Two recent double-blind, randomized trials suggest that megestrol is effective for AIDS-related cachexia [1, 2]. Von Roenn and colleagues [1] concluded that megestrol should be considered for all patients with AIDS and a persistent weight loss of 5% or more of ideal body weight. This conclusion raises several concerns. The first centers on the persistence of objective and subjective improvement. Both studies were of 12 weeks' duration. In one [2], 52% of patients receiving megestrol discontinued treatment during the 12-week period. Comparable data were not provided by the other study. Neither study stated how many patients continued megestrol therapy beyond 12 weeks. Data from patients with cancer indicate that weight gain rapidly reverses after discontinuation of nutritional therapy [3]. Consequently, it is possible that megestrol treatment, unless continued, produces only transitory increases in weight and well-being.

The second issue concerns changes in body composition after megestrol treatment. In both studies, weight gain was primarily caused by increased fat storage. Lean body mass in patients receiving megestrol decreased in one study [2] and increased by 1.14 kg in the other [1]. Lean body mass is known to be an important predictor of mortality. In patients with AIDS, death occurs when lean body mass decreases to 54% of normal [4]. On the basis of these findings, it might be predicted that megestrol would not prolong survival; indeed, it did not.

The third issue concerns whether an increase in body fat, a major therapeutic effect of megestrol, is beneficial. Similarly, Murray and Murray [5] persuasively argued that starvation suppresses infection and that refeeding activates infection. Weight loss may be a host defense mechanism in AIDS. If so, interventions should be aimed at maintaining, but not increasing, weight. The fact that infectious complications did not increase in patients receiving megestrol may be related to the short duration of therapy.

Megestrol seems to be an expensive way to make patients feel better about their appearance, appetite, enjoyment of eating, and, to a lesser extent, quality of life. Whether maintenance of this benefit will require lifetime megestrol treatment is unknown, as is the tolerance of patients to long-term therapy. It therefore seems premature to recommend megestrol for patients with AIDS and significant weight loss.