It has been more than 15 years since the introduction of the highly active antiretroviral therapy (HAART). By reducing HIV-related morbidity and mortality,1-3 HAART has transformed HIV-1 infection to a manageable chronic condition.4 In developed countries, over 70% of HIV-infected individuals have received HAART.4-6 CD4+ T-cell counts (CD4+ counts) and plasma HIV-1 RNA (RNA) concentration after HAART initiation strongly predicts AIDS and death,7-11 highlighting the importance of characterizing these biomarkers post HAART. Most published reports have studied CD4+ counts and RNA in the first 5 years after HAART initiation,7,10,12-14 and the few that have studied these markers after 5 years of HAART15-18 have not delineated the factors related to their heterogeneity with long-term HAART use.
We describe the CD4+ count and plasma RNA distributions and identify factors associated with these 2 biomarkers, at 5-12 years after HAART initiation in the Multicenter AIDS Cohort Study (MACS). We hypothesized that age, HIV-1 disease stage before HAART initiation, immunological and virological responses to HAART in the first 5 years after HAART, and types of HAART regimens would significantly affect later CD4+ count and RNA levels among long-term HAART users.
MATERIALS AND METHODS
Population and Study Design
The MACS is an ongoing prospective study of HIV-1 infection among men who have sex with men in the United States.19-21 A total of 6972 men were recruited: 5622 before 1991 and 1350 in 2001-2003 in Baltimore, MD; Chicago, IL; Los Angeles, CA; and Pittsburgh, PA. Study questionnaires are available at http://www.statepi.jhsph.edu/macs/forms.html. MACS study protocols were approved by institutional review boards at each center, and informed consent was obtained from all participants.
Six monthly visits include standardized interviews, physical examinations, and collection of blood for concomitant laboratory testing and storage. Positive enzyme immunoassays with confirmatory Western blot tests were used to determine HIV-1 seropositivity. RNA levels were determined using Roche Ultrasensitive RNA PCR assay (Hoffman-LaRoche, Nutley, NJ) with a detection limit of 50 copies per milliliter, and CD4+ levels were quantified using standardized flow cytometry.22,23 For 2001-2003 recruits who initiated HAART before study entry, historical information on CD4+ counts, RNA measurements, and use of antiretroviral therapy (ART) was abstracted from medical records. Hepatitis B virus (HBV) infection was determined by a positive surface antigen or core antibody. Hepatitis C virus (HCV) infection was defined by a positive antibody. Presence of an AIDS-defining illness (AIDS) was based on the clinical conditions listed in the 1993 Centers for Disease Control and Prevention case definition24 but did not include cases identified only by low CD4+ counts. Self-reported use of ART at each visit was summarized to define use of HAART according to the DHHS/Kaiser Panel guidelines.25 The date of HAART initiation was the midpoint between the last study visit without report of HAART use (last no-HAART) and the first visit at which HAART use was reported (first HAART). Adherence to HAART was dichotomized as 100% or less than 100% based on self-reported adherence to current medications over the last 4 days.26
We used data collected up to September 30, 2009 (end of visit 51). Specifically, our study population was restricted to HIV-positive men who (1) initiated HAART on or after July 1, 1995, with ≤1 year between the last no HAART and first HAART visits; (2) reported using HAART for ≥80% of the visits post HAART initiation; and (3) had ≥1 CD4+ count and RNA measurement available at 5-12 years after HAART initiation. The measurements at visits where participants did not report using HAART (n = 393 person-visits) were excluded from the analysis.
We examined 3 outcome variables: (1) continuous CD4+ count, (2) binary indicator of suppressed RNA (<50 copies/mL), and (3) log10 (RNA) for those with detectable RNA measured at 5-12 years after HAART initiation. CD4+ counts above 1500 cells per microliter (ie, above the normal range) were set to 1500 to reduce variance.
Linear regression analyses were conducted for the continuously distributed outcomes, and logistic regression was used for RNA suppression. Factors with a P value ≤0.1 were kept in final models. SAS GENMOD procedure (SAS Institute, Cary, NC) with generalized estimating equations was used to control for within-subject correlation.
For each outcome, 2 models were used. First, we examined pre-HAART factors including last known CD4+ count (≤200, 201-350, 351-500, or >500 cells/μL), RNA level (≤1000, 1001-10,000, 10,001-100,000, or >100,000 copies/mL), AIDS (yes/no), experience of any ART, HBV coinfection, and HCV coinfection. We also examined factors in the first 5 years on HAART, including CD4+ slope (per 50 cells/μL change per year), <50 RNA copies per milliliter for ≥50% of measurements (yes/no), and total number of times regimens were switched (0, 1, 2+). Age at HAART initiation (<40, 40-49.9, or ≥50 years), race, cohort entry (pre 2001 or 2001-2003), calendar year of HAART initiation, number of years since HAART initiation (each year treated linearly, centered at 8 years), and adherence to HAART (100% vs. <100%) in the 4 days before the study visit were also evaluated. MACS center was included to control for site variability. Because CD4+ measurements may be difficult to obtain in resource-limited settings, we also examined the effects of total lymphocyte count (TLC) below 1200 cells per microliter (yes/no) before HAART and its occurrence in the first 5 years on HAART.27 For this analysis, we excluded 2001-2003 recruits who initiated HAART before study entry because pretherapy TLC was not available. For the analysis of TLC as predictor, we excluded the CD4+ count as a covariate.
Second, we assessed the effects of HAART regimens used between visit(i) and visit(i+1) on the biomarkers at visit(i+1). We examined the 3 backbones of HAART: (1) protease inhibitor (PI)-HAART (HAART regimens containing any PI, taken as the reference group), (2) nonnucleoside reverse transcriptase inhibitor (NNRTI)-HAART (HAART regimens containing any NNRTI without PI), (3) triple nucleoside reverse transcriptase inhibitor (NRTI)-HAART (HAART regimens containing only NRTIs). We then examined the 15 most frequently used HAART regimens by replacing the HAART backbones in the model with an HAART regimen variable (15 categories). We also included race, center, year of cohort entry (pre-2001, 2001-2003), pre-HAART HBV and HCV coinfections, calendar year of HAART initiation, and number of years since HAART initiation. Because current disease stage and adherence may predict regimen type, we also included time-varying covariates that were updated at each visit: age (<40, 40-49.9, or ≥50 years), CD4+ count and RNA detection at the prior visit, cumulative number of times regimens were switched between HAART initiation and the prior visit (0, 1, 2+), and adherence. We also repeated the analyses, with restriction to the men who were ART naive before HAART initiation.
A flow chart of the study population is shown in Supplemental Digital Content 1 (see Figure, http://links.lww.com/QAI/A169); 614 MACS participants contributed 4431 CD4+ and RNA measurements while using HAART for 5-12 years. At HAART initiation, 47% were younger than 40 and 12% were older than 50 years. Before initiating HAART, 40% were ART naive and more than half had <350 CD4+ cells per microliter and ≥10,000 RNA copies per milliliter. In the first 5 years on HAART, 70% had a positive CD4+ slope. The 614 men analyzed were similar to the men excluded due to missing data (n = 188) or on HAART for <80% of visits (n = 200) in terms of the calendar year and age at HAART initiation, race, cohort entry, CD4+ and RNA levels, and proportion of ART naive before HAART initiation.
Of the 4431 person-visits with CD4+ counts and RNA measurements, 53% were classified as PI-HAART related. Of those visits, ritonavir-boosted regimens were reported at 70%, with lopinavir/ritonavir (Kaletra) accounting for about one-third. The other boosted PIs were atazanavir (25%), indinavir (17%), and saquinavir (9%). Among the 42% NNRTI-based HAART person-visits, efavirenz was reported at 67% and nevirapine at 32%. Triple NRTI-HAART regimens were reported at 5% of person-visits. Figure 1 shows the frequency of the most reported HAART regimens by calendar time.
CD4+ T-Cell Count at 5-12 Years After HAART Initiation
Sixty-three percent of the 614 men contributed at least 5 CD4+ counts 5-12 years after HAART initiation (Table 1). The unadjusted median CD4+ count during this interval was 585 (interquartile range, 412-791) cells per microliter. Figure 2 shows the trajectories of median CD4+ count from pre-HAART to 12 years after HAART initiation and the overall adjusted mean CD4+ count in the 5-12 years by pre-HAART levels. In the first 5 years after HAART initiation, the CD4+ counts in each group increased significantly (P < 0.01) but were stable at 5-12 years on HAART for those with >350 CD4+ cells per microliter pre-HAART. Adjusting for factors shown in Table 2, men with ≤350 CD4+ cells per microliter pre-HAART had a mean annual increase of 11 cells per microliter (P < 0.01) while on HAART for 5-12 years.
In multivariate analysis (Table 2), men with lower pre-HAART CD4+ counts or with HBV or HCV coinfections had lower counts at 5-12 years on HAART. CD4+ slope in the first 5 years on HAART predicted subsequent CD4+ counts only for those with <500 CD4+ cells per microliter before starting HAART. Those who suppressed virus for ≥50% of when measured or remained on their initial or second regimen (compared with those who switched their HAART regimens ≥ 2 times) in the first 5 years on HAART had significantly higher CD4+ counts in the subsequent 8 years.
Age at HAART initiation also affected long-term CD4+ counts. The adjusted mean CD4+ count for those 50 years and older at HAART initiation was significantly lower than that for participants younger than 40 years. Men younger than 40 years who started HAART with 201-350 CD4+ cells per microliter (reference group) had an adjusted mean cell count of 670 cells per microliter in the 10-12 years after HAART initiation compared with 578 cells per microliter in those 50 years and older with the same starting level of CD4+ cells (P < 0.01). To achieve equivalent CD4+ counts as the younger men, those aged 50 years and older at HAART initiation needed to start HAART with 351-500 CD4+ cells per microliter (adjusted mean CD4+ counts at 10-12 years post HAART was 643 cells/μL, P = 0.45 for comparing to the reference group). Restricting to men who were ART naive before HAART initiation, older men still had 60 fewer CD4+ cells per microliter on average when compared with the younger men (Table 2). The lack of statistical significance for this difference most likely resulted from a reduced sample size. The average long-term CD4+ count (653 cells/μL) for men aged 40-49.9 years who initiated HAART with 201-350 CD4+ cells per microliter did not differ significantly (P = 0.39) from that of the younger men starting with the same cell count.
In a separate model to examine the effect of TLC on long-term CD4+ counts, those with pre-HAART TLC <1200 cells per microliter had an adjusted mean CD4+ count of 514 cells per microliter at 10-12 years on HAART, which was 122 cells per microliter lower than that of those with ≥1200 cells per microliter (P < 0.01). Similarly, those with a nadir TLC <1200 cells per microliter in the first 5 years after HAART initiation had an adjusted mean of 522 CD4+ cells per microliter at 10-12 years on HAART, 134 cells per microliter lower than those whose TLC was always above 1200 cells per microliter in the first 5 years. The effects of the other factors examined persisted in this model except that the effect of having switched regimens ≥2 times was smaller (−11 cells/μL with P = 0.74, compared with the reference group).
When examining the effects of recently used HAART regimens on CD4+ counts, we controlled for potential confounding by including time-varying age, CD4+ counts, detection of RNA, cumulative number of times switched regimen from HAART initiation to the prior visit, and adherence. Race, study center, year of cohort entry, and pre-HAART HBV and HCV coinfection status were also included. The adjusted CD4+ counts did not differ significantly for those reporting PI-, NNRTI-, or triple NRTI-based HAART regimens. Furthermore, specific HAART regimens did not differentiate the CD4+ counts at 5-12 years after HAART initiation (data not shown).
Because follow-up times varied across the cohort and individuals followed for 10 years may have differed from those with less follow-up, we repeated the above analyses with restriction to the 314 men who were followed for at least 10 years under HAART and observed similar findings (see Table, Supplemental Digital Content 2, http://links.lww.com/QAI/A170). Specifically, differences in the estimates of the effects on the CD4+ count shown in Table 2 between the full and the restricted analysis were all less than 50 cells cubic millimeter and within the confidence limits of the estimates. Furthermore, the median CD4+ counts over time in Figure 2 did not differ significantly when using the restricted sample (see Figure, Supplemental Digital Content 3, http://links.lww.com/QAI/A171).
Factors Associated With Suppressed RNA at 5-12 Years After HAART Initiation
Overall, 78% of RNA measurements in 5-12 years after HAART were undetectable (<50 copies/mL) and this percentage increased over time (bottom of Fig. 2). Age, center, pre-HAART CD4+ counts, RNA levels, and AIDS status had little to no effect on this proportion after accounting for RNA in the first 5 years after initiation. Recent HAART initiators were more likely to suppress RNA (Table 3). After including this variable, the effect of being ART naive when starting HAART was no longer significant. Participants enrolled in 2001-2003 also were more likely to suppress RNA when compared with the pre-2001 recruits (Table 3). However, among men who were ART naive at HAART initiation, cohort entry year did not affect long-term RNA suppression and the effect of switching regimens also attenuated. As shown in Table 3, the RNA response to HAART in the first 5 years after initiation predicted subsequent viral suppression, as did the number of times regimens were switched in those 5 years and current adherence to HAART. Non-white participants were less likely to suppress RNA, significantly so for the African American group [odds ratio (OR) = 0.42, P < 0.01]. RNA suppression was more likely with increased time on HAART (OR = 1.28 per year increase, P < 0.01).
In a separate model, persons using NNRTI-based regimens were as likely to suppress RNA as those on PI-based regimens, controlling for current age (OR = 1.47, P = 0.04 for comparing age ≥50 vs. <40), race, CD4+ counts and RNA detectability at the prior visit, adherence, and number of years after HAART initiation. Participants using triple NRTI-HAART were significantly less likely to suppress RNA than those using PI-HAART (OR = 0.42, P < 0.01) or NNRTI-HAART (OR = 0.40, P < 0.01). Similar results were observed among those who were ART naive at HAART initiation (data not shown). Compared with the reference regimen (tenofovir + emtricitabine + efavirenz), some regimens containing nevirapine were associated with a greater likelihood of detectable RNA. We then restricted the analyses to those using NNRTI regimens (without any PIs) and observed lower odds of viral suppression for those using nevirapine compared with efavirenz (OR = 0.67, P = 0.08). Not surprisingly, a higher percentage (80%) of undetectable RNA was observed among those with complete adherence compared with 70% for those with less adherence (adjusted OR was 1.34, P < 0.01).
Restricting the analysis to those 314 men who were followed for ≥10 years under HAART, most of the ORs for suppressed HIV RNA were further from null, but all were in the same direction as for the full population (see Table, Supplemental Digital Content 4, http://links.lww.com/QAI/A172).
Factors Associated With Level of Detectable RNA
Among the 22% (n = 970) of person-visits with detectable RNA while on HAART for 5-12 years, the median RNA was 958 (interquartile range, 149-9180) copies per milliliter. RNA level before HAART initiation affected long-term RNA level; adjusted median RNA at 5-12 years was 955 and 330 copies per milliliter for those with pre-HAART RNA of >10,000 and ≤1000 copies per milliliter, respectively (P < 0.01). The effect of RNA while under HAART in the first 5 years on subsequent RNA levels was also significant; adjusted median RNA at 5-12 years after HAART initiation was 1181 and 223 copies per milliliter for those with <50% and ≥50% undetectable RNA measurements, respectively (P < 0.01).
In this study, several factors were found to be associated with lower CD4+ counts in men who had received HAART for 5-12 years. Important modifiable factors were older age (>50 years) and lower CD4+ counts at the time of HAART initiation. Other significant factors were the annual change in CD4+ count in the first 5 years of HAART (for those who initiated HAART with <500 CD4+ cells/μL), the number of switches in HAART regimens in the first 5 years, and HBV and HCV infections.
The importance of age and baseline CD4+ count at HAART initiation in the immunological response to HAART 5-12 years later extends the findings of Kaufmann et al12 that these parameters were significantly associated with CD4+ count at 5 years after HAART. In the present study, we show that to have an equivalent long-term immunological response as younger men, older men needed to start HAART at higher CD4+ counts. Specifically, those 50 years or older who initiated HAART with 351-500 CD4+ cells per microliter had equivalent CD4+ counts at 10-12 years after HAART to those younger than 40 years who started HAART with 201-350 T cells per microliter. Adjusting for age, we also show that long-term CD4+ counts in men who started HAART at different CD4+ count levels follow almost parallel trajectories. MACS investigators also reported similar parallel trajectories over the 3½ years after HAART initiation3 and in a joint analysis with the Women's Interagency HIV-1 Study (WIHS) up to 7 years post HAART.15 In the present study, we show that CD4+ counts increased during the years 5-12 on HAART among men who started HAART with ≤350 CD4+ cells per microliter, consistent with the earlier report from the MACS and WIHS15 and with a recent report16 that showed that those with lower pre-HAART CD4+ counts took longer to reach CD4+ >500 cells per microliter than those with higher counts. Age also was negatively associated with CD4+ count increases after 4 years of HAART.16 The association of older age with lower CD4+ count might be due to diminution of thymic T-cell production with age.28
Older age is associated with a more rapid progression to AIDS and a shorter survival both in the absence of HAART29,30 and in the presence of HAART. 2,31-34 It is also well known that immune function declines with age. Therefore, initiating HAART at relatively high CD4+ counts might be of more benefit in older HIV-1-infected individuals than in younger ones. In Europe, but not in the US,25 age >50 years is included in the guidelines35 for initiation of HAART at CD4+ counts >350 cells per microliter. Age may explain some of the heterogeneity in response to HAART among those starting treatment at relatively high CD4+ counts, and the data from this study support the inclusion of age in treatment guidelines.
In the present study, the median CD4+ count was 585 cells per microliter and RNA was undetectable (<50 copies/mL) in 78% of the measurements that occurred 5-12 years under HAART. Among the 22% with detectable RNA, the median was 958 copies per milliliter. Given that more than half of the men had <350 CD4+ cells per microliter and almost 90% had ≥1000 RNA copies per milliliter when HAART was initiated, this study shows that the effectiveness of HAART persists for up to 12 years.
Immunological and virological responses in the first 5 years after HAART initiation strongly influenced these responses in the subsequent 5-7 years. Also, individuals who responded well to the first or second HAART regimens maintained higher CD4+ counts and lower HIV RNA than those who switched regimens at least twice. This may reflect a suboptimal response to HAART, the presence of drug-resistant viral strains, and/or adverse effects of HAART that limit therapy.36
TLC, stratified as greater or less than 1200 cells per microliter, was evaluated as a surrogate for CD4+ count, and a lower TLC was also associated with lower CD4+ counts at 5-12 years on HAART. This finding supports inclusion of TLC in HIV treatment guidelines, as in the Joint United Nations Programme on HIV/AIDS (UNAIDS) guidelines, and may be useful for regions where measurement of CD4+ counts is not available. However, factors that can affect TLC such as infectious diseases and nutritional deficiencies37,38 may not be present in the MACS. Therefore, optimal cutoff values for use of TLC in treatment guidelines might vary regionally.
We did not observe significant differences in long-term CD4+ counts by race, although African Americans were approximately 50% less likely than whites to suppress RNA. Virologic and immunologic responses to HAART did not significantly differ by race in the WIHS.39 However, we and others have reported racial differences in adherence to HAART,26,40 which predicts better RNA suppression.41-43 It is notable that decreased achievement of viral suppression in African Americans in the present study did not translate into an effect on long-term CD4+ counts. Similarly, consonant with the literature and guidelines, we observed that triple NRTI therapy25,44,45 and nevirapine relative to efavirenz46-48 were less likely to suppress HIV RNA.
The data on which the present study is based, including the characteristics before HAART was initiated, were obtained using standardized longitudinal methods That is, they were collected without bias that may arise when disease progression may direct the frequency of measurements that are obtained.49,50 Insufficient numbers and follow-up time limit our ability to evaluate new HAART regimens.
The results presented here were obtained after those with the fastest progression of HIV disease after initiating HAART were removed because our analysis did not begin until 5 years after HAART initiation. The goal of this study was to describe the experience of the subgroup of HAART initiators who are long-term users. Thus, the present results are generalizable only to such people. This restriction should represent a conservative bias because CD4+ count at HAART initiation strongly predicts progression in the first 5 years after HAART initiation,3,7-10 that is, if they had been included, the effect of CD4+ count at HAART initiation on immunological response should be even stronger. The removal of those who progressed within the first 5 years after HAART initiation may explain why this study did not observe the plateauing of CD4+ count after 2 years that was previously reported in the MACS.4 The results presented here may not be generalizable to those who have started HAART in the last 5 years. Our population started HAART early in the HAART era with a variety of initial regimens that are different from the regimens most commonly started today. A further limitation is that our population had largely received monotherapy or dual therapy before initiating HAART. However, this is probably not a major limitation because restriction of the present analysis to the 40% men who were ART naive prior to HAART initiation yielded similar trends as those observed in the overall population. To evaluate the long-term effects of new HAART regimens, analyses similar to the present study should be performed with each new cohort of HAART initiators, keeping in mind that the results of these analyses will always be limited to historical initiators likely to be unrepresentative of recent initiators.
In summary, over half of the men in this study who reported HAART use for more than 5 years had >500 CD4+ cells per microliter and three-fourths had undetectable plasma RNA. Our data support using age in the guidelines for initiating HAART, such that persons who are older than 50 years should start treatment at higher CD4+ counts, for example, 350-500 cells per microliter. This interaction of age and immunological stage when starting treatment and the effects of specific regimens on long-term response to treatment should be further evaluated by large collaborative observational cohorts.
Data in this article were collected by the MACS with centers (Principal Investigators) at The Johns Hopkins University Bloomberg School of Public Health (J. B. Margolick and L. P. Jacobson), Howard Brown Health Center and Northwestern University Medical School (J. P. Phair and Steven M. Wolinsky), University of California, Los Angeles (Roger Detels and Otto Martínez-Maza), and University of Pittsburgh (C. R. Rinaldo and Lawrence A. Kingsley). Web site located at http://www.statepi.jhsph.edu/macs/macs.html.
1. Detels R, Muñoz A, McFarlane G, et al. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. Multicenter AIDS Cohort Study Investigators. JAMA
2. Jacobson LP, Li R, Phair J, et al. Evaluation of the effectiveness of highly active antiretroviral therapy in persons with human immunodeficiency virus using biomarker-based equivalence of disease progression. Am J Epidemiol
3. Tarwater PM, Margolick JB, Jin J, et al. Increase and plateau of CD4 T-cell counts in the 3(1/2) years after initiation of potent antiretroviral therapy. J Acquir Immune Defic Syndr
4. Schneider MF, Gange SJ, Williams CM, et al. Patterns of the hazard of death after AIDS through the evolution of antiretroviral therapy: 1984-2004. AIDS
5. Lohse N, Hansen AB, Pedersen G, et al. Survival of persons with and without HIV infection in Denmark, 1995-2005. Ann Intern Med
6. Valcour VG, Sacktor NC, Paul RH, et al. Insulin resistance is associated with cognition among HIV-1-infected patients: the Hawaii Aging With HIV Cohort. J Acquir Immune Defic Syndr
7. Anastos K, Barron Y, Cohen MH, et al. The prognostic importance of changes in CD4+
cell count and HIV-1 RNA level in women after initiating highly active antiretroviral therapy. Ann Intern Med
8. Chene G, Sterne JA, May M, et al. Prognostic importance of initial response in HIV-1 infected patients starting potent antiretroviral therapy: analysis of prospective studies. Lancet
9. MacArthur RD, Perez G, Walmsley S, et al. Comparison of prognostic importance of latest CD4+
cell count and HIV RNA levels in patients with advanced HIV infection on highly active antiretroviral therapy. HIV Clin Trials
10. Tarwater PM, Gallant JE, Mellors JW, et al. Prognostic value of plasma HIV RNA among highly active antiretroviral therapy users. AIDS
11. Lanoy E, May M, Mocroft A, et al. Prognosis of patients treated with cART from 36 months after initiation, according to current and previous CD4 cell count and plasma HIV-1 RNA measurements. AIDS
12. Kaufmann GR, Furrer H, Ledergerber B, 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
13. Kaufmann GR, Perrin L, Pantaleo G, 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
14. Li X, Chu H, Gallant JE, et al. Bimodal virological response to antiretroviral therapy for HIV infection: an application using a mixture model with left censoring. J Epidemiol Community Health
15. Chu H, Gange SJ, Yamashita TE, et al. Individual variation in CD4 cell count trajectory among human immunodeficiency virus-infected men and women on long-term highly active antiretroviral therapy: an application using a Bayesian random change-point model. Am J Epidemiol
16. Kelley CF, Kitchen CM, Hunt PW, et al. Incomplete peripheral CD4+
cell count restoration in HIV-infected patients receiving long-term antiretroviral treatment. Clin Infect Dis
17. Mocroft A, Phillips AN, Gatell J, 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
18. 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
19. Detels R, Phair JP, Saah AJ, et al. Recent scientific contributions to understanding HIV/AIDS from the Multicenter AIDS Cohort Study. J Epidemiol (Japan)
20. Dudley J, Jin S, Hoover D, et al. The Multicenter AIDS Cohort Study: retention after 9 1/2 years. Am J Epidemiol
21. Kaslow RA, Ostrow DG, Detels R, et al. The Multicenter AIDS Cohort Study: rationale, organization, and selected characteristics of the participants. Am J Epidemiol
22. Giorgi JV, Cheng HL, Margolick JB, et al. Quality control in the flow cytometric measurement of T-lymphocyte subsets: the multicenter AIDS cohort study experience. The Multicenter AIDS Cohort Study Experience. Clin Immunol Immunopathol
23. Schenker EL, Hultin LE, Bauer KD, et al. Evaluation of a dual-color flow cytometry immunophenotyping panel in a multicenter quality assurance program. Cytometry
24. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Recomm Rep
25. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. Department of Health and Human Services; December 1, 2009:1-161. Available at: http://aidsinfo.nih.gov/contentfiles/AdultandAdolescentGL.pdf
. Accessed July 21, 2010.
26. Kleeberger CA, Phair JP, Strathdee SA, et al. Determinants of heterogeneous adherence to HIV-antiretroviral therapies in the Multicenter AIDS Cohort Study. J Acquir Immune Defic Syndr
27. Daka D, Loha E. Relationship between total lymphocyte count (TLC) and CD4 count among peoples living with HIV, Southern Ethiopia: a retrospective evaluation. AIDS Res Ther
28. Teixeira L, Valdez H, McCune JM, et al. Poor CD4 T-cell restoration after suppression of HIV-1 replication may reflect lower thymic function. AIDS
29. Darby SC, Ewart DW, Giangrande PL, et al. Importance of age at infection with HIV-1 for survival and development of AIDS in UK haemophilia population. UK Haemophilia Centre Directors' Organisation. Lancet
30. Butt AA, Dascomb KK, DeSalvo KB, et al. Human immunodeficiency virus infection in elderly patients. South Med J
31. Woldemichael G, Christiansen D, Thomas S, et al. Demographic characteristics and survival with AIDS: health disparities in Chicago, 1993-2001. Am J Public Health
. 2009;99(suppl 1):S118-S123.
32. Sabin CA, Smith CJ, d'Arminio Monforte A, et al. Response to combination antiretroviral therapy: variation by age. AIDS
33. Grabar S, Kousignian I, Sobel A, et al. Immunologic and clinical responses to highly active antiretroviral therapy over 50 years of age. Results from the French Hospital Database on HIV. AIDS
34. Grabar S, Weiss L, Costagliola D. HIV infection in older patients in the HAART era. J Antimicrob Chemother
36. Hammer SM, Eron JJ Jr, Reiss P, et al. Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society—USA panel. JAMA
37. Lukito W, Wattanapenpaiboon N, Savige GS, et al. Nutritional indicators, peripheral blood lymphocyte subsets and survival in an institutionalised elderly population. Asia Pac J Clin Nutr
38. Patrick L. Nutrients and HIV: part two—vitamins A and E, zinc, B-vitamins, and magnesium. Altern Med Rev
39. Anastos K, Schneider MF, Gange SJ, et al. The association of race, sociodemographic, and behavioral characteristics with response to highly active antiretroviral therapy in women. J Acquir Immune Defic Syndr
40. Heckman BD, Catz SL, Heckman TG, et al. Adherence to antiretroviral therapy in rural persons living with HIV disease in the United States. AIDS Care
41. Parienti JJ, Massari V, Descamps D, et al. Predictors of virologic failure and resistance in HIV-infected patients treated with nevirapine- or efavirenz-based antiretroviral therapy. Clin Infect Dis
42. Roge BT, Barfod TS, Kirk O, et al. Resistance profiles and adherence at primary virological failure in three different highly active antiretroviral therapy regimens: analysis of failure rates in a randomized study. HIV Med
43. Tuboi SH, Harrison LH, Sprinz E, et al. Predictors of virologic failure in HIV-1-infected patients starting highly active antiretroviral therapy in Porto Alegre, Brazil. J Acquir Immune Defic Syndr
44. Gulick RM, Ribaudo HJ, Shikuma CM, et al. Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection. N Engl J Med
45. van Leeuwen R, Katlama C, Murphy RL, et al. A randomized trial to study first-line combination therapy with or without a protease inhibitor in HIV-1-infected patients. AIDS
46. Cozzi-Lepri A, Phillips AN, d'Arminio Monforte A, et al. Virologic and immunologic response to regimens containing nevirapine or efavirenz in combination with 2 nucleoside analogues in the Italian Cohort Naive Antiretrovirals (I.Co.N.A.) study. J Infect Dis
47. Keiser P, Nassar N, White C, et al. Comparison of nevirapine- and efavirenz-containing antiretroviral regimens in antiretroviral-naive patients: a cohort study. HIV Clin Trials
48. Matthews GV, Sabin CA, Mandalia S, et al. Virological suppression at 6 months is related to choice of initial regimen in antiretroviral-naive patients: a cohort study. AIDS
49. Lau B, Gange SJ, Moore RD. Interval and clinical cohort studies: epidemiological issues. AIDS Res Hum Retroviruses
50. Lau B, Gange SJ, Kirk GD, et al. Evaluation of human immunodeficiency virus biomarkers: inferences from interval and clinical cohort studies. Epidemiology