Jules Levin, NATAP

Judith Currier reported data from this study (ACTG 362), in which median follow-up was 56 weeks, that no cases of MAC occurred for those receiving MAC prophylaxis and 2 cases occurred for those receiving no prophylaxis. As a result she said this suggests MAC prophylaxis is not needed in the population studied (no prior MAC disease) and discontinuing MAC prophylaxis does not pose a significant risk. A remaining question relates to viral load. CD4s increased to above 100 (235 median in MAC study) and 340 median in PCP study, but only 62% in MAC study and 80% in PCP study had undetectable viral load. If a person’s CD4s increase substantially due to HAART but viral load is high (for example- 20,000 copies/ml), is it safe to discontinue prophylaxis? Some doctors have expressed less confidence about this situation, but if CD4s are increased and viral load is undetectable they are more confident in discontinuing prophylaxis.

Weekly prophylaxis with azithromycin has been shown to reduce the risk of MAC in patients with advanced HIV when compared to placebo and directly to rifabutin. Subsequently the standard of care has been to recommend azithromycin or clarithromycin prophylaxis for individuals with CD4s < 50. Data from observational studies since 1996 suggest that rates of MAC have declined among patients receiving potent anti-retroviral therapy. This raises the question of whether MAC prophylaxis is needed when CD4s increase due to HAART. This study addresses the question.

The primary study objective was to compare the time to development of MAC disease in subjects receiving azithromycin to those receiving placebo. This is a randomized, double-blinded, placebo controlled study of 643 participants with no prior MAC disease who were receiving antiretroviral therapy. To be eligible for the study people had to have at some time in past documented increase in CD4 from < 50 to > 100 on two occasions, at least 4 weeks apart. Patients were equally randomized to azithromycin 1200 mg/week or placebo.

Every 8 weeks in study clinical evaluations and MAC blood cultures were performed. HIV RNA PCR (Amplicor) was done every 16 weeks. CD4 every 8 weeks. A patient could re-start open label azithromycin if their CD4s went below 50. The median follow-up at this point is 56 weeks.

Baseline—There were no major differences between the two groups. 36% in each group (AZ or placebo) had no prior prophylaxis. 45% had prior MAC prophylaxis. The nadir (lowest prior) CD4 count was 21 in both groups. The median baseline CD4 was 235 in AZ arm and 219 in Placebo arm. About 1/3 had between 100 and 199 CD4s at baseline. At baseline the median CD4s were broken down to 34% and 40% had between 100 and 199 CD4s in AZ and placebo arms, respectively. 33% had between 200-299 in both groups. About 23% had between 300-500 CD4s.

At baseline, 62% in the AZ arm and 64% in the placebo arm had <500 copies/ml; 19% and 15% had, respectively, had 500-20,000 copies/ml; 12% and 18% had >20,000 copies/ml.

18.6% discontinued from the AZ group and 17% discontinued from the placebo group. No real difference. But a higher proportion of the discontinuations due to toxicity were in the AZ group. Of note there were less grade 3/4 side effects (14.6% vs 20.2%) in the AZ arm.

There were no cases of MAC that occurred in the study in the azithromycin arm (n=322) and 2 cases in the placebo arm (n=321).

Both people who developed MAC were responding well to HAART with <500 copies/ml. One person’s CD4s were 306 when MAC was seen, while their CD4 nadir was < 20 and had previously been on prophylaxis; of note the patient had no systemic signs or symptoms. They had a history of PCP. The other’s was 203 when MAC developed. Their CD4 nadir was <10 CD4s.

There will be continued follow-up to monitor risk over longer term. So far at 96 week median CD4s are increasing in population and viral loads remain steady.

J.C. Lopez spoke for a Spanisg research group in reporting the data from this study. This was an open-label randomized study where patients were assigned to withdraw from or to continue ongoing PCP prophylaxis. Patients were included if they had <200 CD4s previously or previous PCP episode, receiving ongoing PCP prophylaxis and having a sustained response to HAART (including PI): CD4s >200 and viral load <5,000 copies/ml for at least 3 months. People who discontinued HAART, had two CD4s counts less than 200, or developed an AIDS defining event were off study.

At baseline, the no PCP prophylaxis group (n=247) had CD4s <200 for average 33 months, a CD4 nadir of 102 and 24 previously had PCP. In the arm continuing PCP prophylaxis (n=241) they had CD4s < 200 for 32 months, nadir CD4 of 90, and 18 had prior PCP.

When study started CD4s were 340 on average in the no prophylaxis arm and 330 in those receiving prophylaxis. 38% in the no prophylaxis arm and 35% in the prophylaxis arm had 200-300 CD4s. About 80% in each arm had <500 copies/ml. And all were on PCP prophylaxis about 34 months. They were on HAART about 15 months. About 30% of all patients. About 90% of patients were taking TMP-SMX. About 80% were taking 3TC, 65% d4T, and 61-67% were taking indinavir.

The median follow-up was 11 months but 20% of patients had >15 months. 15 patients were taken out of study: 6 in no proph. group and 9 in proph. group; 13 because their CD4 fell below 200: 5 in no proph. arm and 8 in proph. arm. One in each arm discontinued HAART. 5 in the no proph arm and 3 in proph arm were lost to follow-up.

After 9 months, CD4s were 411 and 425 in the no prop and proph arms, respectively. The percentages below 500 were 79% and 72% after 9 months in the no proph and proph arms, respectively. 29 patients in the no proph and 27 in the proph group changed their HAART regimens. 7 in no proph and 8 in proph arms had bacterial infections. There were no AIDS defining events.

So far there have been NO CASES of PCP in either group.