of therapy resulted in correction of hyperglycaemia, followed by recurrence when therapy was challenged. To our knowledge, the case reported here is the first case of diabetes associated with indinavir complicated initially by ketoacidosis and partial insulin resistance. The mechanism of this side-effect might be the inhibition of the protease converting proinsulin to insulin. Both serum peptide C (0.24 mmol/l) and glucagon test showed low insulin secretion. The mechanism of the initial insulin resistance is not known. Patients under protease inhibitor therapy and their practitioners should be aware of this potentially severe adverse effect. Plasma glucose should be monitored, and weight loss, polyuria and polydipsia should be detected early. New studies are needed to determine whether protease inhibitor therapy should be pursued or discontinued in these cases.