The number of persons initiating antiretroviral therapy (ART) in low and middle income countries has increased nearly eight-fold since 2004, with the World Health Organization estimating that 3 million persons were receiving ART by the end of 2007 . CD4 cell count and HIV RNA viral load in response to ART are important measures of the efficacy of ART in individual patients and of the effectiveness of ART in populations of patients enrolled in HIV care and treatment programs. However, few data exist on long-term CD4 response to ART among patients receiving care in resource-limited settings, where HIV RNA testing is not generally available or conducted. Several studies in Europe and North America have reported robust improvements in CD4 cell counts following ART initiation in clinical trials and in observational studies [2–18]. In addition, CD4 cell count at the time of ART initiation is an important determinant of the degree of immunologic and virologic response [17,19–22], as well as subsequent risk of morbidity and mortality [8,23,24]. Among those patients who are able to remain on ART, robust immunologic responses can be maintained for long periods [3,4,11,13] and the risk of serious morbidity and mortality may eventually diminish to levels observed in the general population .
Although data from resource-limited settings are less commonly available, some investigators of research and scale-up cohorts in sub-Saharan Africa [26–32], Barbados , Brazil [34,35], China , Thailand , and Cambodia  have reported effects of ART on clinical and immunologic outcomes that were comparable to those observed in resource-rich settings. However, the majority of the studies in developing countries have had follow-up times of 1–2 years, Thus, although it has been shown in developed countries [3,13,18], the degree to which CD4 responses can be maintained for longer periods after ART initiation in developing countries has not been demonstrated. This analysis was conducted to describe the determinants of CD4 response trajectories up to 5 years after initiation of ART in resource-limited settings.
The Antiretroviral Therapy in Low-Income Countries (ART-LINC) Collaboration of the International Databases to Evaluate AIDS (IeDEA) (see www.art-linc.org and www.iedea-hiv.org) is a network of HIV/AIDS treatment programs in Africa, Latin America, and Asia, which has been described in detail elsewhere [39,40]. Briefly, HIV care and treatment programs from low-income and middle-income countries were approached to determine their interest and capacity to collaborate. Thirty-two treatment programs were approached, 29 agreed to participate, and 27 contributed data to the present analysis.
All patients initiating ART who were ART naive, aged 15 years or older, and enrolled on a such a date that they had the potential to contribute at least 6 months of follow-up data on ART were eligible for this analysis. Patients with a baseline CD4 cell count of 500 cells/μl or above or missing data in key variables were excluded. Baseline CD4 cell count was defined as the CD4 measurement occurring closest to the date of starting ART, within a window of 6 months prior to 1 week after ART initiation. The rate of CD4 testing varied by site, but was generally about two CD4 tests per patient per year. Data collection was approved by Institutional Review Boards or ethics committees for all cohorts. ART was defined as any combination of at least three antiretroviral drugs.
Patients were classified as being dead, lost to program, or presumed to be alive and on ART based on all available information at the time of the closure of the study database. Lost to follow-up was defined as not returning to clinic for 1 year or longer after the last recorded visit, as described elsewhere for this study population . Time on ART was measured from the date of ART initiation to the date of the last known encounter. Data were censored at the closing date of each cohort to a maximum of 5 years after ART initiation. Descriptive analyses were done to examine CD4 cell count distributions over time on ART and trends in the mean and median CD4 cell count over time on ART.
Linear mixed regression models for repeated measures were used to evaluate the association between independent variables and CD4 response over time on ART. We constructed a multilevel model (CD4 measurements within patients nested within cohorts) with random intercepts for each patient and cohort. CD4 cell count response values were square-root transformed to better approximate a normal distribution. Baseline CD4 cell count was included in the models as a categorical variable (0–24, 25–49, 50–99, 100–149, 150–199, 200–299, ≥300 cells/μl). To allow flexible modeling of the CD4 response with time on ART, we tested a series of fractional polynomial functions up to the third order [42,43]. The fractional polynomial function was allowed to vary between baseline CD4 categories by including interaction terms. We further allowed for random variation in the slope of the fractional polynomial function between individuals. The model with the highest gain in fit, a second-order fractional polynomial with time transformed to the power −0.5 and the natural logarithm of time (fractional polynomial [2, (−0.5, ln)]), was used in final modeling.
The following variables were considered for their influence on CD4 trajectory: sex, age at ART initiation, clinical stage of disease, year of ART initiation, and initial ART regimen. For age, we used three categories: 15–29, 30–39, and 40–65 years. Clinical stage was categorized as less advanced (CDC stage A/B, WHO stage I/II), more advanced (CDC stage C, WHO stage III/IV), or not assessed/unknown. Year of HAART initiation was entered as before 2001, 2001, 2002, 2003, 2004, 2005, 2006, and 2007. Type of ART regimen was classified as protease inhibitor based [two nucleoside reverse transcriptase inhibitors (NRTIs) and one protease inhibitor, including ritonavir-boosted protease inhibitor], nonnucleoside reverse transcriptase inhibitor (NNRTI) based (two NRTIs and one NNRTI), and other or unknown regimens.
Some patients had missing information on baseline CD4 cell count and baseline clinical stage. In order to account for missing data, we used multiple imputation by chained equations to impute the baseline CD4 count conditional on the first CD4 cell count after the start of HAART, the time from start of HAART until the first CD4 response, as well as the other predictor variables [44,45]. Modeling of the CD4 response over time as described above was on five imputed datasets; results were combined using Rubin's rules. Model results were similar in raw and imputed data sets. Analyses were done using SAS version 9.1 (SAS Institute Inc., Cary, North Carolina, USA) and Stata version 10.0 (Stata Corporation, College Station, Texas, USA).
Of 35 010 treatment-naive patients aged 15 and over who initiated ART in 1995 or later at one of the ART-LINC collaborating centers, 5835 (17%) were ineligible because the maximum follow-up time they could contribute was less than 6 months and 264 (0.8%) were ineligible because the baseline CD4 cell count was 500 cells/μl or more. Of the 28 911 eligible patients, 8933 (26%) were excluded because they had either no recorded CD4 cell count (n = 2291) or had a baseline CD4 cell count, but no follow-up counts (n = 6642). An additional 11 patients were excluded due to missing information on their sex. The 19 967 patients included in analyses contributed a total of 71 067 CD4 cell count measurements (median per patient 3, range 1–27) up to 5 years after ART initiation. Fifteen thousand, seven hundred and seventy-eight patients (79.0%) had a baseline and more than one follow-up CD4 cell count, 2670 (13.4%) had more than one follow-up CD4 count but no baseline measurement, and 1519 (7.6%) had one follow-up CD4 cell count but no baseline measurement. The overall time on ART for patients included in this analysis was 39 200 person-years.
Patients excluded from analyses due to missing CD4 cell counts (n = 8933) were more likely to be men (43 vs. 40%, P < 0.001) and of either a less advanced (50 vs. 57%, P < 0.001) or unknown stage of disease (17 vs. 9%, P < 0.001) when compared with those included in the analysis. Of the 8933 excluded, 4370 (49%) were presumed to be alive and attending clinic; 1352 (15%) were known to have died (86% within 6 months following ART initiation); and 3111 (35%) were considered lost to follow-up as of the closing date of the database.
Characteristics of study population
Characteristics of the 19 967 patients in the cohort are described in Table 1. There were 16 170 patients from 22 centers in Africa (81%), 1398 from three centers in Latin America (7%), 2399 from two centers in India, and Southeast Asia (12%). The majority of patients (60%) was women and the median age at ART initiation was 35 years. Among those with a baseline weight recorded (80% of patients), the median was 55 kg for women (IQR 48–63 kg) and 58 for men (IQR 52–65 kg). Among patients with a baseline CD4 cell count (79% of patients), the median was 114 cells/μl; 121 cells/μl for women (IQR 56–186) and 104 for men (IQR 45–179). The range of the median CD4 cell count at ART initiation varied widely by cohort (from 61 to 181 cells/μl) with higher counts in women than in men in all but two sites (Fig. 1). Among those with less advanced clinical stage and a baseline CD4 cell count (n = 5460), 4221 (77%) had a CD4 count below 200 cells/μl. The majority of patients initiated ART in an advanced clinical stage of HIV disease and most patients started on regimens with two NRTIs and one NNRTI (92%), with 6% starting with two NRTIs and a protease inhibitor (Table 2).
During follow-up, 568 patients died (3%) and 1790 were lost to follow-up (9%); 820 patients (4%) from eight centers were followed up beyond 4 years after starting ART.
CD4 cell count trajectories after antiretroviral therapy initiation
The median time on ART was 1.6 years (IQR 1.1–2.7 years). The median rate of CD4 cell count testing over the follow-up period was 1.8 tests per person-year on ART (range across centers 0.4–3.5). The median CD4 cell count was 114 cells/μl among all patients at baseline and increased to 395 cells/μl among those remaining on ART for 5 years. Figure 2 shows the crude median CD4 over time on ART stratified by sex, age, clinical stage, initial ART regimen, baseline CD4, and status of patient at closure of the database. On average, the increases were steeper in women than in men, with differences widening with time on ART. Increases were also more pronounced in younger patients than in older patients. Smaller differences were observed between patients initiating ART in more advanced and less advanced clinical stages, and patients starting ART with different regimens. The baseline CD4 cell count was the most important factor differentiating CD4 trajectories: patients initiating ART with higher CD4 cell counts tended to achieve and maintain higher levels up to 5 years. CD4 cell counts increased substantially in all patients except in those who died. Of note, those who were lost to follow-up had CD4 trajectories that were more similar to those who were known to be alive than to those who were known to have died (Fig. 2).
In multivariable regression models, baseline CD4 cell count and age continued to independently predict CD4 trajectories, whereas the effect of sex virtually disappeared. Table 2 shows the mixed-effects model for CD4 (square-root transformed). Tables 3 and 4 give predicted mean CD4 cell counts over time by age and baseline CD4 cell count for men and women. Only patients initiating ART at 200 cells/μl and higher would be expected attain a CD4 cell count near or above 500 cells/μl after 5 years of therapy. Figure 3 shows the predicted trajectory in median CD4 cell counts by baseline CD4 and the observed median CD4 cell count at each time point. Starting ART at CD4 cell counts below 100 cells/μl was associated with substantially worse predicted CD4 trajectories compared with starting ART at higher CD4 cell count. The model fit was good when comparing predicted with observed data.
This study combined data from 27 centers from resource-limited settings in Africa, Latin America, and Asia to assess CD4 trajectories after ART initiation among patients at sites engaged in the scale-up of HIV care and treatment. The data demonstrate robust CD4 responses to ART that are sustained over several years. Our results are thus encouraging regarding the long-term effectiveness of ART in resource-limited settings, but they naturally are only applicable to those who are able to remain on ART for extended periods.
The availability of information on CD4 cell count to the health care providers at these sites in and of itself was likely a critical factor in getting many patients on to ART earlier than they otherwise would have in these programs. Were the programs to rely solely on clinical staging criteria (77% of patients with less advanced clinical stage had CD4 cell counts less than 200 cells/μl), many patients would have likely started therapy even later or not in time, limiting the full potential to benefit from therapy. Thus, extending access to CD4 testing in programs that do not currently have it would likely result in earlier ART initiation and improve treatment outcomes substantially.
Apart from the amount of time on ART, the single most important factor determining CD4 trajectories and the maximum CD4 cell count reached was the baseline CD4 cell count. Patients with higher CD4 cell counts at ART initiation achieved a higher CD4 cell count in the following months and years. Although this has been shown by other investigators in both resource-rich [13,21,46] and resource-limited settings [29,47], the importance of this observation cannot be overstated. The baseline CD4 cell count, second only to subsequent medication adherence (which we could not measure), is the most important predictor of clinical progression and survival after ART initiation [8,23,24,31,48–50]. Patients with lower baseline CD4 cell count remain at risk for opportunistic infections for a substantially longer period than patients starting ART at higher CD4 cell counts, increasing their risk for serious morbidity and death. In this analysis, 35% of the patients had a baseline CD4 cell count below 100 cells/μl and 29% had a count between 100 and 199 cells/μl. Fortunately, a recent analysis of the ART-LINC database indicated that median CD4 cell counts at the start of ART, although still low in most of the cohorts, have increased in recent years . However, our model predictions suggest that, after several years of therapy, only those few patients initiating ART at 200 cells/μl or higher could be expected to achieve CD4 cell counts near or above 500 cells/μl or higher (Tables 3 and 4), the level at which their risk of mortality may diminish to that observed in the general population in some settings .
Although most patients quickly achieved CD4 cell counts above the important clinical milestone of 200 cells/μl, we also note that there remains increased risk of morbidity and mortality even at CD4 cell counts above 200 cells/μl, especially in developing countries . In our study of survival in these cohorts, we showed that the risk of mortality continues to diminish with increasing CD4 cell count, even among patients with baseline CD4 cell counts above 200 cells/μl . For example, relative to those patients with baseline CD4 cell counts below 25 cells/μl, we found an RR of 0.67 in patients with baseline CD4 cell count between 100 and 200 cells/μl, 0.44 in patients with baseline CD4 cell count between 200 and 349 cells/μl, and 0.26 in patients with baseline CD4 cell counts above 350 cells/μl. Additionally, in one study of South African patients on ART for 3 or more years, the risk of incident tuberculosis was still five to 10 times higher than in the general population . Finally, the Strategies for Management of Antiretroviral Therapy (SMART) study recently reported a higher incidence of opportunistic illness in those patients who interrupted therapy, almost all of whom had CD4 cell counts above 200 cells/μl .
Our findings on long-term CD4 response appear to be consistent with those of other long-term investigations in Switzerland , the United States , and the Netherlands , which examined immunologic response up to 4, 6, and 7 years after ART initiation, respectively. However, these studies were restricted to continuously treated, virologically suppressed patients, making a comparison with our patients difficult. Other factors may further confound this comparison, such as lower CD4 cell counts in the general populations in developing than in developed countries, which could be due to other differences such as the prevalence of tuberculosis or helminth coinfections. Nonetheless, an important next step, therefore, is to conduct more direct comparisons of CD4 response between developed and developing countries within strata of baseline CD4 while controlling for other differences between the patient populations.
An intriguing finding in our investigation and others is the differences in trajectories by sex in crude analyses. We have previously reported sex differences in the CD4 cell count at ART initiation . In the present study, women had higher baseline CD4 cell counts in 25 of the 27 cohorts included in the analysis. Similar to other investigators in Spain , we found that sex differences in CD4 trajectories after starting ART were largely explained by differences in CD4 cell counts at baseline.
Given that baseline CD4 cell count is such an important determinant of CD4 trajectories after ART initiation, it is important to gain a better understanding of the determinants of baseline CD4 cell count among patients initiating ART in resource-limited settings. These determinants likely operate at multiple levels, starting with the knowledge of being at risk for HIV infection among persons in the community, access to and uptake of HIV testing and counseling, intensity of active screening for HIV in the healthcare setting and in the community, entry points in to care, availability of CD4 testing, and, among programs providing pre-ART care, frequency and intensity of clinical monitoring and CD4 testing. Some of these factors are more easily modifiable than others. A recent investigation in the Netherlands suggested that entry into care with low CD4 cell counts explained a substantial portion of the variation in mortality rates across HIV care and treatment centers . Clearly, further studies are needed to help inform efforts aimed at getting patients on to ART with higher baseline CD4 cell counts.
Most national care and treatment programs have adapted the WHO criteria for ART eligibility. Updated in 2006, they recommend initiating ART in patients in WHO stage IV, all patients with CD4 cell counts below 200 cells/μl (irrespective of WHO stage), and in patients with WHO stage II/III and CD4 cell counts below 350 cells/μl . In our analysis, only 6% of patients were eligible based on having less advanced disease with CD4 cell counts between 200 and 350 cells/μl. In other words, 94% of the patients who initiated ART at ART-LINC sites were in the advanced clinical and immunologic stages by the time they initiated treatment. Our analysis clearly suggests that there is substantial room for improvement in earlier initiation of ART across a diversity of geographic settings.
Our study has several strengths. Twenty-seven sites on three continents were represented in the analysis, and the overall finding (substantial improvements in CD4 cell counts after ART initiation) was consistent across the diversity of geographical settings and contexts. Our results should thus be applicable to a wide range of lower-income settings. In addition, this investigation benefited from a large amount of follow-up time (up to 5 years for some patients). Although the number of patients remaining in follow-up decreased substantially with time since the start of ART, there were still 820 patients from eight clinics who were actively followed up in the fifth year. These findings are thus encouraging for patients, providers, and program implementers involved in the scale-up of HIV care and treatment service delivery in some of the most affected areas of the world.
Our study also has limitations. We had to exclude a substantial number of patients due to lack of follow-up CD4 cell counts. These patients differed systematically from those who were included in the analysis: they were more likely to die or be lost before a follow-up CD4 cell count could be measured. Our analysis will thus probably overestimate the true impact of ART on CD4 response in all patients initiating ART, particularly in the first 6 months after ART initiation. This is illustrated by the CD4 trajectories of patients who were known to have died or were lost to follow-up (Fig. 2). However, because excluded patients had no follow-up CD4 cell count measurements, we had no way of taking this into account in our analysis. Of note, the median baseline CD4 cell count among patients excluded from analyses because follow-up CD4 cell counts were not measured was 82 cells/μl for patients who died and 117 cells/μl for patients lost to follow-up. This is very comparable to the median counts in patients included in analyses who died or were lost to follow-up (79 and 121 cells/μl, respectively). Nevertheless, we emphasize the point that our findings apply only to those patients who survived and remained in care with follow-up CD4 cell counts.
The range of calendar years during which participating sites contributed data varied. Although there was substantial overlap between sites, data were not contributed uniformly over the entire time period. Although we controlled for site and calendar year of ART initiation in our statistical model, it is possible that this may have influenced the results. Finally, the sites participating in the ART-LINC collaboration of IeDEA represent a convenience sample, which are unlikely to be representative of all treatment programs in their respective countries. Although there are no data that we can use to assess this directly, the ART-LINC sites are almost exclusively in urban settings and more likely to represent secondary and tertiary care facilities or centers of excellence than other sites not participating in the collaboration. Similar studies using data from other scale-up sites in primary health centers and rural settings are therefore needed.
In conclusion, our study demonstrates robust CD4 response to ART among patients in multiple treatment programs in resource-limited settings. The response appears to be sustained among those remaining in programs for up to 5 years. Our results thus support the notion that a programmatic, public health approach to scaling up ART in resource-limited settings using a limited repertoire of drugs can result in sustained immunologic and virologic outcomes that are comparable with those in industrialized countries . However, given that the most important determinant of long-term CD4 response was the baseline CD4 cell count at ART initiation, our data also suggest that many patients in developing countries must be started much earlier in order to achieve an optimal on treatment CD4 cell count over the long term. In light of evidence suggesting normal life expectancy among ART patients who achieve a CD4 cell count of 500 cells/μl or higher  programs and clinicians in resource-limited settings should strive to initiate more patients on ART at higher CD4 cell counts, such that CD4 levels closer to 500 cells/μl can be achieved after 3–5 years of treatment. Thus, our study has important implications for development of guidelines and for future trials regarding when to start ART.
We are indebted to the patients and clinic staff for their contributions to these research efforts.
Sponsorship: The ART-LINC Collaboration of the International epidemiological Databases to Evaluate AIDS is funded by the US National Institutes of Health (Office of AIDS Research and National Institute of Allergy and Infectious Diseases) and the French Agence Nationale de Recherches sur le Sida.
Roles of authors: D.N., M.W.G.B., and M.E. designed the analysis and wrote the paper. M.W.G.B., M.K., and D.N. conducted the statistical analysis, and M.W.G.B. conducted the statistical modeling and multiple imputation. M.M. and R.H. contributed to the statistical modeling approach. O.K., F.D., R.W., E.S., M.S., and M.E. contributed ideas for analysis and to the writing and editing of the manuscript.
Writing committee: Denis Nash, Monica Katyal, Martin W.G. Brinkhof, Olivia Keiser, Margaret May, Rachael Hughes, Francois Dabis, Robin Wood, Eduardo Sprinz, Mauro Schechter, Matthias Egger for the ART-LINC Collaboration of IeDEA.
Central coordinating team: Martin W.G. Brinkhof, Eric Balestre, Claire Graber (project manager), Catherine Seyler, Hapsatou Touré, François Dabis (principal investigator), Matthias Egger (principal investigator), Mauro Schechter (principal investigator).
Steering committee: Kathryn Anastos (Kigali), David Bangsberg (Mbarara/Kampala), Andrew Boulle (Cape Town), Jennipher Chisanga (Lusaka), Eric Delaporte (Dakar), Diana Dickinson (Gaborone), Ernest Ekong (Lagos), Kamal Marhoum El Filali (Casablanca), Mina Hosseinipour (Lilongwe), Charles Kabugo (Kampala), Silvester Kimaiyo (Eldoret), Mana Khongphatthanayothin (Bangkok), N. Kumarasamy (Chennai), Christian Laurent (Yaounde), Ruedi Luthy (Harare), James McIntyre (Johannesburg), Timothy Meade (Lusaka), Eugene Messou (Abidjan), Denis Nash (New York), Winstone Nyandiko Mokaya (Eldoret), Margaret Pascooe (Harare), Larry Pepper (Mbarara), Papa Salif Sow (Dakar), Sam Phiri (Lilongwe), Mauro Schechter (Rio de Janeiro), John Sidle (Eldoret), Eduardo Sprinz (Porto Alegre), Besigin Tonwe-Gold (Abidjan), Siaka Touré (Abidjan), Stefaan Van der Borght (Amsterdam), Ralf Weigel (Lilongwe), Robin Wood (Cape Town).
Participating centers: Service des Maladies Infectieuses, Casablanca, Maroc; Moi Teaching and Referral Hospital, Eldoret, Kenya; Immune Suppression Syndrome clinic, Mbarara, Uganda; Generic Antiretroviral Treatment Project, Kampala, Uganda; Centre de Prise en Charge, de Recherche et de Formation sur le VIH/SIDA (CEPREF), Abidjan, Côte d'Ivoire; ANRS 1215/1290 Study Group, Dakar, Senegal; Independent Surgery, Gaborone, Botswana; Lighthouse Trust Clinic, Lilongwe, Malawi; Gugulethu ART Programme, Gugulethu, South Africa; Perinatal HIV Research Unit (PHRU), Soweto, South Africa; Khayelitsha ART Programme, Khayelitsha, South Africa; Helen Joseph Hospital Themba Lethu Clinic, Johannesburg, South Africa; CorpMed Medical Centre, Lusaka, Zambia; Connaught Clinic, Harare, Zimbabwe; Prospective Evaluation in the Use and Monitoring of Antiretrovirals in Argentina (PUMA), Buenos Aires, Argentina; Rio HIV Cohort, Rio de Janeiro, Brazil; South Brazil HIV Cohort (SOBRHIV), Hospital de Clinicas, Porto Alegre, Brazil; YRG Care, Chennai, India; Thai Red Cross AIDS Research Centre, Bangkok, Thailand; International Center for AIDS Care and Treatment Programs, MTCT-Plus Initiative, Mailman School of Public Health, Columbia University, New York, USA; Heineken ART Programme, Amsterdam, The Netherlands.
1. WHO. Towards Universal access: scaling up priority HIV/AIDS interventions in the health sector. Geneva: WHO; 2008.
2. Wolbers M, Battegay M, Hirschel B, Furrer H, Cavassini M, Hasse B, et al
. CD4+ T-cell count increase in HIV-1-infected patients with suppressed viral load within 1 year after start of antiretroviral therapy. Antivir Ther 2007; 12:889–897.
3. Moore RD, Keruly JC. CD4+ cell count 6 years after commencement of highly active antiretroviral therapy in persons with sustained virologic suppression. Clin Infect Dis 2007; 44:441–446.
4. Mocroft A, Phillips AN, Gatell J, Ledergerber B, Fisher M, Clumeck N, et al
. Normalisation of CD4 counts in patients with HIV-1 infection and maximum virological suppression who are taking combination antiretroviral therapy: an observational cohort study. Lancet 2007; 370:407–413.
5. Hecht FM, Wang L, Collier A, Little S, Markowitz M, Margolick J, et al
. A multicenter observational study of the potential benefits of initiating combination antiretroviral therapy during acute HIV infection. J Infect Dis 2006; 194:725–733.
6. Goicoechea M, Smith DM, Liu L, May S, Tenorio AR, Ignacio CC, et al
. Determinants of CD4+ T cell recovery during suppressive antiretroviral therapy: association of immune activation, T cell maturation markers, and cellular HIV-1 DNA. J Infect Dis 2006; 194:29–37.
7. Kaufmann GR, Furrer H, Ledergerber B, Perrin L, Opravil M, Vernazza P, et al
. Characteristics, determinants, and clinical relevance of CD4 T cell recovery to <500 cells/microL in HIV type 1-infected individuals receiving potent antiretroviral therapy. Clin Infect Dis 2005; 41:361–372.
8. Bonnet F, Thiebaut R, Chene G, Neau D, Pellegrin JL, Mercie P, et al
. Determinants of clinical progression in antiretroviral-naive HIV-infected patients starting highly active antiretroviral therapy. Aquitaine Cohort, France, 1996–2002. HIV Med 2005; 6:198–205.
9. Smith CJ, Sabin CA, Youle MS, Kinloch-de Loes S, Lampe FC, Madge S, et al
. Factors influencing increases in CD4 cell counts of HIV-positive persons receiving long-term highly active antiretroviral therapy. J Infect Dis 2004; 190:1860–1868.
10. Pulido F, Arribas JR, Miro JM, Costa MA, Gonzalez J, Rubio R, et al
. Clinical, virologic, and immunologic response to efavirenz-or protease inhibitor-based highly active antiretroviral therapy in a cohort of antiretroviral-naive patients with advanced HIV infection (EfaVIP 2 study). J Acquir Immune Defic Syndr 2004; 35:343–350.
11. Garcia F, de Lazzari E, Plana M, Castro P, Mestre G, Nomdedeu M, et al
. Long-term CD4+ T-cell response to highly active antiretroviral therapy according to baseline CD4+ T-cell count. J Acquir Immune Defic Syndr 2004; 36:702–713.
12. Al-Harthi L, Voris J, Patterson BK, Becker S, Eron J, Smith KY, et al
. Evaluation of the impact of highly active antiretroviral therapy on immune recovery in antiretroviral naive patients. HIV Med 2004; 5:55–65.
13. Kaufmann GR, Perrin L, Pantaleo G, Opravil M, Furrer H, Telenti A, et al
. CD4 T-lymphocyte recovery in individuals with advanced HIV-1 infection receiving potent antiretroviral therapy for 4 years: the Swiss HIV Cohort Study. Arch Intern Med 2003; 163:2187–2195.
14. Hunt PW, Deeks SG, Rodriguez B, Valdez H, Shade SB, Abrams DI, et al
. Continued CD4 cell count increases in HIV-infected adults experiencing 4 years of viral suppression on antiretroviral therapy. AIDS 2003; 17:1907–1915.
15. Demeter LM, Hughes MD, Coombs RW, Jackson JB, Grimes JM, Bosch RJ, et al
. Predictors of virologic and clinical outcomes in HIV-1-infected patients receiving concurrent treatment with indinavir, zidovudine, and lamivudine. AIDS Clinical Trials Group Protocol 320. Ann Intern Med 2001; 135:954–964.
16. Kaufmann GR, Bloch M, Zaunders JJ, Smith D, Cooper DA. Long-term immunological response in HIV-1-infected subjects receiving potent antiretroviral therapy. AIDS 2000; 14:959–969.
17. Staszewski S, Miller V, Sabin C, Schlecht C, Gute P, Stamm S, et al
. Determinants of sustainable CD4 lymphocyte count increases in response to antiretroviral therapy. AIDS 1999; 13:951–956.
18. Gras L, Kesselring A, Griffin J, van Sighem A, Fraser C, Ghani A, et al
. CD4 cell counts of 800 cells/mm3
or greater after 7 years of highly active antiretroviral therapy are feasible in most patients starting with 350 cells/mm3
or greater. J Acquir Immune Defic Syndr 2007; 45:183–192.
19. Chene G, Sterne JA, May M, Costagliola D, Ledergerber B, Phillips AN, et al
. Prognostic importance of initial response in HIV-1 infected patients starting potent antiretroviral therapy: analysis of prospective studies. Lancet 2003; 362:679–686.
20. Michael CG, Kirk O, Mathiesen L, Nielsen SD. The naive CD4+ count in HIV-1-infected patients at time of initiation of highly active antiretroviral therapy is strongly associated with the level of immunological recovery. Scand J Infect Dis 2002; 34:45–49.
21. Kaufmann GR, Bloch M, Finlayson R, Zaunders J, Smith D, Cooper DA. The extent of HIV-1-related immunodeficiency and age predict the long-term CD4 T lymphocyte response to potent antiretroviral therapy. AIDS 2002; 16:359–367.
22. Rizzardi GP, Tambussi G, Bart PA, Chapuis AG, Lazzarin A, Pantaleo G. Virological and immunological responses to HAART in asymptomatic therapy-naive HIV-1-infected subjects according to CD4 cell count. AIDS 2000; 14:2257–2263.
23. Egger M, May M, Chene G, Phillips AN, Ledergerber B, Dabis F, et al
. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360:119–129.
24. Baillargeon J, Grady J, Borucki MJ. Immunological predictors of HIV-related survival. Int J STD AIDS 1999; 10:467–470.
25. Lewden C, Chene G, Morlat P, Raffi F, Dupon M, Dellamonica P, et al
. HIV-infected adults with a CD4 cell count greater than 500 cells/mm3
on long-term combination antiretroviral therapy reach same mortality rates as the general population. J Acquir Immune Defic Syndr 2007; 46:72–77.
26. Sow PS, Otieno LF, Bissagnene E, Kityo C, Bennink R, Clevenbergh P, et al
. Implementation of an antiretroviral access program for HIV-1-infected individuals in resource-limited settings: clinical results from 4 African countries. J Acquir Immune Defic Syndr 2007; 44:262–267.
27. Charalambous S, Innes C, Muirhead D, Kumaranayake L, Fielding K, Pemba L, et al
. Evaluation of a workplace HIV treatment programme in South Africa. AIDS 2007; 21(Suppl 3):S73–S78.
28. Stringer JS, Zulu I, Levy J, Stringer EM, Mwango A, Chi BH, et al
. Rapid scale-up of antiretroviral therapy at primary care sites in Zambia: feasibility and early outcomes. JAMA 2006; 296:782–793.
29. Lawn SD, Myer L, Bekker LG, Wood R. CD4 cell count recovery among HIV-infected patients with very advanced immunodeficiency commencing antiretroviral treatment in sub-Saharan Africa. BMC Infect Dis 2006; 6:59.
30. Erhabor O, Ejele OA, Nwauche CA. The effects of highly active antiretroviral therapy (HAART) of stavudine, lamivudine and nevirapine on the CD4 lymphocyte count of HIV-infected Africans: the Nigerian experience. Niger J Clin Pract 2006; 9:128–133.
31. Ferradini L, Jeannin A, Pinoges L, Izopet J, Odhiambo D, Mankhambo L, et al
. Scaling up of highly active antiretroviral therapy in a rural district of Malawi: an effectiveness assessment. Lancet 2006; 367:1335–1342.
32. Laurent C, Ngom Gueye NF, Ndour CT, Gueye PM, Diouf M, Diakhate N, et al
. Long-term benefits of highly active antiretroviral therapy in Senegalese HIV-1-infected adults. J Acquir Immune Defic Syndr 2005; 38:14–17.
33. Kilaru KR, Kumar A, Sippy N, Carter AO, Roach TC. Immunological and virological responses to highly active antiretroviral therapy in a nonclinical trial setting in a developing Caribbean country. HIV Med 2006; 7:99–104.
34. Tuboi SH, Harrison LH, Sprinz E, Albernaz RK, Schechter M. Predictors of virologic failure in HIV-1-infected patients starting highly active antiretroviral therapy in Porto Alegre, Brazil. J Acquir Immune Defic Syndr 2005; 40:324–328.
35. Marins JR, Jamal LF, Chen SY, Barros MB, Hudes ES, Barbosa AA, et al
. Dramatic improvement in survival among adult Brazilian AIDS patients. AIDS 2003; 17:1675–1682.
36. Dai Y, Qiu ZF, Li TS, Han Y, Zuo LY, Xie J, et al
. Clinical outcomes and immune reconstitution in 103 advanced AIDS patients undergoing 12-month highly active antiretroviral therapy. Chin Med J (Engl) 2006; 119:1677–1682.
37. Srasuebkul P, Ungsedhapand C, Ruxrungtham K, Boyd MA, Phanuphak P, Cooper DA, Law MG, et al
. Predictive factors for immunological and virological endpoints in Thai patients receiving combination antiretroviral treatment. HIV Med 2007; 8:46–54.
38. Madec Y, Laureillard D, Pinoges L, Fernandez M, Prak N, Ngeth C, et al
. Response to highly active antiretroviral therapy among severely immuno-compromised HIV-infected patients in Cambodia. AIDS 2007; 21:351–359.
39. Dabis F, Balestre E, Braitstein P, Miotti P, Brinkhof WG, Schneider M, et al
. Cohort Profile: Antiretroviral Therapy in Lower Income Countries (ART-LINC): international collaboration of treatment cohorts. Int J Epidemiol 2005; 34:979–986.
40. Braitstein P, Brinkhof MWG, Dabis F, Schechter M, Boulle A, Miotti P, et al
. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet 2006; 367:817–824.
41. Brinkhof MWG, Dabis F, Myer L, Bangsberg DR, Boulle A, Nash D, et al
. Early loss to program in HIV-infected patients starting potent antiretroviral therapy in lower-income countries. Bull World Health Organ 2008; 86:559–567.
42. Royston P, Sauerbrei W. A new approach to modelling interactions between treatment and continuous covariates in clinical trials by using fractional polynomials. Stat Med 2004; 23:2509–2525.
43. Royston P, Altman D. Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling. Appl Stat 1994; 43:429–467.
44. Wood AM, White IR, Royston P. How should variable selection be performed with multiply imputed data? Stat Med 2008; 27:3227–3246.
45. van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of missing blood pressure covariates in survival analysis. Stat Med 1999; 18:681–694.
46. Le Moing V, Thiebaut R, Chene G, Leport C, Cailleton V, Michelet C, et al
. Predictors of long-term increase in CD4(+) cell counts in human immunodeficiency virus-infected patients receiving a protease inhibitor-containing antiretroviral regimen. J Infect Dis 2002; 185:471–480.
47. Kabugo C, Bahendeka S, Mwebaze R, Malamba S, Katuntu D, Downing R, et al
. Long-term experience providing antiretroviral drugs in a fee-for-service HIV clinic in Uganda: evidence of extended virologic and CD4+ cell count responses. J Acquir Immune Defic Syndr 2005; 38:578–583.
48. Etard JF, Ndiaye I, Thierry-Mieg M, Gueye NF, Gueye PM, Laniece I, et al
. Mortality and causes of death in adults receiving highly active antiretroviral therapy in Senegal: a 7-year cohort study. AIDS 2006; 20:1181–1189.
49. Badri M, Lawn SD, Wood R. Short-term risk of AIDS or death in people infected with HIV-1 before antiretroviral therapy in South Africa: a longitudinal study. Lancet 2006; 368:1254–1259.
50. Lawn SD, Myer L, Orrell C, Bekker LG, Wood R. Early mortality among adults accessing a community-based antiretroviral service in South Africa: implications for programme design. AIDS 2005; 19:2141–2148.
51. Keiser O, Anastos K, Schechter M, Balestre E, Myer L, Boulle A, et al
. Antiretroviral therapy in resource-limited settings, 1996 to 2006: patient characteristics, treatment regimens and monitoring in sub-Saharan Africa, Asia and Latin America. Trop Med Int Health 2008; 13:870–879.
52. Lawn SD, Myer L, Bekker LG, Wood R. Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control. AIDS 2006; 20:1605–1612.
53. El-Sadr WM, Lundgren JD, Neaton JD, Gordin F, Abrams D, Arduino RC, et al
. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med 2006; 355:2283–2296.
54. Braitstein P, Boulle A, Nash D, Brinkhof MW, Dabis F, Laurent C, et al
. Gender and the use of antiretroviral treatment in resource-constrained settings: findings from a multicenter collaboration. J Womens Health (Larchmt) 2008; 17:47–55.
55. Collazos J, Asensi V, Carton JA. Sex differences in the clinical, immunological and virological parameters of HIV-infected patients treated with HAART. AIDS 2007; 21:835–843.
56. Smit C, Hallett T, Lange J, Garnett G, de Wolf F. Late entry to HIV care limits the impact of anti-retroviral therapy in the Netherlands. PLoS One 2008; 3:1–4.
57. WHO. Antiretroviral therapy for HIV infection in infants and children: towards universal access. Recommendations for a public health approach
. Geneva: WHO; 2006.
58. Keiser O, Orrell C, Egger M, Wood R, Brinkhof MWG, Furrer H, et al
. Public health and individual approach to antiretroviral therapy: township South Africa and Switzerland compared. PLoS Med 2008; 5:e148.