Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Cumulative Viral Load and Virologic Decay Patterns after Antiretroviral Therapy in HIV-Infected Subjects Influence CD4 Recovery and AIDS

  • Vincent C. Marconi ,

    vcmarco@emory.edu (VCM); kulkarnih@uthscsa.edu (HK)

    Affiliations Division of Infectious Diseases, Emory University School of Medicine, Atlanta, Georgia, United States of America, Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America

  • Greg Grandits,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, Division of Biostatistics, University of Minnesota, Minneapolis, Minnesota, United States of America

  • Jason F. Okulicz,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, Infectious Disease Service, San Antonio Military Medical Center, Brooke Army Medical Center, Fort Sam Houston, Texas, United States of America

  • Glenn Wortmann,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, Infectious Disease Service, Walter Reed Army Medical Center, Washington, D.C., United States of America

  • Anuradha Ganesan,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, Infectious Disease Clinic, National Naval Medical Center, Bethesda, Maryland, United States of America

  • Nancy Crum-Cianflone,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, Infectious Disease Clinic, Naval Medical Center San Diego, San Diego, California, United States of America

  • Michael Polis,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, National Institute for Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland, United States of America

  • Michael Landrum,

    Affiliations Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America, Infectious Disease Service, San Antonio Military Medical Center, Brooke Army Medical Center, Fort Sam Houston, Texas, United States of America

  • Matthew J. Dolan,

    Affiliation Henry M. Jackson Foundation, Wilford Hall United States Air Force Medical Center, Lackland Air Force Base, Texas, United States of America

  • Sunil K. Ahuja,

    Affiliations Veterans Administration Research Center for AIDS and HIV-1 Infection, South Texas Veterans Health Care System, San Antonio, Texas, United States of America, Department of Medicine, University of Texas Health Science Center, San Antonio, Texas, United States of America, Department of Microbiology and Immunology, and Biochemistry, University of Texas Health Science Center, San Antonio, Texas, United States of America

  • Brian Agan,

    Affiliation Infectious Disease Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, United States of America

  • Hemant Kulkarni ,

    vcmarco@emory.edu (VCM); kulkarnih@uthscsa.edu (HK)

    Affiliations Veterans Administration Research Center for AIDS and HIV-1 Infection, South Texas Veterans Health Care System, San Antonio, Texas, United States of America, Department of Medicine, University of Texas Health Science Center, San Antonio, Texas, United States of America

  • the Infectious Disease Clinical Research Program (IDCRP) HIV Working Group

    Membership of the IDCRP HIV Working Group is provided in the Acknowledgments.

Abstract

Background

The impact of viral load (VL) decay and cumulative VL on CD4 recovery and AIDS after highly-active antiretroviral therapy (HAART) is unknown.

Methods and Findings

Three virologic kinetic parameters (first year and overall exponential VL decay constants, and first year VL slope) and cumulative VL during HAART were estimated for 2,278 patients who initiated HAART in the U.S. Military HIV Natural History Study. CD4 and VL trajectories were computed using linear and nonlinear Generalized Estimating Equations models. Multivariate Poisson and linear regression models were used to determine associations of VL parameters with CD4 recovery, adjusted for factors known to correlate with immune recovery. Cumulative VL higher than the sample median was independently associated with an increased risk of AIDS (relative risk 2.38, 95% confidence interval 1.56–3.62, p<0.001). Among patients with VL suppression, first year VL decay and slope were independent predictors of early CD4 recovery (p = 0.001) and overall gain (p<0.05). Despite VL suppression, those with slow decay during the first year of HAART as well as during the entire therapy period (overall), in general, gained less CD4 cells compared to the other subjects (133 vs. 195.4 cells/µL; p = 0.001) even after adjusting for potential confounders.

Conclusions

In a cohort with free access to healthcare, independent of established predictors of AIDS and CD4 recovery during HAART, cumulative VL and virologic decay patterns were associated with AIDS and distinct aspects of CD4 reconstitution.

Introduction

The initial goal of highly-active antiretroviral therapy (HAART) was to improve AIDS-free survival and attempt to mitigate the harmful effects of treatment. Immune reconstitution via CD4 recovery served as an intermediate marker for response to HAART because of its predictive capacity for AIDS events and death.[1], [2], [3] Thereafter, virologic suppression became the primary target for therapy because it was shown to be an appropriate, early predictor of immunologic response and clinical outcomes.[4], [5], [6], [7] Furthermore, it was demonstrated that incomplete suppression of viral replication allowed for the emergence of drug resistance and ultimately virologic failure.[8], [9] These findings led to recommendations in the U.S. Department of Health and Human Services guidelines that patients should achieve complete virologic suppression (viral load [VL] <400 copies/mL by 24 weeks or <50 copies/mL by 48 weeks) and maintain suppression thereafter.[10]

Even among patients reaching these virologic targets, there are significant inter-individual differences in the recovery of CD4+ T cells and risk of clinical events, suggesting that other factors may relate to these outcomes.[11], [12], [13], [14], [15], [16] Age at HAART initiation, pre-HAART VL and CD4 cell count, magnitude of and time to VL suppression all have been shown to influence CD4 recovery and clinical outcomes. [4], [13], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26] Although the relationship of virologic decay patterns with VL changes during HAART has been described,[23], [27], [28], [29] the impact of these decay patterns on CD4 reconstitution and risk of subsequent clinical AIDS events has not been fully elucidated. Furthermore, it is also conceivable that the overall VL burden, represented as the cumulative VL during HAART, may also influence CD4 recovery and risk of AIDS events. Hence, we determined whether the patterns of virologic decay and the cumulative VL during HAART were associated with AIDS and CD4 recovery after HAART initiation independent of the currently recommended dichotomous measures of VL suppression[10] within a large, observational cohort with free access to medications and care, high rates of adherence, and low rates of injection drug use.[26], [30] If virologic decay measures are independently associated with outcomes, this could provide some explanation as to why some individuals experience inadequate treatment response despite achieving virologic suppression. Additionally, cumulative viral load could serve as a sensitive marker for risk of AIDS after HAART beyond traditional measures.

Materials and Methods

Study Participants

The U.S. Military HIV Natural History Study (NHS) is a prospective multicenter observational study of HIV-infected active duty military personnel and other beneficiaries (spouses, dependents, and retired military personnel) from the Army, Navy/Marines and Air Force. Seroconverters (SC) were defined as patients having a documented HIV seronegative date prior to the first positive HIV date (see Table S1). The estimated date of seroconversion for SC was defined as the midpoint between the two dates. All CD4 count, VL, and other measurements were done as part of routine clinical care.[31] The clinically-approved methodology for this testing varied by site and over time. Prior ARV use referred to any antiretroviral therapy not meeting the NHS definition of HAART.[26] HAART initiation was the date when HAART was first prescribed.

Ethics Statement

Participants who provided written informed consent and initiated HAART through July 1, 2008 regardless of regimen continuation were included in the present study. The NHS and this substudy have been approved by each center's Institutional Review Board and the Uniformed Services University of the Health Sciences Institutional Review Board.

Statistical Analysis

VL Parameters.

A primary aim of this study was to capture and summarize the overall and early VL dynamics in such a manner as to permit their eventual use in clinical practice. In that regard, we made the following assumptions: i) by the time HAART is typically initiated for an individual in the NHS a natural steady state VL exists; ii) once potent HAART is initiated there is a rapid decline in the VL followed by a slower decline; and iii) such a typical pattern of VL can be explained on the basis of an exponential decay in the circulating VL. The definitions of the parameters used in this study are shown in Table S1, and the theoretical bases for the estimation of these parameters are further described in Note S1. The composite “virologic decay” refers to the application of an exponential decay equation which has been fitted to all viral loads available for an individual after the initiation of HAART. For a majority of participants in this cohort who have a high level of adherence, the virologic decay pattern corresponds to the concatenation of each "classical" (first, second, etc.) phase of decay for that individual. For some participants, their virologic decay does not follow these patterns due to suboptimal adherence, inadequate drug levels, drug resistance, and treatment interruption.

We computed four VL parameters at the level of each individual: (i–ii) exponential decay constant of VL change during entire duration of HAART (overall) and during the first year of HAART, respectively; (iii) VL slope during the first year of HAART obtained using linear Generalized Estimating Equations (GEE) models; and (iv) cumulative VL (Table S1). The VL parameters described above in i, ii, and iii are designated as VL kinetic parameters. Similarly, we computed the following four CD4 count parameters at the level of the individual: (i–ii) slope of the CD4 count change during and after the first two years of HAART; (iii) mean CD4 count after the first two years of HAART; and (iv) overall gain in CD4 counts (Table S1).

Cohort level analyses.

The cohort-level analyses made use of all available CD4+ T cell count and VL data for all subjects to generate time-trend lines or curves using linear and non-linear GEE models, assuming an equal correlation structure. The time-trend curves derived by non-linear GEE modeling were refined further using spline smoothed curves with knots at the end of each year since HAART initiation. The resulting curves describe an overall or composite VL pattern for the cohort.

Association analyses.

We estimated the parameters detailed in Table S1 for each individual. The association of these individual level parameters with the risk of AIDS (defined using 1993 clinical criteria[32] but did not include a CD4 count <200 cells/µL as an endpoint) was assessed by Poisson regression models, and with recovery of CD4 counts by linear regression models. In these models we accounted for the potential confounding due to VL suppression by HAART by including two covariates - achievement of VL suppression (as defined in Table S1) and the time taken to achieve VL suppression from the start of HAART. As described in the results, we ran these multivariate analyses for the VL parameters that were estimated (i) by including all VL measurements after HAART and separately (ii) by restricting to only those measurements after HAART but prior to the occurrence of the first AIDS event. Statistical significance was evaluated at a type I error rate of 0.05. All statistical analyses were conducted using Stata 7.0 (Stata Corp., College Station, TX).

Results

Cohort-level VL and CD4 changes after HAART initiation

Characteristics of the 2278 participants who initiated HAART are in Table 1. The average follow up time after HAART for participants was 5.63 years (SD 3.98). Cohort-level non-linear GEE modeling of VL from time-of-HAART initiation in all subjects revealed the following pattern: a precipitous decline in VL during the first year, a temporary rebound at ∼1.6 years post-HAART, followed by a relatively steady-state VL thereafter (Fig. 1A). The VL trajectory of subjects who developed AIDS during HAART versus those who remained AIDS-free differed significantly as a decline in VL after HAART initiation was not observed in patients who developed AIDS (Fig. 1B). In all subjects (Fig. 1C) and in those who attained VL suppression (Fig. 1D), VL trajectories differed according to the tertiles of the pre-HAART VL such that those who started with higher VLs (upper and middle tertiles of pre-HAART VL) displayed a sharper decline in VL than those subjects categorized to the lower pre-HAART VL tertile (Fig. 1C-D, Table S2).

thumbnail
Figure 1. CD4+ T cell count and VL trajectories during HAART.

(A) Overall (population level) VL and CD4 trajectories after HAART initiation. The curves for CD4 counts (blue and corresponding to the right Y axis) and VL (red and corresponding to the left Y-axis) are superimposed to provide a common temporal view of the trajectories from time of HAART initiation (x-axis). (B) VL trajectories in subjects who developed or did not develop AIDS during HAART. (C–D) VL trajectories after HAART initiation based on the tertiles of pre-HAART VL in (C) all subjects and (D) those who achieved VL suppression. (E–F) CD4+ T cell count trajectories after initiation of HAART according to (E) attainment of VL suppression and (F) VL kinetics among VL suppressers. In panel F, slow/slow indicates subjects who had slow (less than median) rate of VL decay estimated either using all the VL measurements or using those during the first year after HAART initiation. The remaining three groups (rapid/slow, slow/rapid and rapid/rapid) showed similar trajectories and were, therefore, grouped into a single category. All trajectories shown were modeled using non-linear GEE and spline smoothing assuming equal-correlation structure. In panels, A, B, E and F the central thick line represents the mean and the two straddling thin lines represent the edge of the 95% confidence interval band. N, number of subjects; M, number of CD4 or VL measurements.

https://doi.org/10.1371/journal.pone.0017956.g001

The cohort-level trajectories in CD4 counts during HAART revealed two phases of CD4 count changes In phase I, for all subjects initiating HAART there was a rapid increase in CD4 counts during the first two years, followed in phase II by a slower, sustained gain in CD4 cells (Fig. 1A). We stratified the cohort-level changes in CD4 count gains according to whether subjects attained VL suppression (Fig. 1E). This analysis revealed that during the first year of HAART, rapid and similar gains in CD4 counts (∼200 cells on average) were observed in those who did (brown curve) or did not (black curve) attain VL suppression (Fig. 1E). However, in contrast to those who attained VL suppression, the initial gains in CD4 counts were not durable among those who did not achieve VL suppression (Fig. 1E).

VL kinetic parameters and AIDS risk after HAART

The association of the three VL kinetic parameters and cumulative VL with risk of developing AIDS during HAART was evaluated in separate multivariate models adjusted for length of follow up. For these and the other analyses described later, we dichotomized subjects based on the VL parameters using the median value of the parameter as the cut-off. We included into each multivariate model additional covariates that have been shown to be predictive of immunologic recovery during HAART, including time to VL suppression (Table 2).[26], [33], [34], [35], Nadir CD4 was used as a surrogate for pre-HAART CD4 counts because the median time from the nadir CD4 to HAART initiation was short (Table 1).

thumbnail
Table 2. Association of VL parameters with risk of AIDS development after initiation of HAARTa.

https://doi.org/10.1371/journal.pone.0017956.t002

The slope and exponential decay constant for VL during the first year of HAART were not predictive of AIDS, whereas a slower overall VL decay showed a statistical trend towards predicting AIDS during HAART, independent of the other covariates (RR 1.38, p = 0.058). A higher than average cumulative VL during HAART (RR = 2.38, 95% CI = 1.56–3.62) was associated with the greatest risk of developing AIDS. These results were similar when the VL parameters were estimated excluding the VL measurements recorded after the AIDS event occurred (Table 2) or when the analyses were restricted to seroconverters only (Table S3).

In separate analyses, where attainment of VL suppression at any point during HAART was replaced with VL suppression by 6 or 12 months, overall VL decay constant was a significant independent predictor of AIDS in patients who attained VL suppression at 6 (RR = 1.65, p = 0.007, 95% CI = 1.14–2.38) and 12 (RR = 1.43, p = 0.055, 95% CI = 0.99–2.05) months, whereas in these models, the VL slope or exponential VL decay during the first year were not predictive of AIDS. The cumulative VL remained highly predictive of AIDS risk in those who attained VL suppression during 6 (RR = 1.96, p = 0.004, 95% CI = 1.24–3.13) and 12 (RR = 2.33, p = 0.001, 95% CI = 1.44–3.80) months of HAART. Collectively, these data indicated that a slow overall VL exponential decay and a high cumulative VL during HAART increased AIDS risk after initiation of HAART.

VL parameters and CD4 Recovery

We next determined whether the VL kinetic and other parameters that were included in the models to assess AIDS risk during HAART also associated with the rate of CD4 gain (Table 3). We found that the VL parameters predicted different aspects of CD4 count recovery even after accounting for factors that we found to be highly predictive of AIDS risk, including prior history of AIDS, nadir CD4, age at HAART initiation and time to VL suppression. The overall decay rate constant was not predictive of rate of CD4 gain in the first 2 years, but was significantly associated with the rate of CD4 gain after two years of HAART, the mean CD4 count two years after HAART, and the overall gain of CD4 cells (Table 3). The decay constant and VL slope in the first year of HAART were mostly predictive of the rate of CD4 cell gain during the first two years and the overall gain in CD4 cells (Table 3). By contrast, the cumulative VL was only predictive of rate of CD4 gains after 2 years and not the overall gain in the CD4 count (Table 3).

thumbnail
Table 3. Association of VL parameters with CD4 recovery after HAART initiation in subjects who did not develop AIDS.

https://doi.org/10.1371/journal.pone.0017956.t003

The aforementioned data suggested that a slower VL decay during the first year of HAART is associated with both a reduced rate of CD4 gain in the first two years of HAART and overall gain in CD4 cells (Table 3). By contrast, a slower overall VL decay is more predictive of a reduced rate of CD4 gain after 2 years of HAART, lower mean gains in CD4 counts after 2 years of HAART as well as a reduced overall gain in CD4 cells (Table 3). On the basis of these results, we posited that VL suppressers who had a slow VL decay in the first year of HAART and the entire therapy course (overall) would fare the worst with respect to CD4 recovery. To test this, we categorized VL suppressors into two groups: those with a slow decay in the first year and slow overall decay were categorized into one group, whereas the remainder (rapid/rapid, rapid/slow, slow/rapid decay in the first year and overall decay, respectively) were grouped together because they had very similar CD4 count trajectories (data not shown). Subjects categorized to the slow early/slow overall decay group were similar to other subjects with respect to age at HAART initiation, ethnicity and nadir CD4 (all p values >0.2).

Notably, VL suppressers categorized to the slow/slow decay group had a significantly muted CD4 recovery during HAART compared with all other subjects (Fig. 1F). Concordantly, VL suppressors categorized to the slow/slow decay category had a slower rate of CD4 recovery in the first two years (54.6 vs. 80.2 cells/µL/yr, p = 0.02) and after two years (−14.8 vs. 16.5 cells/µL/yr, p = 0.002) of HAART, a lower mean CD4 count after two years (564.6 vs. 614.8 cells/µL, p = 0.005) and a lower absolute CD4 gain (133 vs. 195.4 cells/µL p = 0.001). We also conducted these analyses for subjects who achieved VL suppression within 6 and 12 months and found highly concordant results (data not shown).

Discussion

Not all patients on HAART display robust CD4 cell gains, despite VL suppression.[11], [12], [13], [14], [15], [16] This has been attributed previously to factors such as pre-HAART VL and nadir CD4, age at HAART initiation, and depth of and time to VL suppression. In this study, we modeled the VL decay and cumulative VL and applied these relatively unique parameters to a large well-characterized cohort in order to determine whether these factors were associated with AIDS risk and CD4 recovery during HAART independent of currently recommended benchmarks of VL suppression at 6 and 12 months.[10] In the participants that we evaluated, the initiation of HAART was associated with a predictable decline in VL that was concomitantly associated with an increase in CD4 counts. Subjects who did or did not achieve VL suppression both experienced, on average, a gain of 200 CD4 cells/µL during the first year of HAART. However, in contrast to those who attained VL suppression, these gains were not sustainable among non-VL suppressers. Notwithstanding the importance of attaining VL suppression or minimizing the time to VL suppression, our data show that in addition to these endpoints, both a slow early (first year of HAART) and slow overall (during entire treatment period) VL decay were independently associated with both a slower rate of and lower absolute CD4 cell gain during HAART. Furthermore, a slower overall VL decay in those who attained VL suppression within 6 and 12 months of HAART initiation, and a higher cumulative VL during HAART were each independent predictors of increased AIDS risk during HAART. These findings suggest that the patterns of VL decay are important factors in addition to VL suppression for CD4 reconstitution and risk of AIDS during HAART.

The pre-HAART VL predicted the subsequent rate of decay during the first year of HAART. Since most patients were able to achieve suppression by 6–12 months, it is not surprising that the rate of decay would be greater for patients with higher initial VLs. This may also suggest that patients with higher initial VLs have a larger proportion of actively replicating, productively-infected cells that are more susceptible to HAART. This is consistent with previous studies that examined decay for patients receiving HAART.[39] It was also intriguing that the cohort-level analyses also revealed that after the initial precipitous decline in VL there was a transient rebound, regardless of initial VL (Fig. 1C). This is due to a combination of individual profiles including a proportion of patients experiencing virologic rebound with subsequent resuppression, a small percentage experiencing rebound and not achieving resuppression, and some patients experiencing blips. It is unclear if this temporary, population-level rebound represents a specific temporal relationship with average time to medication fatigue and/or the development of virologic drug resistance as nearly half of treated patients experience a change in therapy around this time both as reported in this cohort[26] and elsewhere.[40] Even in the absence of complete rebound from poor adherence or drug resistance, periods of increased replication can occur due to pharmacologic changes or altered drug activity in a particular compartment.[29] Modeling data from structured treatment interruption trials have shown that parametric resonance, such as that seen in our study, can occur even in the absence of drug resistance and complete virologic rebound.[41] This phenomenon can be seen when a system undergoing small oscillations over time (such as during the dynamic equilibrium of viral load setpoint) undergoes a significant dampening (HAART initiation) and then experiences brief periods of external perturbation (brief treatment interruptions).

Although the importance of early virologic suppression and virologic failure on CD4 recovery has been well-described,[21], [22], [24], [42], much less is known about the impact of rate of decay or detectable VL after initial suppression of VL on CD4 recovery.[25], [38], [45], [46], [47], [48] In this study, we incorporated several of these elements into a single parameter of overall virologic decay. This parameter provides information on the early trajectory as well as the durability of the VL response after the initial decay. We also evaluated cumulative VL because it could be argued that it is the overall exposure to virus that influences CD4 recovery and AIDS.[38], [49], [50], [51] We found that cumulative VL was a stronger predictor of AIDS risk than CD4 recovery after HAART initiation. Additionally, VL decay or slope within the first year of HAART was not predictive of AIDS, whereas the overall VL decay predicted AIDS even among patients who attained VL suppression during 6 and 12 months of HAART. Thus, it is striking that the risk of AIDS is not impacted by the initial rapid phase of virologic decay, but rather by longitudinal assessments such as the overall decay or cumulative VL. These findings suggest that risk of AIDS during HAART is more sensitive to the VL over time rather than events that occur during the first year of HAART as has been suggested previously.[52], [53], [54], [55] In contrast to this study which examined the impact after HAART, Cole et al. recently found that cumulative VL predicted AIDS or death in absence of HAART independent of known risk factors in the Multicenter AIDS Cohort Study.[56] As the number of serious non-AIDS events during HAART increases relative to the number of AIDS events over time, it will be important to determine the association of overall virologic decay and cumulative VL with serious non-AIDS events as has been demonstrated with cancer[57], [58] and renal impairment.[59] This data would also suggest that perhaps the cumulative VL even prior to HAART could be associated with clinical events during HAART, supporting the notion that earlier diagnosis and treatment would further reduce the number of these adverse outcomes.

Even among subjects who attained VL suppression, and after adjustment for time to VL suppression, a slow overall VL decay was predictive of late/long-term CD4 changes (rate of CD4 gain, mean CD4 count after two years of HAART, and overall gain in CD4 cells), but not early CD4 changes (rate of CD4 gain during the first two years of HAART). In contrast, a slower VL decay within the first year of HAART associated with early but not later/long-term CD4 changes during HAART. These results suggest that, although among VL suppressers the pace and extent of CD4 gain during the early phases of HAART may be highly correlated with both the early and overall VL decay patterns, durable gains in CD4 cells after two years of HAART may be highly dependent on the overall VL decay pattern. These findings demonstrated that VL suppressors could be stratified into two categories such that those with both a slow early and overall VL decay (slow early/slow overall decay) will achieve CD4 recovery, but the gain in CD4 cells would be significantly muted relative to all other subjects (Fig. 1F). Because studies have identified polymorphisms that track the durability of CD4 recovery, it will be important to evaluate whether the patterns of VL decay are in part related to such host factors.[11]

The association of the extent of CD4 recovery was strongest with the overall VL decay and not the first year VL decay or the cumulative VL, suggests that these VL parameters may be capturing different aspects of VL changes during HAART (early trajectory and maintenance of suppression). The overall decay provides information throughout the duration of treatment and is not limited to one year of information. Hence, it is probable that the decay pattern occurring after virologic suppression (third phase of VL decay)[60] indexed to the decay patterns that occur immediately after HAART initiation[23] together contribute to the ability of a patient to experience durable immune reconstitution. It remains unclear if the latter phases of immune reconstitution are affected by “blips” or primarily by more substantial viral rebound.[61], [62]

In contrast to the overall decay pattern, the cumulative VL is a coarse measure of overall VL burden (total virus exposure) during HAART and does not account for VL decay patterns. For example, a patient who suppresses early but has late rebound might have a comparable cumulative VL to that of a patient with predominantly late virologic suppression. This may partly explain why this parameter as computed may not associate strongly with CD4 recovery. However, another explanation hinges on the use of detectable VLs to compute this parameter. Certainly, patients with complete or repetitive virologic rebounds may experience a loss of CD4 recovery; however, the vast majority of patients in this cohort achieved suppression within the first year and the rate of rebound was low.[26] Therefore, at the frequency of available measurements, the cumulative VL may not capture some of the intermittent or ongoing low-level viremia during HAART which may represent actual viral replication in the setting of periodic HAART interruption. Hence, it is conceivable that computation of the cumulative VL using more frequent measurements and/or single copy assays that assess VL below the detectable threshold of commercial assays might reveal that the cumulative VL is a more sensitive marker of not only AIDS risk but also CD4 recovery.

We investigated a large number of prospectively evaluated subjects who have equal access to healthcare and high rates of adherence to HAART.[26], [30] This afforded an excellent opportunity to observe the impact of virologic parameters on CD4 recovery in a setting outside of a clinical trial, making these results more generalizable to the HIV-infected population at large. There are some limitations of this study. This study did not attempt to dissect the components of the VL decay[24], [63] and determine what baseline and subsequent factors contribute to these components. For example, different regimens, the existence of drug resistance, variable pharmacokinetics and adherence patterns can result in different rates for the first and second phases of VL decay, respectively.[24], [28] To this end, we used HAART era as a covariate in the multivariate model to adjust for regimen potency and prior single or dual ART. Furthermore, although rates of adherence in this cohort[26] are high, the rationale of these analyses was not to understand the impact of adherence on the rate of decay but instead how the decay patterns alone influence subsequent clinical/immunologic outcomes regardless of the level of adherence which in a clinical setting can often be unreliable. We also acknowledge that we studied a total of 52 multivariate models (shown in Tables 2 and 3) and at a global type I error rate of 0.05, 2–3 observed associations are likely to be erroneous. Given the fact, however, that we observed a total of 23 associations to be significant at 0.05 type I error rate, our study results are unlikely to have been influenced by false positive associations due to multiple testing. Finally, although the impact of drug resistance and pharmacokinetic interactions was not examined in this study, prior ARV use was used as a surrogate marker of baseline resistance in the multivariate models.

In summary, our findings underscore that the early and overall patterns of VL decay among VL suppressed patients is an independent determinant of CD4 recovery. In addition, the cumulative VL is a determinant of AIDS risk during HAART. Thus, inter-individual differences in VL decay patterns may partly explain the wide variability in CD4 recovery even among those individuals achieving VL suppression within the recommended timeframe. These results also suggest that regimens that produce the most rapid virologic decay and durable suppression could lead to better clinical/immunologic responses. These parameters could be further developed to enhance clinical trial assessment of ARV regimens and assist clinicians with identifying patients at risk for adverse events beyond standard indicators.

Supporting Information

Table S1.

Definitions for various parameters used in this study.

https://doi.org/10.1371/journal.pone.0017956.s001

(DOCX)

Table S2.

Modeling of VL kinetics based on tertiles of pre-HAART VL.

https://doi.org/10.1371/journal.pone.0017956.s002

(DOCX)

Table S3.

Association of VL parameters with risk of AIDS development after initiation of HAART among seroconverters.

https://doi.org/10.1371/journal.pone.0017956.s003

(DOCX)

Note S1.

Statistical concepts in VL parameter estimation.

https://doi.org/10.1371/journal.pone.0017956.s004

(DOCX)

Acknowledgments

The authors would like to thank our patients for their enormous contributions over the years and the IDCRP HIV Working Group including: Susan Banks, Mary Bavaro, MD, Helen Chun, MD, Cathy Decker, MD, Connor Eggleston, Tomas Ferguson, MD, Heather Hairston, Cliff Hawkes, MD, Arthur Johnson, MD, Erica Johnson, MD, Alan Lifson, MD, MPH, Grace Macalino, PhD, Jason Maguire, MD, Scott Merritt, Robert O'Connell, MD, Sheila Peel, PhD, John Powers, MD, Roseanne Ressner, MD, Edmund Tramont, MD, Timothy Whitman, MD, and Michael Zapor, MD. The content of this publication is the sole responsibility of the authors and does not necessarily reflect the views or policies of the NIH or the Department of Health and Human Services, the DoD or the Departments of the Army, Navy or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government. This work is original and has not been published elsewhere. Portions were presented at the 17th Conference on Retroviruses and Opportunistic Infections, February 2010, San Francisco, California (Abstract #506).

Author Contributions

Conceived and designed the experiments: VCM HK SKA JFO MD. Analyzed the data: HK VCM GG JFO. Contributed reagents/materials/analysis tools: HK SKA BA MP. Wrote the paper: VCM HK SKA GG. Gathered clinical data: VCM JFO GW AG NCC ML MD BA IDCRP-HWG. Analysis interpretation and manuscript review: VCM GG JFO GW AG NCC MP ML SKA BA HK.

References

  1. 1. Anastos K, Barron Y, Miotti P, Weiser B, Young M, et al. (2002) Risk of progression to AIDS and death in women infected with HIV-1 initiating highly active antiretroviral treatment at different stages of disease. Arch Intern Med 162: 1973–1980.
  2. 2. Jacobson LP, Li R, Phair J, Margolick JB, Rinaldo CR, et al. (2002) 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 155: 760–770.
  3. 3. Egger M, May M, Chene G, Phillips AN, Ledergerber B, et al. (2002) Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 360: 119–129.
  4. 4. Anastos K, Barron Y, Cohen MH, Greenblatt RM, Minkoff H, et al. (2004) 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 140: 256–264.
  5. 5. Lundgren JD, Mocroft A, Gatell JM, Ledergerber B, D'Arminio Monforte A, et al. (2002) A clinically prognostic scoring system for patients receiving highly active antiretroviral therapy: results from the EuroSIDA study. J Infect Dis 185: 178–187.
  6. 6. Ghani AC, de Wolf F, Ferguson NM, Donnelly CA, Coutinho R, et al. (2001) Surrogate markers for disease progression in treated HIV infection. J Acquir Immune Defic Syndr 28: 226–231.
  7. 7. Lewden C, Raffi F, Cuzin L, Cailleton V, Vilde JL, et al. (2002) Factors associated with mortality in human immunodeficiency virus type 1-infected adults initiating protease inhibitor-containing therapy: role of education level and of early transaminase level elevation (APROCO-ANRS EP11 study). The Antiproteases Cohorte Agence Nationale de Recherches sur le SIDA EP 11 study. J Infect Dis 186: 710–714.
  8. 8. Hermankova M, Ray SC, Ruff C, Powell-Davis M, Ingersoll R, et al. (2001) HIV-1 drug resistance profiles in children and adults with viral load of <50 copies/ml receiving combination therapy. Jama 286: 196–207.
  9. 9. Barbour JD, Wrin T, Grant RM, Martin JN, Segal MR, et al. (2002) Evolution of phenotypic drug susceptibility and viral replication capacity during long-term virologic failure of protease inhibitor therapy in human immunodeficiency virus-infected adults. J Virol 76: 11104–11112.
  10. 10. Panel on Antiretroviral Guidelines for Adults and Adolescents (2009) Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Human Services.
  11. 11. Ahuja SK, Kulkarni H, Catano G, Agan BK, Camargo JF, et al. (2008) CCL3L1-CCR5 genotype influences durability of immune recovery during antiretroviral therapy of HIV-1-infected individuals. Nat Med 14: 413–420.
  12. 12. Tan R, Westfall AO, Willig JH, Mugavero MJ, Saag MS, et al. (2008) Clinical outcome of HIV-infected antiretroviral-naive patients with discordant immunologic and virologic responses to highly active antiretroviral therapy. J Acquir Immune Defic Syndr 47: 553–558.
  13. 13. Gazzola L, Tincati C, Bellistri GM, Monforte A, Marchetti G (2009) The absence of CD4+ T cell count recovery despite receipt of virologically suppressive highly active antiretroviral therapy: clinical risk, immunological gaps, and therapeutic options. Clin Infect Dis 48: 328–337.
  14. 14. Torti C, Cologni G, Uccelli MC, Quiros-Roldan E, Imberti L, et al. (2004) Immune correlates of virological response in HIV-positive patients after highly active antiretroviral therapy (HAART). Viral Immunol 17: 279–286.
  15. 15. Gilson R, Man SL, Copas A, Rider A, Forsyth S, et al. (2009) Discordant responses on starting highly active antiretroviral therapy: suboptimal CD4 increases despite early viral suppression in the UK Collaborative HIV Cohort (UK CHIC) Study. HIV Med 11: 152–60.
  16. 16. Kelley CF, Kitchen CM, Hunt PW, Rodriguez B, Hecht FM, et al. (2009) Incomplete peripheral CD4+ cell count restoration in HIV-infected patients receiving long-term antiretroviral treatment. Clin Infect Dis 48: 787–794.
  17. 17. Weverling GJ, Lange JM, Jurriaans S, Prins JM, Lukashov VV, et al. (1998) Alternative multidrug regimen provides improved suppression of HIV-1 replication over triple therapy. Aids 12: F117–122.
  18. 18. Polis MA, Sidorov IA, Yoder C, Jankelevich S, Metcalf J, et al. (2001) Correlation between reduction in plasma HIV-1 RNA concentration 1 week after start of antiretroviral treatment and longer-term efficacy. Lancet 358: 1760–1765.
  19. 19. Ghani AC, Ferguson NM, Fraser C, Donnelly CA, Danner S, et al. (2002) Viral replication under combination antiretroviral therapy: a comparison of four different regimens. J Acquir Immune Defic Syndr 30: 167–176.
  20. 20. Maggiolo F, Migliorino M, Pirali A, Pravettoni G, Caprioli S, et al. (2000) Duration of viral suppression in patients on stable therapy for HIV-1 infection is predicted by plasma HIV RNA level after 1 month of treatment. J Acquir Immune Defic Syndr 25: 36–43.
  21. 21. Demeter LM, Hughes MD, Coombs RW, Jackson JB, Grimes JM, et al. (2001) 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 135: 954–964.
  22. 22. Casado JL, Perez-Elias MJ, Antela A, Sabido R, Marti-Belda P, et al. (1998) Predictors of long-term response to protease inhibitor therapy in a cohort of HIV-infected patients. Aids 12: F131–135.
  23. 23. Wu H, Kuritzkes DR, McClernon DR, Kessler H, Connick E, et al. (1999) Characterization of viral dynamics in human immunodeficiency virus type 1-infected patients treated with combination antiretroviral therapy: relationships to host factors, cellular restoration, and virologic end points. J Infect Dis 179: 799–807.
  24. 24. Wu H, Huang Y, Acosta EP, Rosenkranz SL, Kuritzkes DR, et al. (2005) Modeling long-term HIV dynamics and antiretroviral response: effects of drug potency, pharmacokinetics, adherence, and drug resistance. J Acquir Immune Defic Syndr 39: 272–283.
  25. 25. Gutierrez F, Padilla S, Masia M, Iribarren JA, Moreno S, et al. (2006) Clinical outcome of HIV-infected patients with sustained virologic response to antiretroviral therapy: long-term follow-up of a multicenter cohort. PLoS One 1: e89.
  26. 26. Marconi VC, Grandits GA, Weintrob AC, Chun H, Landrum ML, et al. (2010) Outcomes of highly active antiretroviral therapy in the context of universal access to healthcare: the U.S. Military HIV Natural History Study. AIDS Res Ther 7: 14.
  27. 27. Kuritzkes DR, Sevin A, Young B, Bakhtiari M, Wu H, et al. (2000) Effect of zidovudine resistance mutations on virologic response to treatment with zidovudine-lamivudine-ritonavir: genotypic analysis of human immunodeficiency virus type 1 isolates from AIDS clinical trials group protocol 315.ACTG Protocol 315 Team. J Infect Dis 181: 491–497.
  28. 28. Kuritzkes DR, Ribaudo HJ, Squires KE, Koletar SL, Santana J, et al. (2007) Plasma HIV-1 RNA dynamics in antiretroviral-naive subjects receiving either triple-nucleoside or efavirenz-containing regimens: ACTG A5166s. J Infect Dis 195: 1169–1176.
  29. 29. Loveday C, Kaye S, Tenant-Flowers M, Semple M, Ayliffe U, et al. (1995) HIV-1 RNA serum-load and resistant viral genotypes during early zidovudine therapy. Lancet 345: 820–824.
  30. 30. Weintrob AC, Grandits GA, Agan BK, Ganesan A, Landrum ML, et al. (2009) Virologic Response Differences Between African Americans and European Americans Initiating Highly Active Antiretroviral Therapy With Equal Access to Care. J Acquir Immune Defic Syndr 52: 574–80.
  31. 31. Weintrob AC, Fieberg AM, Agan BK, Ganesan A, Crum-Cianflone NF, et al. (2008) Increasing age at HIV seroconversion from 18 to 40 years is associated with favorable virologic and immunologic responses to HAART. J Acquir Immune Defic Syndr 49: 40–47.
  32. 32. Centers for Disease Control and Prevention (1993) 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. JAMA 269: 729–730.
  33. 33. Giorgi JV, Lyles RH, Matud JL, Yamashita TE, Mellors JW, et al. (2002) Predictive value of immunologic and virologic markers after long or short duration of HIV-1 infection. J Acquir Immune Defic Syndr 29: 346–355.
  34. 34. Mellors JW, Munoz A, Giorgi JV, Margolick JB, Tassoni CJ, et al. (1997) Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 126: 946–954.
  35. 35. Phillips AN, Sabin CA, Elford J, Bofill M, Janossy G, et al. (1994) Use of CD4 lymphocyte count to predict long-term survival free of AIDS after HIV infection. Bmj 309: 309–313.
  36. 36. Vlahov D, Graham N, Hoover D, Flynn C, Bartlett JG, et al. (1998) Prognostic indicators for AIDS and infectious disease death in HIV-infected injection drug users: plasma viral load and CD4+ cell count. Jama 279: 35–40.
  37. 37. Hulgan T, Shepherd BE, Raffanti SP, Fusco JS, Beckerman R, et al. (2007) Absolute count and percentage of CD4+ lymphocytes are independent predictors of disease progression in HIV-infected persons initiating highly active antiretroviral therapy. J Infect Dis 195: 425–431.
  38. 38. Tomasoni LR, Patroni A, Torti C, Paraninfo G, Gargiulo F, et al. (2003) Predictors of long-term immunological outcome in rebounding patients on protease inhibitor-based HAART after initial successful virologic suppression: implications for timing to switch. HIV Clin Trials 4: 311–323.
  39. 39. Notermans DW, Goudsmit J, Danner SA, de Wolf F, Perelson AS, et al. (1998) Rate of HIV-1 decline following antiretroviral therapy is related to viral load at baseline and drug regimen. Aids 12: 1483–1490.
  40. 40. Vo TT, Ledergerber B, Keiser O, Hirschel B, Furrer H, et al. (2008) Durability and outcome of initial antiretroviral treatments received during 2000–2005 by patients in the Swiss HIV Cohort Study. J Infect Dis 197: 1685–1694.
  41. 41. Breban R, Blower S (2006) Role of parametric resonance in virological failure during HIV treatment interruption therapy. Lancet 367: 1285–1289.
  42. 42. Katzenstein DA, Hammer SM, Hughes MD, Gundacker H, Jackson JB, et al. (1996) The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200 to 500 CD4 cells per cubic millimeter. AIDS Clinical Trials Group Study 175 Virology Study Team. N Engl J Med 335: 1091–1098.
  43. 43. Kitchen CM, Kitchen SG, Dubin JA, Gottlieb MS (2001) Initial virological and immunologic response to highly active antiretroviral therapy predicts long-term clinical outcome. Clin Infect Dis 33: 466–472.
  44. 44. Marschner IC, Collier AC, Coombs RW, D'Aquila RT, DeGruttola V, et al. (1998) Use of changes in plasma levels of human immunodeficiency virus type 1 RNA to assess the clinical benefit of antiretroviral therapy. J Infect Dis 177: 40–47.
  45. 45. Le Moing V, Thiebaut R, Chene G, Leport C, Cailleton V, et al. (2002) 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 185: 471–480.
  46. 46. Kousignian I, Abgrall S, Duval X, Descamps D, Matheron S, et al. (2003) Modeling the time course of CD4 T-lymphocyte counts according to the level of virologic rebound in HIV-1-infected patients on highly active antiretroviral therapy. J Acquir Immune Defic Syndr 34: 50–57.
  47. 47. Havlir DV, Bassett R, Levitan D, Gilbert P, Tebas P, et al. (2001) Prevalence and predictive value of intermittent viremia with combination hiv therapy. Jama 286: 171–179.
  48. 48. Easterbrook PJ, Ives N, Waters A, Mullen J, O'Shea S, et al. (2002) The natural history and clinical significance of intermittent viraemia in patients with initial viral suppression to <400 copies/ml. Aids 16: 1521–1527.
  49. 49. Sproat M, Pozniak AL, Peeters M, Winters B, Hoetelmans R, et al. (2005) The influence of the M184V mutation in HIV-1 reverse transcriptase on the virological outcome of highly active antiretroviral therapy regimens with or without didanosine. Antivir Ther 10: 357–361.
  50. 50. Kravcik S, Magill A, Sanghvi B, Ogden R, Cameron WD, et al. (2001) Comparative CD4 T-cell responses of reverse transcriptase inhibitor therapy with or without nelfinavir matched for viral exposure. HIV Clin Trials 2: 160–170.
  51. 51. Kim S, Hughes MD, Hammer SM, Jackson JB, DeGruttola V, et al. (2000) Both serum HIV type 1 RNA levels and CD4+ lymphocyte counts predict clinical outcome in HIV type 1-infected subjects with 200 to 500 CD4+ cells per cubic millimeter. AIDS Clinical Trials Group Study 175 Virology Study Team. AIDS Res Hum Retroviruses 16: 645–653.
  52. 52. HIV Surrogate Marker Collaborative Group (2000) Human immunodeficiency virus type 1 RNA level and CD4 count as prognostic markers and surrogate end points: a meta-analysis. AIDS Res Hum Retroviruses 16: 1123–1133.
  53. 53. Journot V, Chene G, Joly P, Saves M, Jacqmin-Gadda H, et al. (2001) Viral load as a primary outcome in human immunodeficiency virus trials: a review of statistical analysis methods. Control Clin Trials 22: 639–658.
  54. 54. Zoufaly A, Stellbrink HJ, Heiden MA, Kollan C, Hoffmann C, et al. (2009) Cumulative HIV viremia during highly active antiretroviral therapy is a strong predictor of AIDS-related lymphoma. J Infect Dis 200: 79–87.
  55. 55. El-Sadr WM, Lundgren JD, Neaton JD, Gordin F, Abrams D, et al. (2006) CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med 355: 2283–2296.
  56. 56. Cole SR, Napravnik S, Mugavero MJ, Lau B, Eron JJ Jr, et al. (2010) Copy-years viremia as a measure of cumulative human immunodeficiency virus viral burden. Am J Epidemiol 171: 198–205.
  57. 57. Guiguet M, Boue F, Cadranel J, Lang JM, Rosenthal E, et al. (2009) Effect of immunodeficiency, HIV viral load, and antiretroviral therapy on the risk of individual malignancies (FHDH-ANRS CO4): a prospective cohort study. Lancet Oncol 10: 1152–1159.
  58. 58. Bruyand M, Thiebaut R, Lawson-Ayayi S, Joly P, Sasco AJ, et al. (2009) Role of uncontrolled HIV RNA level and immunodeficiency in the occurrence of malignancy in HIV-infected patients during the combination antiretroviral therapy era: Agence Nationale de Recherche sur le Sida (ANRS) CO3 Aquitaine Cohort. Clin Infect Dis 49: 1109–1116.
  59. 59. Choi AI, Shlipak MG, Hunt PW, Martin JN, Deeks SG (2009) HIV-infected persons continue to lose kidney function despite successful antiretroviral therapy. Aids 23: 2143–2149.
  60. 60. Siliciano JD, Kajdas J, Finzi D, Quinn TC, Chadwick K, et al. (2003) Long-term follow-up studies confirm the stability of the latent reservoir for HIV-1 in resting CD4+ T cells. Nat Med 9: 727–728.
  61. 61. Nettles RE, Kieffer TL, Kwon P, Monie D, Han Y, et al. (2005) Intermittent HIV-1 viremia (Blips) and drug resistance in patients receiving HAART. Jama 293: 817–829.
  62. 62. Rong L, Perelson AS (2009) Modeling latently infected cell activation: viral and latent reservoir persistence, and viral blips in HIV-infected patients on potent therapy. PLoS Comput Biol 5: e1000533.
  63. 63. Coffin JM (1995) HIV population dynamics in vivo: implications for genetic variation, pathogenesis, and therapy. Science 267: 483–489.