Megestrol acetate, a synthetic orally active progestational agent used widely for the treatment of metastatic breast cancer, has been reported to stimulate appetite and weight gain. Using conventional doses (160 mg/d), approximately 30% of patients with breast cancer gained weight [6]. In patients with hormone-insensitive tumors, megestrol acetate (160 mg/d) induced a similar degree of weight gain [7]. A phase I to II study [8] of high-dose megestrol acetate (480 to 1600 mg/d) for the treatment of advanced breast cancer reported marked appetite stimulation, and 81% of patients had weight gain of more than 2 kg. Several controlled randomized studies [9-11] have subsequently shown the benefit of megestrol acetate for treating patients with cancer-related cachexia.

On the basis of these data, a pilot study of megestrol acetate for the treatment of HIV-related cachexia was initiated in 1987 [12]. Twenty-one of 22 patients with HIV infection and cachexia who were treated with megestrol acetate (320 to 640 mg/d) gained weight. Because of its perceived appetite-enhancing effect and excellent tolerability, we did a double-blind, randomized, placebo-controlled trial of megestrol acetate in patients with AIDS-associated weight loss.

Methods

Top Methods Results Discussion References

Study Design

Our multicenter, double-blind, placebo-controlled study was designed to assess the safety and efficacy of megestrol acetate (Megace; Bristol-Myers Oncology Division, Princeton, New Jersey) for the treatment of AIDS-related anorexia and cachexia. Persons included were 18 years of age or older, had an AIDS diagnosis [13], a life expectancy of 20 weeks or more, weight loss perceived as a detriment to well-being, anorexia, a Karnofsky performance status of 50% or more, and a history of clinically significant weight loss. Clinically significant weight loss was defined as a decrease of 20% from usual body weight, or as 10% below ideal body weight for patients whose premorbid weight was greater than ideal body weight, or as a loss of 10% or more of usual body weight for those whose premorbid weight was below ideal body weight. Exclusion criteria were difficulty in swallowing, impaired digestive or absorptive function, contraindications to high-dose megestrol acetate (poorly controlled hypertension, heart failure, or deep venous thrombosis), dementia, a history of substance abuse with questionable current or future abstinence, severe diarrhea despite symptomatic treatment (defined as 5 or more watery stools per day for at least 7 days), a history of steroid or marijuana use in the 2 months before the study, previous use of megestrol acetate, active uncontrolled systemic infections, disease exacerbation associated with weight loss within the 2 weeks before study enrollment, or the presence of ascites or pleural effusions. Women who were pregnant or menstruating were also excluded. Patients were not started on a new antiviral therapy that might affect appetite or weight (or both) during the 3-month study period. The baseline evaluation included a history and physical examination, complete blood counts, laboratory chemistry tests, lymphocyte phenotyping, and urine analysis. Laboratory evaluations were done at monthly intervals. Delayed hypersensitivity reactions were measured using the Merieux Institute Antigen Reactive Skin Test (Multitest-CMI).

Patients were randomly assigned in an unbalanced randomization (1:2:2:2) to receive placebo or megestrol acetate in an oral suspension of 100 mg, 400 mg, or 800 mg/d. Patients with stable weight or excessive weight gain could be removed from the study after completing the 12-week trial period. Patients otherwise continued on their assigned treatment as long as they did not have additional weight loss of more than 10% of their baseline body weight; development of disabling or life-threatening toxicity; initiation of treatment designed to affect HIV replication or expected to affect weight, appetite, or quality of life; or interruption of planned therapy for more than 2 weeks.

The primary efficacy criterion was weight gain. Weight measurements were made at 4-week intervals on a standardized balance scale. Serial anthropometric measurements were used to estimate changes in lean body mass and body fat. Lean body mass and total body water were estimated by bioelectrical impedance analysis using standard tetrapolar methods [14, 15]. The degree of anorexia and its change from baseline was assessed every 4 weeks by 3-day caloric and protein intake diaries, an appetite assessment scale, and a questionnaire of factors affecting food intake. A linear analog self-assessment questionnaire evaluated patients' perceived benefit from megestrol acetate treatment.

Biostatistical Methods and Analysis

To be evaluable for the analysis of efficacy, patients must have met all inclusion and exclusion criteria and have had no substantial protocol violations. Patients must also have had a baseline weight measurement and at least two additional weight measurements during a 12-week period. Patients who had a baseline evaluation and one additional evaluation during therapy but who dropped out of the study because of therapeutic failure (for example, weight loss, worsening of condition because of infection, death, or toxicity) were also considered evaluable for the analysis of efficacy. An intent-to-treat analysis included all patients who took study medication and had at least one follow-up visit. This analysis included the following variables: weight, weight change, and nutritional variables.

All statistical tests were two-sided. All values are reported as mean (±SE) except where otherwise indicated. For efficacy variables, a significance level of P < 0.05 was used for the primary comparison of placebo with 800 mg of megestrol acetate. Additional pairwise comparisons were done using a Bonferroni- adjusted significance level of 0.001 [16]. For safety measures, a significance level of P < 0.10 was used. All statistical methods were adjusted for site where possible. Sites with nine or fewer evaluable patients were combined for the statistical analysis. All computations were done using SAS statistical software 6.06.

The Cochran-Mantel-Haenszel statistic, chi-square statistic, and the Fisher exact test were used for the analysis of categorical variables [17]. Linear models containing effects for treatment group, site, and treatment-by-site interaction were used for the analysis of continuous variables [18].

The Benard van Elteren test [18] was used to compare the placebo recipients and patients receiving 800 mg of megestrol acetate with respect to maximum weight change. Maximum weight change was also analyzed using a dose-response model. In this analysis, maximum weight change was expressed as a linear function of the dose of megestrol acetate, allowing the effect of other variables on weight change to be measured. The model contained effects for dose, site, baseline weight, and age.

The correlation between maximum weight change and well-being was assessed by calculating the Spearman correlation coefficient for each question [17].

The number of patients reporting adverse events was summarized by body system and treatment group using the COSTART term. Treatment groups were compared with respect to each adverse event reported by 10% or more of all the patients.

Results

Top Methods Results Discussion References

Patient Characteristics

Consecutive, ambulatory patients with AIDS-related weight loss who met all of the study eligibility criteria were enrolled in this trial. A total of 271 patients were randomly assigned, of whom 270 were evaluable for safety and 195 were evaluable for efficacy. The 75 patients who were not evaluable for the efficacy analysis either did not meet the premorbid weight loss requirement (27 patients), had no follow-up visits (46 patients), or had only one follow-up visit (7 patients). Some patients did not meet premorbid weight loss criteria and had insufficient follow-up for evaluation for the efficacy analysis (5 patients). No statistically significant differences were noted between treatment groups with respect to the percentage of evaluable patients. Characteristics of all patients who received study medication are summarized in Table 1. No statistically significant differences were noted between treatment groups with respect to sex, race, mean age, mean prestudy weight loss, individual patient weight loss, performance status, baseline caloric intake, history of opportunistic infections or types of infections, use of previous antiretroviral therapy, number or type of other prophylactic medicines, baseline lymphocyte counts, or skin reactivity tests. Patient assessment of well-being at baseline showed no statistically significant differences between treatment groups with respect to any of the questions posed.

Weight Change

Mean weight change from baseline to the last evaluation for the evaluable patients during the 12 weeks of study was –0.726,0.863, 1.91, 3.54 kg ( –1.6,+1.9, +4.2, and +7.8 pounds) in the placebo, 100-mg, 400-mg, and the 800-mg megestrol acetate groups, respectively (Table 2). A difference was noted between the group receiving 800 mg of megestrol acetate and the group receiving placebo (P < 0.001) with respect to maximum weight change during 12 weeks of treatment. A difference was observed in percent weight change from baseline to time of maximum weight change in the 12 study weeks (P < 0.001). The percentage of patients with a maximum weight gain of 2.27 kg (5 pounds) or more from baseline to last evaluation was 21.4%, 44.3%, 56.6%, and 64.2% in the placebo, 100-mg, 400-mg, and 800-mg groups, respectively. A weight gain of 2.27 kg (5 pounds) or more was observed in a greater percentage of patients receiving 800 mg of megestrol acetate when compared with placebo recipients (P < 0.001) (Figure 1).

View larger version (17K): [in this window] [in a new window]   Figure 1. Weight gain of patients during the first 12 weeks of the study. A greater percentage (64.2%) of patients in the group receiving 800 mg of megestrol acetate than in the placebo group (21.4%) had a maximum weight gain of at least 2.27 kg (5 pounds) (P < 0.001). Mean values with 95% CIs (bars) are shown.

The intent-to-treat analysis (224 patients) was similar to the analysis of evaluable patients for changes in mean weight and percentages of patients gaining 2.27 kg (5 pounds) or more. Twenty-five percent (8 of 32) of patients receiving placebo gained 2.27 kg (5 pounds) or more compared with 62.3% (38 of 61) of patients receiving 800 mg of megestrol acetate. Differences for the intent-to-treat analysis were observed between groups receiving placebo and 400 mg of megestrol acetate (P = 0.005) and between groups receiving placebo and 800 mg of megestrol acetate (P = 0.002) with regard to the number of patients who gained 2.27 kg (5 pounds) or more (Table 3).

A dose-response model was fitted to the maximum weight change with factors for dose (mg), site, baseline weight (mean overall baseline weight, 132.26 pounds), age (mean overall age, 39.01 years), dose x site, and dose x age (mean overall age, 39.01 years). Patients receiving megestrol acetate had an increased weight gain in a dose-dependent manner (P < 0.001; regression coefficient, 0.011). An additional dose-response analysis was done substituting the ratio of dose to baseline weight (mg/lb) for dose. This model had factors for dose/baseline weight, site, baseline weight/age, dose/weight, site and dose/weight, and age. For each patient, dose was divided by the baseline weight, and this new variable was included in the model in place of dose. Dose correlated with baseline weight (regression coefficient, P < 0.001); the maximum weight change increased as the dose-to-baseline weight ratio increased. These analyses were also done on percentage maximum weight change, and similar results were seen.

Time to Maximum Weight Change

In patients treated with 800 mg of megestrol acetate, the time to maximum weight change occurred after 2 to 6 weeks in 26.4% of patients, after 6 to 10 weeks in 39.6% of patients, and after more than 10 weeks in 34% of patients. In the placebo group, the time to maximum weight change occurred after 2.6 weeks in 35.7% of patients, after 6 to 10 weeks in 39.3%, and after 10 weeks in 25%. The differences between the treatment groups were not statistically significant with respect to this variable.

Bioelectrical Impedance Analysis

This analysis was done to define changes in body composition responsible for weight change. Lean body mass from baseline to last evaluation increased across the treatment groups from lowest with placebo to highest with 800 mg of megestrol acetate. The mean change in lean body mass from baseline to last evaluation was 1.14 kg (2.5 lb) in the 800-mg megestrol acetate group, 0.68 kg (1.5 lb) in the 400-mg group, –0.14 kg ( –0.3 lb) in the 100-mg group, and –0.772 kg ( –1.7 lb) in the placebo group. The difference from baseline to last evaluation between treatment with 800 mg of megestrol acetate compared with placebo (P < 0.001) was significant. No other pair-wise comparisons were statistically significant. An analysis of changes in total body water with therapy showed no statistically significant changes between baseline and last evaluation during the 12 study weeks and time of maximum weight change.

Anthropometric Measurements

Mid-arm circumference decreased 0.8 cm from baseline to last evaluation in the placebo group. This decrease was attenuated by megestrol acetate therapy. A mean increase in mid-arm circumference of 0.9 cm was observed in patients receiving megestrol acetate, 800 mg/d. More dramatic trends were seen in triceps skinfold results. Megestrol acetate treatment prevented the decrease of triceps skinfold thickness in patients receiving placebo and resulted in an increase in all doses tested. Mean changes in total body muscle mass at last evaluation were an increase of 1 kg in the 800-mg megestrol acetate group and decreases of 0.8 kg, 1.1 kg, and 0.6 kg in the placebo, 100-mg, and 400-mg megestrol acetate groups, respectively. The difference between the 800-mg megestrol acetate group and the placebo group was significant (P = 0.001). Differences were also seen between the 800-mg compared with 100-mg megestrol acetate groups (P = 0.006) and the 800-mg compared with 400-mg megestrol acetate (P = 0.005) groups.

Perception of Well-Being Measurements

The patient's assessment of the effect of therapy on their overall sense of well-being was evaluated using a nine-item linear analog self-assessment questionnaire. Questions addressed the effect of therapy on weight change (questions 1 and 2), appearance (questions 3 and 4), appetite (questions 5 and 6), and overall perception of the effect of the study treatment (questions 7, 8, and 9). For every question, greater improvement in the patient's perception of well-being was reported by patients receiving 800 mg of megestrol acetate when compared with placebo (P 0.033). In fact, at the time of last evaluation, patients reported a dose-related increase in the degree of improvement. Figure 2 highlights the relation among megestrol acetate therapy, dose, and effect of therapy on overall sense of well-being. Compared with patients in the placebo group, patients in the 800-mg megestrol acetate group felt that change in weight had a substantial positive effect on their health, patients were less concerned about their weight than when they started treatment, and they felt better overall.

View larger version (18K): [in this window] [in a new window]   Figure 2. Well-being questionnaire responses. Patient responses to representative questions from the self-assessment questionnaire to evaluate the effect of therapy on perception of well-being are charted by group. For every question, significantly greater improvement in the patient's perception of well-being was reported by patients receiving 800 mg of megestrol acetate when compared with those receiving placebo. The degree of improvement reported by patients increased as the dose of megestrol acetate increased. There was a statistically significant quadratic trend (P < 0.001) in the mean patient well-being scores as the dose increased from 0 to 800 mg of megestrol acetate for each question presented. Mean values are presented with bars for 95% CIs.

Pearson correlation coefficients were calculated between actual weight perception and each question about perception of well-being at the last evaluation for all patients evaluable for efficacy during the 12 study weeks. Actual weight at last evaluation correlated with question 1 (P = 0.045) about the perception of the treatment's effect on health and correlated with questions 3 and 4 about the extent the patient's appearance had changed according to self-report and the report of others (P = 0.002 and P = 0.034), respectively. Actual weight at last evaluation also correlated with question 7 about whether the patient felt better or worse overall (P = 0.022).

Caloric Intake and Appetite

A difference in the change in daily caloric intake from baseline to time of maximum weight change was seen in the 800-mg megestrol acetate group when compared with the placebo group (P = 0.001). No other pair-wise comparisons were statistically significant. At baseline evaluation, the mean caloric intake was 1920, 1993, 2048, and 2139 calories for the placebo, 100-mg, 400-mg, and 800-mg groups, respectively. No statistically significant differences were noted among treatment groups with respect to baseline caloric intake. The mean changes in caloric intake were 645.6, 307.5, 326.2, and –107 calories in the 800-mg, 400-mg, 100-mg megestrol acetate, and placebo groups, respectively.

Patient-reported improvement in appetite grade at the time of maximum weight change was noted in 92.5% of the patients treated with 800 mg of megestrol acetate, 71.7% of those in the 400-mg group, 70.5% in the 100-mg group, and 50% in the placebo group. At the time of maximum weight change, a greater improvement in appetite was noted in the 800-mg megestrol acetate group compared with the placebo group (P < 0.001).

Safety

All 270 patients who received at least one dose of study medication were evaluable for the analysis of safety. No major drug-related safety issues were observed during this study. A total of 359 adverse events were reported by 138 patients during the 12 weeks of the study. No statistically significant differences were reported among study groups with regard to the number of patients reporting at least one adverse event. Fifty-six percent of adverse events reported were not, or probably not, related to the study drug as assessed by the investigators. Symptoms or adverse events were evaluated at 4-week intervals. The most commonly reported symptoms that developed or worsened during the study in decreasing frequency were cough, fever, fatigue, dyspnea, diarrhea, and weakness. No statistically significant differences were seen among treatment groups with regard to severity of adverse events. Eighty-one percent of adverse events reported in all treatment groups were mild to moderate in severity. Fifty-one severe and 7 life-threatening events were reported by 41 patients; 4 events in the placebo group, 27 events in the 100-mg megestrol acetate group, 24 events in the 400-mg group, and 10 events in the 800-mg group. Deep venous thrombosis was observed in 1 patient in the 400-mg megestrol acetate group. The percentage of patients reporting impotence increased with dose (2.6% of patients receiving placebo, 3.7% receiving 100 mg of megestrol acetate, 5.3% receiving 400 mg, and 12% receiving 800 mg). Differences were seen among treatment groups with regard to edema (P = 0.042). Edema occurred or worsened in 10.5% of placebo recipients and in 4.9%, 12%, and 1.3% of patients treated with 100 mg, 400 mg, and 800 mg of megestrol acetate, respectively.

No statistically significant differences among treatment groups with regard to abnormal laboratory test results, lymphocyte counts, or lymphocyte subsets were found. No statistically significant differences were found between placebo and the megestrol acetate treatment groups with respect to changes in skin reactivity.

No statistically significant difference among study groups was seen at baseline or at last evaluation with regard to changes in Karnofsky performance status.

No statistically significant differences were seen among treatment groups with regard to the percentage of patients who developed opportunistic infections. Of the 270 patients evaluable for safety, 225 (83%) patients did not develop an opportunistic infection during the study. A total of 45 patients developed opportunistic infections during the study; 7 (18.4%) in the placebo group and 10 (12.2%), 17 (22.7%), and 11 (14.7%) in the 100-mg, 400-mg, and 800-mg megestrol acetate groups, respectively. The most common infection that developed during the study was Candida infection, which occurred in a total of 19 patients (7%) (Table 4).

Seventeen patients died of an AIDS-related disease or complication during this study. Three other patient deaths occurred during the study (1, cause unknown; 1, anaphylactic reaction to amikacin; and 1, automobile accident). These 20 patients were distributed by dose as follows: 1 patient in the placebo group (2.6%, 1 of 38), 7 patients in the 100-mg megestrol acetate group (8.4%, 7 of 83), 8 patients in the 400-mg group (10.6%, 8 of 75), and 4 patients in the 800-mg group (5.3%, 4 of 75).

To evaluate the potential effect of megestrol acetate on the course of HIV infection and survival, a follow-up search of patients who participated in the trial was done to specifically assess survival by treatment group. Investigators were notified by mail of the patients entered at their site for whom dates of death were not recorded in the database. They were asked to review clinical records and contact referring physicians for additional follow-up or death (or both) of those patients for whom data were censored. A date of death was recorded for a patient only if the date was known through medical records or a death certificate. Pracon (Reston, Virginia) was retained to contact sites about the vital status of those patients who were lost to follow-up. For the patients lost to follow-up, a date of death was requested through the National Social Security Database, which is current through December 1992. For names with social security numbers that did not have an associated date of death through this database search, patients were censored at the date that was last recorded in the clinical database. Seventeen patients did not have a date of death, and their social security numbers could not be located by investigators.

Patient survival from the start of the study, as estimated by the Kaplan-Meier method, showed no overall significant difference (P = 0.47) in the median survival of 32 weeks, 49 weeks, 36 weeks, and 39 weeks for the placebo and 100-mg, 400-mg, and 800-mg megestrol acetate groups, respectively. Median survival in patients treated with megestrol acetate is equivalent to or longer than median survival in patients treated with placebo.

Discussion

Top Methods Results Discussion References

This randomized, placebo-controlled trial shows the efficacy of megestrol acetate for the treatment of AIDS-related anorexia and cachexia. Sixty-four percent of patients treated with megestrol acetate, 800 mg/d, gained more than 2.27 kg (5 pounds) when compared with 21.4% of patients receiving placebo. None of the patients in the placebo group had weight gain more than 6.8 kg (15 pounds) when compared with 22.6% of patients in the 800-mg megestrol acetate group. Similarly, Loprinzi and colleagues [11], in a randomized trial of 800 mg of megestrol acetate compared with placebo in patients with advanced cancer who had cachexia, reported weight gain of 6.8 kg (15 pounds) or more over baseline levels in 16% of patients receiving megestrol acetate compared with 2% of patients receiving placebo.

The effect of megestrol acetate on the quality of life of patients enrolled in this trial is an important aspect of the cost–benefit ratio of megestrol acetate therapy for patients with AIDS-related cachexia. Patients consistently evaluated therapy as having a positive effect not only on their weight but on appetite, appearance, and overall sense of well-being. Further, the degree of improvement in these aspects of patient perceived sense of well-being increased as the megestrol acetate dose increased. One could argue that the subjective benefit of therapy is of greater value than the more traditional objective end points of food intake and weight gain.

Although numerous investigators have suggested that megestrol acetate results in nonfluid weight gain as a result of appetite stimulation, careful evaluation of changes in body composition and anthropometric measures have not been routinely done [7, 11]. Bruera and colleagues [9], in a double-blind crossover trial of 480 mg of megestrol acetate compared with placebo in malnourished patients with advanced non-hormone-responsive tumors, noted statistically significant increases in weight, triceps skinfold measurements, calf circumference, and caloric intake in patients treated with megestrol acetate. In the current investigation, the bioelectric impedance analysis data, anthropometric measurements, and caloric intake diaries all support the presence of nonfluid weight gain in patients treated with megestrol acetate. Lean body mass increased with megestrol acetate therapy, whereas total body water did not change, implying an increase in cellular body mass. Further support for the hypothesis that the weight gain with megestrol acetate therapy is nonfluid is the absence of a statistically significant increase in edema in those patients receiving megestrol acetate when compared with those receiving placebo.

The current study was not designed to assess the effect of megestrol acetate therapy on survival or immune function in patients with AIDS and weight loss. It has been suggested, on a theoretical basis, that megestrol acetate might up-regulate the expression of HIV and lead to more rapid progression of the underlying HIV infection [19]. No data from the current study suggest this. Twenty patient deaths occurred during the study. Seventeen of the 20 deaths were caused by AIDS-related complications. No significant differences were noted in the number of deaths, the frequency of opportunistic infections, or overall survival on the basis of treatment group.

Despite advances in antiretroviral therapy and treatment of opportunistic infections, progressive wasting remains a major problem in patients with AIDS. Megestrol acetate therapy leads to improved appetite, sense of well-being, and weight gain for most patients treated. No significant toxic reaction or adverse effect on the course of HIV infection was shown with therapy. Megestrol acetate should be considered for all patients with persistent weight loss of 5% or more of ideal body weight, independent of immune status.