Does PCP Prophylaxis Improve Survival in Patients With HIV Infection?
New Data Suggest New Approaches to Clinical Management
Primary and secondary prophylaxis against Pneumocystis carinii pneumonia have achieved unqualified success in reducing the morbidity and mortality that were once associated with episodes of AIDS-related PCP infection.(1-6) Recent data show that, in those who tolerate the drugs, trimethoprim-sulfamethoxazole is superior to dapsone -- which in turn is superior to aerosolized pentamidine.(7-12)
However, despite these clear-cut therapeutic successes there is still some controversy about whether PCP prophylaxis -- which unquestionably delays the onset of pneumocystis in patients with advanced HIV infection -- also significantly improves survival. Several observational studies have shown survival benefits with PCP prophylaxis,(3, 13-16) but with one notable exception -- a study by Fischl et al. that was published in 1988(17) -- survival benefits have not been seen in clinical trials of PCP prophylaxis regimens.(4-6, 8-12)
There appear to be several potential explanations for this dichotomy. The first is that clinical trials often have low death rates from PCP due to limited sample size and follow-up. In almost every trial of PCP prophylaxis conducted to date, the primary endpoint measured was an episode of pneumocystis, rather than survival. As a result, most of these studies were underpowered to detect the impact of prophylaxis on death from PCP.
This phenomenon is shown most clearly in two early studies of PCP prophylaxis. In a randomized study of three doses of aerosolized pentamidine (AP) -- 300 mg monthly, 150 mg biweekly, and 30 mg biweekly -- as secondary PCP prophylaxis, there were 101 PCP episodes available for analysis but only 19 PCP-related deaths.(6) This represents an 80% reduction in power. Not surprisingly, given this study's lack of power for measuring survival endpoints, the higher dose of AP reduced the risk of PCP but was not associated with improved survival.
Girard et al. reported similar results in a smaller randomized study, but in that study only one PCP related death was observed.(4) By contrast, the power to detect survival differences for PCP prophylaxis is significantly greater in cohort studies -- where development of PCP is not necessarily a primary endpoint and studies are continued irrespective of outcome.(3, 13-16)
Another potential explanation for this dichotomy in survival benefit -- seen when the results of observational studies are compared with the results of clinical studies -- is that most clinical trials of PCP prophylaxis compare two or more effective regimens that have similar effects on survival. The majority of recent PCP prophylaxis trials have compared at least two proven prophylaxis regimens (Table 1), and unlike earlier randomized trials and recent observational studies (Table 2) these recent prophylaxis trials have not had a placebo or untreated arm.(6-12, 18-23) Some clinical studies have even tested the benefits of a single prophylaxis regimen, with no comparison arm of any kind.(24-26)
It should come as no surprise that the studies listed in Table 1 show no survival differences, since all prophylaxis regimens are likely to have similar effects on survival and none of these comparison studies have been large enough to detect small differences in survival effect. What is more, in all of the studies which show that oral agents are superior to AP for preventing PCP, the oral agents have also had significantly higher rates of adverse events, events that required either cessation of therapy of crossover to other regimens.(8-12) The high crossover rates noted in these studies undoubtedly contributed to their failure to detect a difference in survival by study regimen.
The impact of earlier and more aggressive treatment
As a rule, observational studies monitor subjects much less closely than do clinical trials. The former often depend on semiannual follow-up visits, whereas most clinical trials typically follow enrolled patients on a monthly or even biweekly basis. Surveillance is much less aggressive in observational studies, and this has an impact on both the diagnosis and outcome of episodes of PCP. In observational studies, subjects often present only after their symptoms are well established; in clinical trials, on the other hand, symptoms of PCP are likely to be detected early and treated promptly.
The result is that episodes of PCP, when they occur in clinical trial cohorts, are usually diagnosed and treated before they become severe, let alone fatal. In community settings, by contrast, episodes of pneumocystis can lead to higher rates of morbidity and mortality -- thereby making the prevention of PCP more critically important in improving survival in such populations.
In addition, participants in observational studies are often treated by primary care practitioners, who may take a less aggressive approach to prophylaxis and treatment than is taken in clinical trials conducted in tertiary-care centers. In combination, these two factors -- closeness of follow-up and aggressiveness of treatment -- are likely to contribute to the observed difference that PCP prophylaxis has on survival in observational studies.
Another factor that almost certainly contributes to the lack of survival benefit noted in clinical trials is that when a predetermined number of PCP events have accumulated, the trials are stopped for ethical reasons. In fact, the two randomized studies that compared AP prophylaxis to no therapy or placebo(4, 5) showed no survival benefit precisely because these studies were carefully monitored for differences in PCP endpoints, and both trials were stopped as soon as significant differences were observed. As a result, both studies had so few survival endpoints that it was impossible to judge the effectiveness of therapy -- which is not the case in epidemiological studies, which have no such ethical imperative.(2, 3, 13-16)
Is there unrecognized bias in observational studies?
It is possible that participants in non-randomized studies of PCP prophylaxis are somehow healthier than those not on therapy, and this potential source of bias may explain the observed differences in survival attributed to therapy. Certainly this could explain the findings in a study that did not adjust for disease stage(14) -- except that a number of other studies(2, 3, 13, 15, 16) did adjust for a variety of prognostic indicators, including CD4 count, symptoms, concurrent therapies, and disease stage, and obtained similar results . Moreover, data from the Multicenter AIDS Cohort Study suggest that those who start PCP prophylaxis are, if anything, sicker than those who do not.(27) Finally, the fact that several completely different studies, conducted in completely different settings, reported similar results argues against unrecognized bias as an explanation for these results.
Happily, two recent studies also support the hypothesis that PCP prophylaxis improves survival. These studies suggest that prophylaxis, particularly with oral agents, reduces the severity of PCP episodes when breakthroughs do occur.(28, 29) If confirmed by larger trials, these findings would favor better outcomes with PCP prophylaxis, including shorter hospitalizations and improved survival.
In summary, six of the eight studies (Table 2) that compared PCP prophylaxis with placebo or no therapy show a significant survival advantage for prophylaxis. Although significant, the duration of this benefit is probably limited to about six months to a year, due to the onset of replacement mortality from end-stage AIDS-related conditions such as MAC, CMV, and severe wasting (Figure). Studies comparing two or more PCP prophylaxis regimens (Table 1) have no control group and were not designed to detect small differences in survival benefit between effective therapies. Other factors, including unrecognized bias in observational studies, probably contribute to the differences in survival attributable to PCP prophylaxis when the results of observational studies are compared to those of clinical trials. Nonetheless, bias appears to be an insufficient explanation for the survival benefit from PCP prophylaxis reported in observational studies.
1. Hoover et al. Clinical manifestations of AIDS in the era of pneumocystis prophylaxis. N Engl J Med 1993; 329: 1922-6.
3. Saah et al. Factors influencing survival after AIDS: report from the Multicenter AIDS Cohort Study (MACS). JAIDS 1999; 7: 287-95.
4. Girard et al. Prevention of Pneumocystis carinii pneumonia relapse by pentamidine aerosol in zidovudine-treated AIDS patients. Lancet 1989; 1: 1348-53.
5. Hirschel. A controlled study of inhaled pentamidine for primary prevention of Pneumocystis carinii pneumonia. N Engl J Med 1991; 324: 1079-83.
6. Leoung et al. Aerosolized pentamidine for prophylaxis against Pneumocystis carinii pneumonia. N Engl J Med 1990; 323: 769-75.
7. Martin et al. A comparison of the effectiveness of three regimens in the prevention of Pneumocystis carinii pneumonia in human immunodeficiency virus-infected patients. Arch Int Med 1992; 152: 523-8.
8. Hardy et al. A controlled trial of trimethoprim-sulfamethoxazole or aerosolized pentamidine for secondary prophylaxis of Pneumocystis carinii pneumonia in patients with acquired immunodeficiency syndrome. N Engl J Med 1992; 327: 1842-8.
9. Schneider et al. A controlled trial of aerosolized pentamidine or trimethoprim-sulfamethoxazole as primary prophylaxis in patients with human immunodeficiency virus infection. N Engl J Med 1992; 327: 1836-41.
10. Podzamczer et al. Thrice-weekly co-trimoxazole is better than weekly dapsone-pyrimethamine for the prevention of Pneumocystis carinii pneumonia in HIV-infected patients. AIDS 1993; 7: 501-6.
11. Bozzette et al. A randomized trial of three anti-pneumocystis agents in patients with advanced human immunodeficiency virus infection. N Engl J Med 1995; 332: 693-9.
12. Podzamczer. Intermittent trimethoprim-sulfamethoxazole compared with dapsone-pyrimethamine for simultaneous primary prophylaxis of Pneumocystis carinii pneumonia and toxoplasmosis in patients infected with HIV. Ann Intern Med 1995; 122: 755-61.
13. Graham et al. Effects on survival of early treatment of human immunodeficiency virus infection. N Engl J Med 1992; 326: 1037-42.
14. Carr et al. Trimethoprim-sulfamethoxazole appears more effective than aerosolized pentamidine as secondary prophylaxis against Pneumocystis carinii pneumonia in patients with AIDS. AIDS 1992; 6: 165-71.
15. Chaisson et al. Pneumocystis prophylaxis and survival in patients with advanced human immunodeficiency virus infection treated with zidovudine. Arch Intern Med 1992; 152: 2009-13.
16. Osmond et al. Changes in AIDS survival time in two San Francisco cohorts of homosexual men, 1983-1993. JAMA 1994; 271: 1083-7.
17. Fischl et al. Safety and efficacy of trimethoprim-sulfamethoxazole chemoprophylaxis for Pneumocystis carinii pneumonia in AIDS. JAMA 1988; 259: 1185-9.
18. Slavin et al. Oral dapsone vs. nebulized pentamidine for Pneumocystis carinii pneumonia prophylaxis: an open randomized prospective trial to assess efficacy and hematological toxicity. AIDS 1992; 6: 1169-74.
19. Golden et al. A randomized comparison of once-monthly or twice-monthly high-dose aerosolized pentamidine prophylaxis. Chest 1993; 104: 743-50.
20. Girard et al. Dapsone-pyrimethamine compared with aerosolized pentamidine as primary prophylaxis against Pneumocystis carinii pneumonia and toxoplasmosis in HIV infection. N Engl J Med 1993; 328: 1514-20.
21. Torres et al. Randomized trial of dapsone and aerosolized pentamidine for the prophylaxis of Pneumocystis carinii pneumonia and toxoplasmic encephalitis. Am J Med 1993; 95: 573-83.
22. May et al. Trimethoprim-sulfamethoxazole vs. aerosolized pentamidine for primary prophylaxis of Pneumocystis carinii pneumonia: a prospective, randomized, controlled clinical trial. JAIDS 1994; 7: 457-62.
23. Blum et al. Comparative trial of dapsone vs. trimethoprim-sulfamethoxazole for primary prophylaxis of Pneumocystis carinii pneumonia. JAIDS 1992; 5: 341-7.
24. Ruskin and La Riviere. Low-dose co-trimoxazole for prevention of Pneumocystis carinii pneumonia in human immunodeficiency disease. Lancet 1991; 337: 468-71.
25. Stein et al. Use of low-dose trimethoprim-sulfamethoxazole thrice weekly for primary and secondary prophylaxis of Pneumocystis carinii pneumonia in human immunodeficiency virus-infected patients. AAC 1991; 35: 1705-9.
26. Jorde et al. Utility of dapsone for prophylaxis of Pneumocystis carinii pneumonia in trimethoprim-sulfamethoxazole-intolerant, HIV-infected individuals. AIDS 1993; 7: 355-9.
27. Graham et al. Access to therapy in the Multicenter AIDS Cohort Study 1989-1992. J Clin Epidemiol 1994; 47: 1003-12.
28. Mallal et al. Severity and outcome of Pneumocystis carinii pneumonia in patients of known and unknown HIV status. JAIDS 1994; 7: 148-53.
29. Fahey et al. Effects of aerosolized pentamidine prophylaxis on the clinical severity and diagnosis of Pneumocystis carinii pneumonia. Am Rev Resp Dis 1992; 146: 844-8.
Neil M.H. Graham, M.B.B.S., M.D., M.P.H., is Associate Professor of Epidemiology, Johns Hopkins School of Hygiene and Public Health and Associate Professor of Medicine, John Hopkins School of Medicine, Baltimore, MD.
Back to the February 1996 HIV Newsline contents page.
Effects of Atovaquone (A) Versus Dapsone (D) in the Prevention of PCP for Patients Intolerant to Trimetheoprim-Sulfa
This article was provided by San Francisco General Hospital. It is a part of the publication HIV Newsline.