Corey, Daniel M MD*; Kim, Hyung Woo PhD*; Salazar, Raul MD†; Illescas, Ricardo MD†; Villena, Juan MD†; Gutierrez, Luis MD†; Sanchez, Jorge MD‡; Tabet, Stephen R MD, MPH*§
The impact of highly active combination or antiretroviral therapy (HAART) on the clinical course of HIV infection is well documented in North American and European populations.1,2 The largest populations of persons with HIV are living in resource-poor areas, however, where much less is known about the effect of HAART on the clinical course and survival of HIV-infected persons.3 In 1996, the Almenara Hospital in Lima, Peru initiated an observational database to record the response to antiretroviral therapy (ART) of all HIV-infected patients attending the hospital. The hospital serves as the major health care facility for persons enrolled in the Social Security System in Peru and was the first center to establish a dedicated clinic for the management of HIV-infected persons in Lima. This article describes the clinical response to ART, including long-term survival, among persons with HIV-1 who were enrolled and followed at the hospital.
The HIV Clinic at Almenara Hospital was established in 1992 and is operated as a special clinic within the Department of Internal Medicine. The clinic is staffed by 10 physicians who have received specialized training in HIV-1 care.4 In November 1997, triple-drug therapy (HAART) became the standard of care for clinic patients. The guidelines for initiating HAART included (1) CD4 counts ≤200 cells/mm3 and (2) symptomatic opportunistic infections (OIs) or plasma HIV RNA levels >55,000 copies/mL.5,6 Because tuberculosis (TB) is prevalent in Peru and BCG vaccination is universal, routine chest radiographs were taken on all persons. Isoniazid (INH) therapy (300 mg once a day) was recommended for 1 year for all persons without active TB. For persons with evidence of active TB, combination TB therapy was initiated first and then followed by HAART, usually 2 months after initiation of TB therapy.7Pneumocystis carinii pneumonia (PCP) prophylaxis was routine in all patients with CD4 counts <200 cells/mm3 and was discontinued when CD4 counts were greater than 200 cells/mm3 on at least 2 consecutive determinations.7 Patients on ART were seen monthly. Antiretroviral drugs were imported into the country from the pharmaceutic company or its subsidiary holding the patent for the medication. Protease inhibitors (PIs) became available in 1997, and efavirenz became available in 2002. These drugs were then immediately incorporated into initial regimens. No generic medications were available during this period. Compliance was not monitored by pill counts; however, because the clinic was a referral clinic through the Peruvian Medical Care System, patients were assessed to be highly compliant with their ART by clinic staff.
Data Collection and Statistical Methods
A database entry, including patient demographics, dates of patient visits, clinical diagnoses listed at each visit, and all antiretroviral medications, was established for each person attending the clinic. Longitudinal data on dates, duration of hospitalization, and discharge diagnoses at each hospitalization and clinic visit were abstracted and put into the database, as were CD4+ T-cell counts. Dosages of such drugs and notation of patient compliance were not recorded in the database.
Statistical analyses were performed using SAS software (SAS Institute, Cary, NC) and R software programs (R Foundation for Statistical Computing, Vienna, Austria). Comparison of categoric data between groups was evaluated by the χ2 test, and comparison of continuous data between groups was evaluated by the student's t-test and Wilcoxon test. Trend in CD4+ T-cell counts since the initiation of the ART regimen was estimated using the locally weighted regression and smoothing scatterplot (LOWESS) smoothing method.8,9 Survival after the initiation of ART was estimated using the method of Kaplan and Meier. All patients were censored at the last follow-up visit. The incidence of opportunistic infections (OIs) was estimated using cumulative incidence estimates. Statistical comparison for time-to-event data was performed using the likelihood ratio test from Cox regression models. All reported P values are 2-sided.
We identified 679 separate HIV-1-infected patients who attended the clinic between 1994 and 2003 (Table 1). Of these 679 patients, 564 received combination (3-drug) ART, 92 were untreated and are the subject of another report,10 and 23 received monotherapy or 2-drug therapy. These 23 patients were excluded from this analysis.
Of the 564 patients who received HAART, 142 received 2 reverse transcriptase (RT) inhibitors and a nonnucleoside reverse transcriptase inhibitor (NNRTI), 371 received 2 RT inhibitors and a PI, and 51 were administered other regimens, (eg, 3 RT inhibitors; see Table 1). The most common RT inhibitors were azidothymidine (AZT)/lamivudine (3TC) (31%), stavudine (d4T)/3TC (16%), or AZT/didanosine (ddI) (28%). Efavirenz was the most common NNRTI, being used in 140 (99%) of the patients on this type of regimen. Among those receiving PI-based regimens, 42% used indinavir, 41% nelfinavir, and 26% ritonavir. More than half of those who received ritonavir were given it as part of a “boosted” PI regimen (see Table 1). PCP prophylaxis was demonstrated in the records by 92% of the patients, and TB prophylaxis was demonstrated by 80%.
CD4+ T-cell counts within the 4 months before initiation of HAART were available on 415 (74%) of the 564 patients: 290 men and 125 women. The demographic, clinical aspects, and treatment regimens of these 415 patients were similar to those of the entire 564-patient population. The mean and median CD4 counts at onset of therapy were 91 cells/mm3. The proportion of persons with a low CD4 count (≤100 cells/mm3) in the PI-containing regimens was significantly higher (159 [56.6%] of 281 patients) than that of the 2 RT plus NNRTI regimen (41 [43.2%] of 95 patients; P = 0.02).
Response to Therapy
The median duration of follow-up was 35 months (range: 3 days to 7 years). Figure 1A illustrates overall survival during the course of follow-up. The overall survival rate was 97% at year 1, 96% at year 2, 95% at year 3, 94% at year 4, and 91% at year 5. Survival rates were higher among those initiating HAART with a CD4 count >100 cells/mm3 than ≤100 cells/mm3 (P = 0.04; see Fig 1B). Among those with CD4 counts ≤100 cells/mm3, the overall survival rate at 3 years was 95%. Overall survival rates for the 415 patients with available data on CD4+ T-cell counts were similar to the overall survival rates for the entire 564-patient cohort.
Overall, patients who received 2 RT inhibitors and a PI regimen had a better survival rate than those with 2 RT inhibitors and an NNRTI (hazard ratio [HR] = 0.29, 95% confidence interval [CI]: 0.12 to 0.71; P = 0.01; Fig. 2A). This trend for increased survival in PI-based regimens was consistent at all levels of CD4+ T-cell counts but was not statistically significant at low CD4+ T-cell counts, likely because of the small number of persons in each subset (see Fig. 2B). The separation of the survival curves between PI and non-PI regimens was greatest within 6 to 12 months of the onset of HAART (see Fig. 2). The causes of death for the 6 patients who died within the first year in the PI-based regimen were progressive wasting (2), persistent TB (2), sepsis attributable to underlying subacute myelitis (1), and unknown (1). The causes of the 10 deaths in the 2 RT inhibitor and NNRTI regimens were progressive wasting (2), lymphoma (1), probable pulmonary Kaposi sarcoma (KS) (1), cryptococcal meningitis (1), undiagnosed pulmonary insufficiency (2), progressive cytomegalovirus (CMV) (1), and unknown (2).
Opportunistic Infections and CD4+ T-Cell Count
Overall, 194 OIs were recorded from 132 of the 564 subjects, and 18% of patients reported an OI within 2 years of initiation of HAART (Fig. 3A). For those who initiated therapy with a CD4 count <100 cells/mm3, the HR for developing an OI in the first year for those on 2 RT inhibitors plus a PI was 0.31 (95% CI: 0.09 to 1.0; P = 0.054) versus those on 2 RT inhibitors plus an NNRTI. The time to the first OI in the first year of therapy tended to be shorter for the PI- than NNRTI-containing regimens (see Fig. 3B), but this difference was not statistically significant (HR = 0.64, 95% CI: 0.37 to 1.11; P = 0.11).
This observational database indicates that ART seems to be highly effective in this developing world clinic. Overall, the survival rate among our patients was 95% at year 3 and 91% at year 5. Among patients who initiated therapy with CD4 counts ≤100 cells/mm3, the mortality rate at 2 years was 3.9%, which is comparable to that seen in European and US populations.1,2 We did not have a randomized untreated control group; hence, the relative survival benefit over no therapy cannot be measured. These rates are far lower than those of untreated disease in the developing world, however.11,12 We did not have a linked database of all sequential deaths to ensure that our case assessment for survival was 100%. Nevertheless, our follow-up and case ascertainment for cause of death over the initial 2 years of therapy were extremely high, even for persons who initiated therapy with baseline CD4 counts ≤100 cells/mm3. The data from this clinic-based study show that ART is feasible in major metropolitan areas in resource-poor settings, that patients can be relied on to adhere to medication, and that ART can achieve a prolonged survival benefit.13
Our cohort is one of the largest observational studies of ART-naive patients from resource-poor settings yet reported. Because the nature of the patients selected for therapy may differ, comparison of our data with those of other observational cohorts is difficult.14-19 Most reported studies, including recent articles, have followed patients for only 6 to 12 months and do not report long-term mortality data.20,21 The 5-year follow-up of our cohort showing sustained benefit provides encouragement to providers and policy makers. Our patients were closely managed by a defined set of clinicians experienced in HIV care.4 Moreover, the Peruvian Social Security Health Care System is severely crowded; hence, patients who attended the clinic were self-selected for their perseverance and interest in treatment, perhaps contributing to the exemplary results.
Our data suggesting improved survival of patients on PI-based regimens over those on 2 RT inhibitors and NNRTI regimens was surprising to us and the clinicians involved in the direct care of these patients, especially because more persons who received PI-based regimens had low CD4+ T-cell counts at the initiation of therapy. These potential differences were seen mainly in the first year after initiation of therapy and were based on only a small number of deaths in each group (6 vs. 10 persons in the PI vs. NNRTI groups, respectively). As such, our data, although provocative, require corroboration.
One deficiency of any observational study is the possibility of selection bias. For example, PI-based regimens were available before NNRTI-based regimens. Moreover, because PIs tended to have more side effects and were, in general, more costly, there was a tendency to reserve these medications for persons who were more ill once NNRTIs became available. Although this perhaps makes the differences in survival we noted even more striking, it also illustrates the selection bias in our cohort.
There are numerous issues that remain in the area of ART among patients in resource-poor countries. The frequency of resistance, the role of compliance in response, the optimal regimen of effectiveness versus cost, and differences in immune reconstitution syndromes between regimens are all important issues that we could not resolve in our analyses. Our data clearly illustrate the extended effectiveness of antiretroviral medications in resource-poor settings and establish the importance of extending ART to HIV-infected persons globally.
The authors acknowledge the Bill and Melinda Gates Foundation for their support of the Puget Sound Partners Program, a program designed to provide educational experience for medical students in resource-poor settings. The article is dedicated to Stephen Tabet, MD, whose career was devoted to the care of persons with HIV infection.
1. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet. 2002;360:119-129.
2. Palella FJ Jr, Deloria-Knoll M, Chmiel JS, et al. Survival benefit of initiating antiretroviral therapy in HIV-infected persons in different CD4+ cell strata. Ann Intern Med. 2003;138:620-626.
3. Attaran A, Sachs J. Defining and refining international donor support for combating the AIDS pandemic. Lancet. 2001;357:57-61.
4. Kitahata MM, Koepsell TD, Deyo RA, et al. Physicians' experience with the acquired immunodeficiency syndrome as a factor in patients' survival. N Engl J Med. 1996;334:701-706.
5. Dybul M, Fauci AS, Bartlett JG, et al. Guidelines for using antiretroviral agents among HIV-infected adults and adolescents. Ann Intern Med. 2002;137:381-433.
6. Yeni PG, Hammer SM, Carpenter CC, et al. Antiretroviral treatment for adult HIV infection in 2002; updated recommendations of the International AIDS Society-USA Panel. JAMA. 2002;288:222-235.
7. Hogan DR, Salomon JA. Prevention and treatment of human immunodeficiency virus/acquired immunodeficiency syndrome in resource-limited settings. Bull World Health Organ. 2005;83:135-143.
8. Cleveland WS. Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc. 1979;74:829-836.
9. Cleveland WS. LOWESS: a program for smoothing scatterplots by robust locally weighted regression. Am Stat. 1981;35:54.
10. Corey D, Kim HW, Salazar R, et al. The natural history of untreated HIV infection in Lima, Peru. Human Vaccines. 2005;1:160-164.
11. Deschamps MM, Fitzgerald DW, Pape JW, et al. HIV infection in Haiti: natural history and disease progression. AIDS. 2000;14:2515-2521.
12. Schneider M, Zwahlen M, Egger M. Natural history and mortality in HIV positive individuals living in resource poor settings: a literature review. UNAIDS Obligation HZ/03/463871, Joint United Nations Programme on HIV/AIDS (UNAIDS), Geneva (2005). Available at: http://www.epidem.org/publications.htm
. Accessed November 2005.
13. Orrell C, Bangsberg DR, Badri M, et al. Adherence is not a barrier to successful antiretroviral therapy in South Africa. AIDS. 2003;17:1369-1375.
14. Sever P, Leger P, Charles M, et al. Antiretroviral therapy in a thousand patients with AIDS in Haiti. N Engl J Med. 2005;353:2325-2334.
15. Farmer P, Leandre F, Mukherjee J, et al. Community-based treatment of advanced HIV disease: introducing DOT-HAART (directly observed therapy with highly active antiretroviral therapy). Bull World Health Organ. 2001;79:1145-1151.
16. Tassie JM, Szumilin E, Calmy A, et al. Highly active antiretroviral therapy in resource-poor settings: the experience of Medecins Sans Frontieres. AIDS. 2003;17:1995-1997.
17. Weidle PJ, Malamba S, Mwebaze R, et al. Assessment of a pilot antiretroviral drug therapy programme in Uganda: patients' response, survival, and drug resistance. Lancet. 2002;360:34-40.
18. Coetzee D, Hildebrand K, Boulle A, et al. Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa. AIDS. 2004;18:887-895.
19. Djomand G, Roels T, Ellerbrock T, et al. Virologic and immunologic outcomes and programmatic challenges of an antiretroviral treatment pilot project in Abidjan, Cote d'Ivoire. AIDS. 2003;17(Suppl 3):S5-S15.
20. Seyler C, Anglaret X, Dakoury-Dogbo N, et al. Medium-term survival, morbidity and immunovirological evolution in HIV-infected adults receiving antiretroviral therapy, Abidjan, Cote d'Ivoire. Antivir Ther. 2003;8:385-393.
21. Laurent C, Diakhate N, Gueye NF, et al. The Senegalese government's highly active antiretroviral therapy initiative: an 18-month follow-up study. AIDS. 2002;16:1363-1370.
© 2007 Lippincott Williams & Wilkins, Inc.