Weintrob, Amy C MD*†; Fieberg, Ann M MS*‡; Agan, Brian K MD*§; Ganesan, Anuradha MD*‖; Crum-Cianflone, Nancy F MD, MPH*¶; Marconi, Vincent C MD*§; Roediger, Mollie MS*‡; Fraser, Susan L MD*#; Wegner, Scott A MD*; Wortmann, Glenn W MD*†
Highly active antiretroviral therapy (HAART) has decreased the morbidity and mortality associated with human immunodeficiency virus (HIV) infection1-3; however, the response to HAART is not uniform. Factors that have been associated with the response to HAART include the following: CD4 cell count and viral load (VL) at the time of HAART initiation,4 history of prior antiretroviral use,5 mode of HIV transmission,4,6 Centers for Disease Control and Prevention stage at the time of HAART initiation,4 hemoglobin levels,7 and certain genetic traits.5,8,9 In some studies, age has also been associated with response to HAART.5,10-16
In the pre-HAART era, older age at HIV seroconversion was associated with decreased survival and increased risk of progression to acquired immunodeficiency syndrome (AIDS).6,17,18 During the HAART era, however, the effect of age on HIV disease progression is less clear. Some studies have evaluated the effect of age at HIV diagnosis on disease progression,10-12 whereas others have studied the effect of age at HAART initiation on disease progression.5,13,14 A common limitation of these studies is the inability to establish the duration of HIV infection before diagnosis or HAART initiation. Potentially, patients who are diagnosed at an older age have been infected for a longer period of time, and it is that factor, rather than age per se, which affects the virologic and immunologic responses to HAART. An example of this potential bias can be found in a study by Grabar et al14 which not only reported that individuals older than 50 years at HAART initiation had smaller CD4 cell increases and faster progression to AIDS or death but also noted that the older persons had more advanced disease at HAART initiation. Without knowing the duration of HIV infection, these results are difficult to interpret.
Another difficulty when comparing studies reporting the interplay of HIV and age is the various definitions of “older age.” Although most studies divide age into 2 groups-those 50 years and younger and those older than 50 years10,11,14,15-others have used 45 years5 or 55-60 years12,13,16 as the standard. For most of these studies, age-related differences within the large group of individuals younger than 50 years were not examined.
The Tri-Service AIDS Clinical Consortium (TACC) Natural History Study (NHS) is a multicenter epidemiologic cohort study that began collecting information on HIV-infected military personnel and other Department of Defense beneficiaries in 1987 and has followed approximately 2400 individuals with defined HIV seroconversion windows. Data from this cohort were used to investigate the effect of age at HIV seroconversion on the virologic, immunologic, and clinical response to HAART.
The TACC NHS is an ongoing, prospective, cohort study of consenting military personnel and beneficiaries with HIV infection and includes participants from the Army, Navy/Marines, and Air Force. Since 1986, routine HIV testing (enzyme-linked immunosorbent assay and confirmatory Western blot) has been used to exclude HIV-infected persons from enlisting for military service or from overseas deployment. Routine testing among active duty members occurs every 1-5 years, resulting in a defined seroconversion window for incident HIV infection. Subjects with HIV infection are referred to 1 of 7 participating medical centers: Walter Reed Army Medical Center, Washington, DC; Wilford Hall Medical Center, San Antonio, TX; National Naval Medical Center, Bethesda, MD; Naval Medical Center San Diego, San Diego, CA; Brooke Army Medical Center, San Antonio, TX; Naval Medical Center Portsmouth, VA; and Tripler Army Medical Center, Honolulu, HI. At these centers, subjects receive evaluation and care and are invited to enroll as research subjects in the TACC NHS.
Those who consent to enroll in the TACC NHS are seen every 6 months at 1 of the above 7 medical centers. Data are collected on demographic characteristics, markers of HIV disease, medication use, and clinical events (including AIDS-defining illnesses according to the 1993 Centers for Disease Control and Prevention definition19) with medical record confirmation. Institutional review boards at each medical center reviewed and approved the TACC NHS. A central institutional review board reviewed and approved this analysis.
Information was collected from the database on individuals who seroconverted after January 1, 1996 (the approximate date when HAART became available in the military health care system), and included dates of the last negative and first positive HIV tests, sex, ethnicity, antiretroviral use before HAART, date of HAART initiation along with first regimen, CD4 counts and VLs before and after HAART initiation, AIDS-defining illnesses after HAART initiation, date of death, and cause of death. Deaths were ascertained by searching the Social Security Registry, the National Death Index, administrative databases, and records at the military institutions or through reports of family or friends.
Age at Seroconversion
The date of seroconversion was defined as the midpoint between the last documented negative HIV test and the first documented positive HIV test. Age at seroconversion was calculated for each subject as the estimated seroconversion date minus their birth date. For descriptive analyses, we compared subjects who were 18-29, 30-39, or 40 years and older (due to a low number of subjects older than 50 years in this cohort) at the time of seroconversion. All baseline comparisons across age groups were performed using the Kruskal-Wallis test for continuous measurements and Fisher exact test for categorical variables.
For this study, HAART refers to a regimen of:
1. Two or more nucleoside reverse transcriptase inhibitors in combination with at least one protease inhibitor (PI) or one nonnucleoside reverse transcriptase inhibitor (NNRTI) or
2. One reverse transcriptase inhibitor with at least one PI and one NNRTI or
3. An abacavir- or tenofovir-containing regimen of 3 or more reverse transcriptase inhibitors in the absence of both PIs and NNRTIs.
Antiretroviral (ART) use refers to therapy with 1 or 2 drugs not meeting the definition of HAART. For purposes of this study, we used an intent-to-treat approach and thus ignored subsequent changes to regimens, treatment interruptions, and treatment terminations. As the TACC NHS is an observational study, therapy was initiated by individual providers based on current treatment guidelines and patient preferences.
Baseline VL was defined as the most recent measurement within 6 months before initiating HAART. Because VL assays with different lower limits of detection were used in the 1990s, an undetectable VL was defined as <500 copies/mL. Viral suppression refers to having an undetectable VL. Kaplan-Meier survival methods were used to estimate the time from HAART initiation to viral suppression and time from viral suppression to the first of 2 consecutive detectable VLs. The log-rank test was used to compare the survival curves across age groups for each outcome. Subjects were censored at the time of their last recorded VL. Cox proportional hazards models were used to examine the effect of age at seroconversion (adjusted for sex, race, rank, baseline VL, baseline CD4 cell count, ART usage before HAART, diagnosis of AIDS before HAART, year of HAART initiation, and PI or NNRTI usage in the first HAART regimen) on each virologic outcome. The R programming language Design package20 was used to test the linearity of the effect of age at seroconversion with respect to each outcome by fitting age using a restricted cubic spline function with 5 knots. If the effect of age on the outcome was found to be significantly nonlinear, then age was fit using the restricted cubic spline function in the final model; otherwise age was modeled as a continuous variable.
CD4 Cell Response
Baseline CD4 cell count was defined as the most recent measurement within 6 months before initiating HAART. CD4 cell response was summarized by the CD4 cell count every 6 months, beginning at the time of HAART initiation and ending at their last recorded CD4 cell count, or 84 months of follow-up. In the post-HAART period for each subject, if multiple CD4 counts within a given 6-month time frame were available, then the value closest to (on or before) the visit date was used. The slope of CD4 cell count from HAART initiation to month 6 was calculated as (CD4 at month 6 − CD4 at baseline)/6 and compared across age groups using analysis of variance. The difference between age groups in the slope of CD4 during the follow-up period of months 6-84 was compared using the PROC MIXED procedure in SAS (SAS Institute Inc, Cary, NC).
The proportion of subjects who died or developed AIDS after HAART initiation was compared across age groups using Fisher exact test. Kaplan-Meier survival methods were used to estimate the time from HAART initiation to death or to the diagnosis of AIDS for each seroconversion age group, and the log-rank test was used to compare the survival curves across age groups. Subjects were censored at the time of their last recorded NHS research visit.
There were 2395 subjects in the TACC NHS database with documented negative followed by documented positive HIV tests. Of these, 1599 were excluded from this study because they had estimated seroconversion dates before January 1, 1996. Of the remaining 796, there were 233 subjects who did not have a documented date of HAART initiation, leaving 563 subjects available for analysis. Of the 563 subjects, 296 seroconverted between the ages of 18 and 29 years, 233 between 30 and 39 years, 37 between 40 and 49 years, and 7 aged 50 years and older.
The baseline characteristics of the 563 subjects are shown in Table 1. The mean age of the 563 subjects was 29.6 years (range 18-58). The majority of individuals were male, 46% were African American, 41% were white, and 12% were of other ethnicities. Race was significantly different between the age groups (P = 0.0002), with a larger percentage of African Americans seroconverting between the ages of 18 and 29 years compared with whites or other ethnicities. Ninety percent of patients had seroconversion windows (the time between their last negative and first positive HIV tests) less than 3.3 years; however, there was a difference between age groups with the older groups having longer windows. Baseline VL, baseline CD4 count, number of clinic visits per year, year of seroconversion, year of HAART initiation, time from seroconversion to initiation of HAART, type of HAART regimen, prior AIDS diagnoses, and ART use before HAART did not significantly differ by age group. There was a significant difference in hepatitis C infection between the age groups; however, only 13 of 563 subjects were infected.
Of the 563 subjects analyzed, 541 had at least one VL post-HAART initiation. The median time from HAART initiation to first follow-up VL was 1.4 months and was not significantly different between the age groups (P = 0.55, data not shown). The number of VLs measured per year of follow-up was also not significantly different between the age groups (Table 1, P = 0.59). Increasing age at seroconversion was significantly associated with shorter time to viral suppression (Fig. 1, P = 0.002, log-rank test). The Kaplan-Meier estimate for attaining an undetectable VL by 6-month post-HAART initiation was 68.8% [95% confidence interval (CI) 63.3 to 74.2] for those who seroconverted between the ages of 18 and 29 years compared with 79.8% (95% CI 74.3 to 84.9) for those aged 30-39 years and 74.4% (95% CI 60.9 to 86.2) for those 40 years and younger at seroconversion. The final adjusted Cox model included 460 subjects with values for all covariates: age at seroconversion, sex, rank, race, baseline CD4 count, baseline log10 VL, year of HAART initiation, previous AIDS-defining illness, prior ART usage, and first HAART regimen (PI vs NNRTI). There was a significant interaction between age and race (P = 0.0005); therefore, interaction terms were also included in the Cox model. The effect of age was found to be nonlinear (test for linearity P = 0.0045), thus the main effect of age and the age effect in the interaction were modeled using a restricted cubic spline function. The log hazard was plotted, overall and by race, and illustrated how the relative probability of suppression changed across all seroconversion ages (plot not shown). For illustrative purposes, Table 2 displays the hazard ratio (HR) of viral suppression, by race and overall, for selected age levels relative to age at seroconversion of 20 years. The plots and HRs in Table 2 demonstrated that for African Americans there was no significant difference in time to viral suppression between 20 and 30 year olds, but the probability of suppression increased after age 30, whereas for whites the probability increased from age 20 to 30 and then decreased. For those of other ethnicities, there was a relatively linear pattern of increasing probability of viral suppression with increasing seroconversion age.
Increasing age at seroconversion was also significantly associated with longer time to loss of virologic response (Fig. 2, P < 0.0001, log-rank test). In the Cox regression model, the effect of age with respect to this outcome was found to be linear (test for linearity P = 0.4); therefore, age was modeled as a continuous variable. The univariate analysis showed that the relative risk of the loss of virologic response was reduced by 32% for every 5-year increase in age above 18 years at seroconversion (HR = 0.68, 95% CI 0.59 to 0.77). The adjusted model produced similar results (HR = 0.65, 95% CI 0.55 to 0.75). There was no significant interaction between age and ethnicity with respect to this outcome.
CD4 Cell Response
There was no significant difference in the slope of the CD4 decline from seroconversion to HAART initiation between the age groups (data not shown). Five hundred nine of the 563 subjects who started HAART had a baseline CD4 measurement available (defined as the most recent CD4 count within 6 months before starting HAART). Follow-up visits from baseline to 84 months were examined, and the mean CD4 count over time for each age group is shown in Figure 3. Across the age groups, there were 2 phases of CD4 count increases after HAART initiation; a steeper increase from baseline to 6-month post-HAART and a more modest increase from 6 to 84 months post-HAART initiation. Because of these different phases, we examined the slopes of the CD4 count increases between age groups from baseline to 6-month post-HAART and the slopes from 6 to 84 months post-HAART. There were no significant differences in the mean slopes (cells/mm3/mo) from baseline to 6-month post-HAART between the age groups (18-29 years = 21.3, 30-39 years = 23.2, and 40+ years = 19.7; P = 0.72). The slopes of CD4 count increases (cells/mm3/mo) over the follow-up period (months 6-84), however, were significantly different between age groups (18-29 years = 0.70, 30-39 years = 1.51, and 40+ years = 2.62; P = 0.0002) with those seroconverting in the older age groups having a steeper CD4 count increase over time. These significant differences remained when the CD4 count analysis was limited only to those who achieved and maintained viral suppression for the duration of their follow-up (data not shown).
Of the 563 subjects who initiated HAART, 40 (7.1%) developed AIDS or died after HAART initiation. The proportion of subjects with at least one AIDS-defining event or death was not significantly different across age groups (P = 0.47). Age at seroconversion was also not significantly associated with time from HAART initiation to AIDS/death (P = 0.52, log-rank test, data not shown).
This study of a prospectively followed cohort of more than 550 adults demonstrates that increasing age at seroconversion from 18 to older than 40 years is associated with better virologic and CD4 cell responses to HAART. This study is unique in that it examines age at seroconversion rather than age at diagnosis or age at treatment initiation, thereby eliminating any effects due to delayed diagnosis of HIV infection. In addition, as subjects in the TACC NHS have equal access to free health care, associations between age and ability to obtain care or medications are greatly reduced as factors in this analysis.
The positive association of increasing age on virologic suppression has been seen in other studies as well.5,14,16,21 A large study of over 3000 subjects from the French Hospital Database on HIV demonstrated that individuals older than 50 years at the time of HAART initiation achieved virologic suppression faster than those younger than 50 years at HAART initiation.14 The study did not evaluate whether there were differential effects of age on time to virologic suppression among individuals younger than 50 years. Another study of 49,921 HIV-infected individuals from 33 European cohorts that examined response to HAART in persons ranging in age from 1 day to 87 years also found that increasing age correlated with an increasing chance of virologic response to ART therapy.21 Again, the duration of HIV infection was unknown in this study.
Besides achieving viral suppression more quickly, individuals who seroconverted at older ages also maintained viral suppression significantly longer before viral rebound than those who seroconverted at younger ages. Studies done to date have not explained why older subjects have better virologic responses to HAART. Possibilities include better adherence to the medications, more favorable pharmacokinetics, or better access to care. The latter possibility should be less likely in the TACC NHS, where all subjects have equal access to free medical care. The TACC NHS did not collect information on adherence, and studies that have evaluated the association between age and adherence have had mixed results.16,22,23 For example, one study that looked at the association of sociodemographic factors and ART adherence found that young persons (aged 20 to younger than 32 years) have worse adherence than those aged 32 to younger than 35 years; however, there was no difference in adherence between those in the young group and those older than 35 years.24 Another study that looked at adherence to PI-based regimens found that older age was associated with >95% adherence (odds ratio 1.1, 95% CI 1.0 to 1.2); however, in this study, age was not associated with virologic failure in the univariate analysis and therefore was not included in the multivariate model.25 A more recent study evaluating the effect of differential adherence on virologic outcome found that age remained a significant predictor of time to virologic failure even when the model was adjusted for adherence.26 Therefore, adherence differences between age groups may not fully explain the differences seen in virologic outcomes.
The current study found that the effect of age on time to virologic suppression differed by ethnicity. These findings suggest that the factors relating age to viral suppression post-HAART initiation, whether they are related to adherence, pharmacokinetics, or other factors, may differ by ethnicity. There have been several studies comparing ART adherence rates between whites and African Americans with mixed findings27-30; however, none of these studies evaluated whether the association between ethnicity and adherence differed depending on age. Likewise, studies evaluating the association between ART drug levels or genetic polymorphisms associated with increased or decreased drug metabolism and virologic outcome have not examined the impact of age on these associations.31-33 Other studies that have evaluated the effect of age on response to HAART did not comment on any interaction between age and ethnicity, and this finding needs to be confirmed by other studies.
Along with improved virologic response to HAART, increasing age at seroconversion also correlated with a faster rate of CD4 cell increase despite similar use of PI- vs NNRTI-based regimens among the different age groups. This finding is in contrast to many other published studies, which have demonstrated that older age correlates with smaller increases in CD4 cells over time.5,13,14,34 In those studies, however, the older age group was comprised of subjects older than 45 or 50 years, and it has been theorized that the smaller increases seen in individuals older than 45 years are due to decreased thymopoiesis. Our cohort had only a small number of subjects older than 45 years, which may explain our discordant findings.
A previous study of the TACC cohort in the pre-HAART era found that older age was associated with faster HIV disease progression.35 In contrast, the current study of the TACC cohort in the HAART era did not reveal any differences in disease progression between the age groups. In fact, the vast majority of subjects did well, with only 7.1% experiencing an AIDS event or death over a median follow-up of 3.4 years. This finding is consistent with that of a large European Collaboration that demonstrated that age at seroconversion had a much smaller effect on progression to AIDS in 1999-2001 compared with the pre-HAART era.6 That study did find that age at seroconversion continued to have an effect on the risk of all causes of death even in the post-HAART era. The authors postulated that if the effect of HAART on mortality is mediated through reducing HIV disease progression, then there may be a lag before the effect on mortality was seen. The current study of the TACC cohort collected data through 2006, 5 years after the CASCADE study and as they predicted, age at seroconversion had little prognostic importance for mortality. Given that individuals in the TACC cohort who seroconverted at younger ages did worse in terms of maintenance of viral suppression and immune reconstitution, it is possible that longer follow-up will reveal that in the HAART era, younger age at seroconversion is associated with faster HIV disease progression.
Several potential limitations of this study should be noted. The small number of subjects older than 50 years at seroconversion limits our ability to make conclusions about this age group. Date of seroconversion was calculated as the midpoint between the last negative HIV test and the first positive HIV test. The window of time between these 2 tests was variable between subjects; however, 90% of subjects have windows less than 3.3 years. We repeated the analyses including only subjects with seroconversion windows less than 3.5 years, and the results remained the same. Lastly, the TACC NHS does not collect information on risk factors for acquiring HIV, which precludes the ability to adjust results for these factors. Given that random drug tests are administered to those on active duty, however, there is little use of illicit drugs in this cohort.
In conclusion, this study demonstrates that increasing age at seroconversion from 18 to 40 years is associated with a favorable rate and duration of response to HAART. Possible reasons why younger individuals did not respond and older ones to HAART include adherence differences or differences in drug absorption, distribution, or metabolism. Further investigation into these factors is imperative to optimizing therapy.
We thank all the military personnel and beneficiaries who have graciously volunteered their time to participate in Infectious Disease Clinical Research Program studies, without whom this work would not have been possible. We also thank John Fieberg, PhD, Lynn Eberly, PhD, and Charles Hicks, MD, for their invaluable reviews of this study.
1. Corey DM, Kim HW, Salazar R, et al. Brief report: effectiveness of combination antiretroviral therapy on survival and opportunistic infections in a developing world setting: an observational cohort study. J Acquir Immune Defic Syndr
2. Crum NF, Riffenburgh RH, Wegner S, et al. Comparisons of causes of death and mortality rates among HIV-infected persons: analysis of the pre-, early, and late HAART eras. J Acquir Immune Defic Syndr
3. Valdez H, Chowdhry TK, Asaad R, et al. Changing spectrum of mortality due to human immunodeficiency virus: analysis of 260 deaths during 1995-1999. Clin Infect Dis
4. 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
5. Yamashita TE, Phair JP, Munoz A, et al. Immunologic and virologic response to highly active antiretroviral therapy in the Multicenter AIDS Cohort Study. AIDS
6. CASCADE Collaboration. Determinants of survival following HIV-1 serconversion after the introduction of HAART. Lancet
7. Lundgren JD, Mocroft A, Gatell JM, et al. A clinically prognostic scoring system for patients receiving highly active antiretroviral therapy: results from the EuroSIDA study. J Infect Dis
8. Valdez H, Purvis SF, Lederman MM, et al. Association of the CCR5delta32 mutation with improved response to antiretroviral therapy. JAMA
9. Ioannidis JPA, Rosenberg RS, Goedert JJ, et al. Effects of CCR5delta32, CCR2-64I, and SDF-1 3A alleles on HIV-1 disease progression: an international meta-analysis of individual-patient data. Ann Intern Med
10. Grimes RM, Otiniano ME, Rodriguez-Barradas MC, et al. Clinical experience with human immunodeficiency virus-infected older patients in the era of effective antiretroviral therapy. Clin Infect Dis
11. Perez JL, Moore RD. Greater effect of highly active antiretroviral therapy on survival in people aged >50 years compared with younger people in an urban observational cohort. Clin Infect Dis
12. Manfredi R, Chiodo F. A case-control study of virological and immunological effects of highly-active antiretroviral therapy in HIV-infected patients with advanced age. AIDS
13. Goetz MB, Boscardin WJ, Wiley D, et al. Decreased recovery of CD4 lymphocytes in older HIV-infected patients beginning highly active antiretroviral therapy. AIDS
14. 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
15. Tumbarello M, Rabagliati R, de Gaetano Donati K, et al. Older HIV-positive patients in the era of highly active antiretroviral therapy; changing of a scenario. AIDS
16. Knobel H, Guelar A, Validecillo G, et al. Response to highly active antiretroviral therapy in HIV-infected patients aged 60 year or older after 24 months of follow up. AIDS
17. Darby SC, Ewart DW, Giangrande PF, et al. Importance of age at infection with HIV-1 for survival and development of AIDS in UK haemophilia population. Lancet. 1996;347:1573-1579.
18. Veugelers PJ, Page KA, Tindall B, et al. Determinants of HIV disease progression among homosexual men registered in the tricontinental seroconverter study. Am J Epidemiol
19. Centers for Disease Control and Prevention. Revised classification system for HIV infection and expanded surveillance case definitions for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep
20. Harrell F Jr. Regression modeling strategies with applications to linear models, logistic regression, and survival analysis. General aspects of fitting regression models. New York, NY: Springer-Verlag; 2001.
21. Sabin C. The COHERE study. Response to combination ART: variation by age . Paper presented at: 14th Conference on Retroviruses and Opportunistic Infections; 2007; Los Angeles.
22. Nilsson SL, Williams ML, Ross MW, et al. Factors associated with suboptimal antiretroviral therapy adherence to dose, schedule, and dietary instructions. AIDS Behav
23. Le Moing V, Chene G, Carrieri MP, et al. Predictors of virological rebound in HIV-1-infected patients initiating a protease inhibitor-containing regimen. AIDS
24. Gordillo V, del Amo J, Soriano V, et al. Sociodemographic and psychological variables influencing adherence to antiretroviral therapy. AIDS
25. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med
26. Gardner LI, Brundage JF, McNeil JG, et al. Predictors of HIV-1 disease progression in early- and late- stage patients: the US Army Natural History Cohort. J Acquir Immune Defic Syndr
27. 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
28. Pence BW, Ostermann J, Kumar V, et al. The influence of psychosocial characteristics and race/ethnicity on the use, duration, and success of antiretroviral therapy. J Acquir Immune Defic Syndr
29. McGinnis KA, Fine MJ, Sharma RK, et al. Understanding racial disparities in HIV using data from the veterans aging cohort 3-site study and VA administrative data. Am J Public Health
30. Schackman BR, Ribaudo HJ, Krambrink A, et al. Racial differences in virologic failure associated with adherence and quality of life on efavirenz-containing regimens for initial HIV therapy. J Acquir Immune Defic Syndr
31. Haas DW, Smeaton LM, Shafer RW, et al. Pharmacogenetics of long-term responses to antiretroviral regimens containing efavirenz and/or nelfinavir: an Adult AIDS Clinical Trials Group study. J Infect Dis
32. Best BM, Goicoechea M, Witt MD, et al. A randomized controlled trial of therapeutic drug monitoring in treatment-naïve and -experienced HIV-1-infected patients. J Acquir Immune Defic Syndr
33. Fellay J, Marzolini C, Meaden ER, et al. Response to antiretroviral treatment in HIV-1 infected individuals with allelic variants of the multidrug resistance transporter 1:a pharmacogenetics study. Lancet
34. Viard JP, Mocroft A, Chiesi A, et al. Influence of age on CD4 cell recovery in human immunodeficiency virus-infected patients receiving highly active antiretroviral therapy: evidence from the EuroSIDA study. J Infect Dis
35. Gardner EM, Sharma S, Peng G, et al. Differential adherence to combination antiretroviral therapy is associated with virological failure with resistance. AIDS
© 2008 Lippincott Williams & Wilkins, Inc.