Improved HIV Survival Indicates Effect of Therapy

On May 1, Johns Hopkins researchers published an epidemiological study of survival of persons who were HIV-positive but did not have any AIDS-defining conditions.(1) They found a significant increase in survival between the years before and the years after widespread availability of pneumocystis prophylaxis and antiviral monotherapy. Because this study compared the time periods of 1985-1988 with 1989-1993, it could only record effects of therapies in wide use in the years between 1989 and 1993 -- ignoring effects of more modern strategies like protease inhibitors, widespread combination antiretrovirals, and viral load tests. (Recent data could not be compared, since several years must elapse to provide accurate estimates of length of survival.)

For those without an AIDS-defining condition and without symptoms of AIDS but with a CD4 (T-helper) count between zero and 100, 2.5-year survival increased from 22% to 54% between the period 1985-1988 to 1989-1993. For those with a CD4 count of 101-200, survival increased from 53% to 71% between those two periods. For those with CD4 in the range of 201-350, the same 2.5-year survival increased from 83% to 91%.

For those with a CD4 count of over 350, the overall three-year survival rate was 94% or greater, and there were too few deaths to see if there was a difference from the earlier to the later period. However, those with CD4 counts over 350 who reported having certain AIDS-related symptoms (less than required for an AIDS diagnosis) had a 1.6 to 2.3 times higher relative hazard of dying than those who had had no such symptoms. (The presence of symptoms had a greater effect on relative hazard at higher CD4 counts than at lower counts, "indicating that the presence of clinical symptoms in the absence of AIDS was a stronger predictor among those with less compromised immune systems.")

This study could not compare the changing risk between the calendar periods for those who started with an AIDS diagnosis, because the data used came from the MACS study (Multicenter AIDS Cohort Study), which enrolled 4,954 gay men in four U.S. cities (both HIV-positive and HIV-negative) in 1984-1985. The MACS entry criteria excluded those who already had AIDS; therefore, the small number who had developed AIDS by the period of 1985-1988 would not have been representative. This study did find that the presence of AIDS (by the old definition, which required at least one opportunistic infection or condition) increased the risk of death by about two to six times, compared to those with the same CD4 count but without AIDS and without other symptoms; however, the data available did not allow accurate estimation of this increase in risk.

This study has limitations. For example, the researchers indicated that there were differences between the earlier and later cohorts which may not have been completely adjusted for by the statistical methods used; therefore, certain biases may have accounted for some of the survival differences. Also, while the two calendar-year cohorts were selected to show a before-therapy to after-therapy change, not everybody in the later group used treatment, and some in the earlier group did, suggesting the possibility that the improved survival due to the treatments available in 1989-1993 could have been greater than this study indicated.

Comment

While the two cohorts in this study were selected to show differences due to pneumocystis prophylaxis and early antiviral therapy, there have also been many other improvements in medical management, especially as physicians have learned more about how to diagnose and treat patients with opportunistic infections. This improved medical management is harder to attribute to particular calendar years, but it would also have contributed to the survival improvement found in this study.

This study suggests that persons without AIDS and with high CD4 counts should also look at AIDS-related symptoms as a factor in planning therapy; these indicated a worse prognosis, and therefore would seem to argue for a somewhat more aggressive approach to therapy than for asymptomatics. The symptoms recorded in this study were "an unintentional weight loss of at least 4.5 kg since the previous visit [visits were every six months], presence of fatigue for at least two weeks, oral thrush or candidiasis for at least two weeks, persistent fever higher than 37.9 degrees C for at least two weeks, and diarrhea for at least two weeks."

References

1. Enger C, Graham N, Peng Y, and others. Survival from Early, Intermediate, and Late Stages of HIV Infection. JAMA. May 1, 1996; volume 275, number 17, pages 1329-1334.