|Home | About | Journals | Submit | Contact Us | Français|
Combination antiretroviral therapy suppresses but does not eradicate human immunodeficiency virus type 1 (HIV-1) in infected persons, and low-level viremia can be detected despite years of suppressive antiretroviral therapy. Short (28d) course antiretroviral intensification of standard combination therapy is a useful approach to determine whether complete rounds of HIV-1 replication in rapidly cycling cells contribute to persistent viremia. We investigated whether intensification with the integrase inhibitor raltegravir decreases plasma viremia in patients on suppressive antiretroviral therapy.
Subjects (N=10) with long term HIV-1 suppression on combination antiretroviral regimens were intensified for 4 weeks with raltegravir. Plasma viremia was determined prior to, during, and following the 4-week intensification period using a sensitive HIV-1 RNA assay (limit of detection of 0.2 copies HIV-1 RNA/ml plasma). A four week intensification course was chosen to investigate potential HIV-1 replication in cells with relatively short (c. 1-14 d) half-lives.
There was no evidence in any subject of decline in HIV-1 RNA levels during the period of raltegravir intensification or rebound after discontinuation. Median levels of HIV-1 RNA prior to (0.17 log10 copies/ml), during (0.04 log10 copies/ml), and following (0.04 log10 copies/ml) raltegravir intensification (p >0.1 for all comparisons in parametric analyses). HPLC/mass spectroscopy studies confirmed that therapeutic levels of raltegravir were achieved in plasma during intensification.
Intensification of antiretroviral therapy with a potent HIV-1 integrase inhibitor did not decrease persistent viremia in subjects on suppressive regimens, indicating that rapidly cycling cells infected with HIV-1 were not present. Eradication of HIV-1 from infected persons will require new therapeutic approaches.
Combination antiretroviral therapy (ART) results in marked suppression of viremia in persons with HIV-1 infection. Therapy is not curative, however, and detectable viremia and replication competent HIV-1 persists despite ART-induced suppression [1-7]. The origin of persistent viremia on therapy is uncertain; potential sources include ongoing complete cycles of HIV-1 replication [8-13], long-lived reservoirs of chronically infected cells, [1, 3, 4, 7, 14-17] sanctuary sites into which antiretrovirals have poor penetration[18-24], or a combination of these possibilities. Understanding the source and mechanisms of viral persistence on antiretroviral therapy has critical implications for future therapeutic approaches. If persistent viremia is derived from cycles of active HIV-1 replication, improving the potency and penetration of drugs that block new cycles of replication is essential. By contrast, if new cycles of viral replication are completely suppressed by current therapy and viremia is derived from reservoirs of long-lived chronically infected cells, new strategies are necessary to cure infection.
Introduction of antiretroviral therapy results in rapid and profound decreases in plasma viral RNA levels, followed by continued declines in viremia with slower rates of decline. As previously reported, viral decay rates are the result of death and elimination of infected cells with short (1-1.2 d), intermediate (14 d) and prolonged (39 week) half- lives[25-28] . Previously we reported a stable level of persistent viremia (essentially infinite half-life) following the third phase decline in patients on suppressive therapy for >3-4 y . These data suggested that all short-, and intermediate-lived HIV-1 infected cells had been eliminated after 3-4 y, and that persistent viremia was the product of long-lived cells with integrated proviruses.
To investigate whether ongoing cycles of HIV-1 infection continue during suppressive therapy, we conducted a trial of antiretroviral intensification using inhibitors of HIV-1 reverse transcriptase (efavirenz, EFV) or protease (atazanavir/ritonavir, ATV/r, or lopinavir/r, LPV/r). These studies demonstrated that using EFV, LPV/r or ATV/r to intensify therapy did not lower viremia, indicating that additional inhibition of either the reverse transcription or protease cleavage steps in viral replication does not further inhibit HIV-1 production. These studies were consistent with the hypothesis that persistent viremia on therapy is the product of long-lived (>14 days) chronically infected cells.
It is possible that unintegrated HIV-1 DNA generated before antiretroviral therapy is started may persist for periods after therapy is initiated, and that slow integration of proviruses over time may continue to provide a population of cells that will produce HIV-1 and contribute to persistent viremia. The presence of a substantial number of short-lived cells has been suggested based on analyses of HIV-1 decay kinetics in patients undergoing therapy with the integrase inhibitor raltegravir[30-32]. To investigate whether inhibition of viral DNA integration decreases persistent viremia, we have studied the effects of therapy intensification with raltegravir, a potent HIV-1 integrase inhibitor, in ten chronically infected patients with persistently detectable viremia on standard combination antiretroviral therapy.
HIV-infected individuals >18 years of age on stable, suppressive antiretroviral therapy with persistent plasma viremia >0.6 copies/ml by single copy assay at screening were enrolled at the University of Pittsburgh Clinical Trials Unit between December 2007 and January 2009. All participants had screening CD4 cell counts >200 cells/μl and were not receiving prophylaxis for opportunistic infections. Study participants had viral RNA levels <50 copies/ml plasma for at least 12 months prior to screening, had no prior exposure to raltegravir, history of HIV-1 drug resistance, febrile illness within 3 weeks prior to enrollment, or vaccination within 6 weeks prior to enrollment. Participants with significant co-morbid illness, such as chronic hepatitis B infection were excluded. Raltegravir was provided by Merck & Co., Inc. Sample size was based on prior experience with single copy assay results and number of suppressed patients who may undergo further suppression on antiretroviral intensification; if ten participants completing the study had no decline in viremia, we estimated with 90% power that the probability of any individual with persistent viremia would be suppressible by raltegravir is <0.21. Based on these estimates, a ceiling of 10 participants completing the study was proposed and approved by NIH and the Pittsburgh Data Safety Monitoring Board.
The study (NCT00618371) was approved by the University of Pittsburgh Institutional Review Board (FWA00006790), and by the NIH Office of Human Subjects Protection, and was performed under Food and Drug Administration Investigational New Drug (IND) application held by the University of Pittsburgh. Participants provided written informed consent.
Study participants had samples collected for single copy assay (SCA) weekly for 3 weeks prior to intensification. On day 0 of intensification, participants underwent phlebotomy, then initiated raltegravir therapy 400 mg orally twice daily. Plasma was sampled on days 7, 14, 21, and day 29, the morning after the last dose of raltegravir. Following intensification, optional additional sampling was performed on days 29, 30, 35, 42, 49, 58. Samples for single copy assay were pelleted immediately following phlebotomy and plasma was frozen at −70°C. Standard immunophenotyping was performed on samples from study days 0, 29, 58. Routine safety laboratory studies were performed and graded for severity using the Division of AIDS table (December 2004 version.
HIV-1 RNA was measured by Amplicor HIV-1 Monitor assay v1.5 (Roche), and by SCA as described. Briefly, 10 ml plasma with RCAS virus added as an internal standard was centrifuged at 100,000 x g and the pellet was extracted and subjected to cDNA synthesis followed by real-time PCR amplification of a 79 bp region of HIV-1 Gag, or a portion of the RCAS genome. Sample recovery was measured by RCAS amplification and detection, and HIV-1 RNA levels were determined using a standard curve constructed with HIV-1 of known RNA copy number.
Samples for raltegravir levels were obtained on days 0, 14, and 29. An assay was developed to quantify raltegravir concentrations in human plasma using liquid–liquid extraction paired with HPLC separation and MS/MS detection. The dynamic range of the assay was 1 to 3000 ng/mL, with a coefficient of determination (r2, mean±SD) of 0.9992±0.0002. The mean precision values for calibration standards ranged from 0.6% to 3.0%, while accuracy values were 96.5–104.3%.
Based on our experience with patients on suppressive antiretroviral therapy, we anticipated that there would be a non-negligible number of measurements below the 0.2 copy/mL limit of quantitation (LOQ) even using the sensitive single copy assay. Consequently, we planned for several different analyses to evaluate the impact of therapy intensification. We first considered plasma HIV-1 RNA level as a binary variable (above or below 0.2 copy/mL) and used repeated-measures regression models with generalized estimating equations to determine if the proportion of samples < 1 copy/ml increased over time, adjusting for the correlation within individuals. Second, we also used parametric mixed regression models with left censoring limits that enabled us to more fully model the distribution of the HIV-1 RNA data. We utilized an extension of the model used by Hughes  for repeated measures as implemented by Thiébaut. In this model, we assume HIV-1 RNA values (log10 copies/ml) can be described by a normal distribution but incorporate both censoring below the 0.2 copy/mL limit and correlation from repeated measurements using a random intercept term.
Thirty one persons were screened for the study and thirteen participants on combination antiretroviral therapy with HIV-1 RNA > 0.6 copies/ml were enrolled; the remainder had viremia ≤0.6 c/ml or other exclusion criteria. Ten participants completed the study and were analyzed; three participants prematurely discontinued the study for personal reasons.
Baseline demographic characteristics of study participants are presented in Table 1; individuals were predominantly male, had high pre-therapy viremia (median HIV-1 RNA 5.0 log10 copies/ml), and long-term suppression of viremia on ART (median of 9 years) with stable CD4+T-cell numbers (median 473.5 cells/μl on day 0 prior to intensification). All participants were on combination antiretroviral therapy with combination antiretroviral therapy; 9 of 10 participants were receiving EFV-based therapy and one participant was receiving NRTI+NNRTI + lopinavir/ritonavir.
Prior to intensification, participants were sampled weekly for four weeks to establish the baseline viremia; the median level of viremia of all ten participants was 1.5 copies/ml. All 10 participants had viral RNA levels ≥1 copy/ml during the screening visit, but only 9 participants had viral RNA levels ≥1 copy/ml for the majority of time points before intensification. Such variation in viremia has been noted previously. All tolerated the raltegravir addition without significant adverse events; no laboratory abnormalities exceeding grade II and no missed doses were reported. Figure 1 shows the longitudinal HIV-1 RNA values for all ten participants. There was no evidence in any subject of decline in HIV-1 RNA levels during the intensification period or rebound following discontinuation. There were no significant differences in median HIV-1 RNA prior to (0.17 log10 copies/ml), during (0.04 log10 copies/ml), or following (0.04 log10 copies/ml) intensification (p >0.1 for all comparisons). As noted above, one participant (#6) had HIV-1 RNA below the limit of quantification for the majority of time points before, during and after intensification. A secondary analysis excluding data from this subject did not change the results of the comparisons. To summarize, neither the individual participant results, nor the aggregate data, showed a statistically significant change in median viremia before, during, or after treatment intensification. These data suggested that HIV-1 was not derived from rapidly cycling, short lived cells (half life 1-10 d) that are responsible for the majority (>99%) of viremia in untreated individuals.
CD4 T cell subsets were determined at entry, on the final day of intensification, and on day 58, four weeks following intensification (data not shown). Median CD4 counts prior to, during and following raltegravir intensification were 580 cell/μl, 623 cells/μl, and 605 cells/μl, respectively; these observed increases were not significantly different from pre-therapy samples (p=0.68-0.86 for all comparisons).
To investigate whether participants undergoing intensification achieved therapeutic levels of raltegravir, plasma samples prior to, during and directly after raltegravir intensification were tested for raltegravir concentrations by HPLC/mass spectroscopy. As shown in figure 3, pre-intensification raltegravir levels were below the limit of quantitation. Therapeutic levels of raltegravir were achieved in all subjects by day 14 of intensification (Figure 3). The mean raltegravir concentration was 1368.4 nM (range 55.8 -3847 nM), which is 30-fold higher than the reported raltegravir IC95 of 33 nM. Raltegravir levels declined after the intensification period, but were still present above the IC95 at 24 hours following discontinuation of raltegravir (Figure 3), consistent with the reported half life of 7-12 h.
Currently approved antiretroviral drugs are designed to inhibit new cycles of viral replication but do not block virus production from cells that are already infected. In various combinations, these drugs reduce viremia >17,000-fold . Despite the potency of these combinations, low level-viremia can be detected in the majority of patients on stable suppressive ART for years. Possible sources of this viremia include virus production from long-lived cells containing integrated proviruses [3, 4, 17] or from ongoing cycles of viral replication in “sanctuary” sites into which some antiretroviral agents penetrate incompletely or not at all[8-13]. To distinguish between these potential sources, we used antiretroviral intensification as means to determine whether residual viremia is affected. In our first intensification trial, we added potent inhibitors of HIV-1 protease or reverse transcriptase as intensification agents and observed that viremia did not decrease during intensification periods of 4-12 weeks. These data strongly suggest that ongoing completes cycles of replication in rapidly cycling cells are not the major source of persistent viremia.
These observations do not exclude the possibility, however unlikely, that unintegrated HIV-1 DNA persists and following integration, could serve as a source of virus production despite standard or intensified therapy with HIV-1 RT or protease inhibitors. In this regard, the half-life of unintegrated HIV-1 DNA has been investigated in a number of studies, and the half-life of unintegrated DNA has been estimated at 4.8 d . Additional support for a similar half-life of unintegrated HIV-1 DNA comes from analysis of viremia decay kinetics following combination therapy with raltegravir [30, 31, 39]. These data suggested that the second phase comprised cells containing long-lived unintegrated DNA that was blocked from becoming integrated by raltegravir, resulting in the absence of virus production from these cells.
We therefore hypothesized that inhibiting integration by intensification with raltegravir could block the potential contribution of relatively long-lived HIV DNA intermediates to virus production from rapidly cycling cells. The results of the current study, which showed no effect of raltegravir on residual viremia, do not support this hypothesis. In addition, this result provides additional evidence that new, complete cycles of HIV-1 replication in rapidly cycling cells are not the major mechanism of persistent viremia in patients on long-term suppressive ART.
One potential limitation of our study is its small size. We intensified 10 participants with raltegravir and no change in viremia was observed in 9 evaluable participants. It is possible that a larger study would identify a subgroup with residual viremia that is responsive to raltegravir intensification. Based on 0 of 9 evaluable participants responding to raltegravir intensification, we estimate that the probability of such a subgroup is <0.21.
The duration of intensification (4 weeks) was chosen to identify changes in viremia arising from virus-producing cells with a short half-life (1-2 days), which comprise >99% of infected cells in untreated HIV-1 infection [25-27]. If all the newly infected cells had a half life > c. 10-14 days, we would be unable to detect a decline in viremia. This scenario is unlikely, however, because it would requires that raltegravir preferentially inhibit new infection of short-lived but not long-lived cells. A larger study with a longer (12 week) intensification period is in progress to investigate effects of prolonged raltegravir therapy in suppressed patients.
The majority of our participants were taking RT inhibitor therapy with the combination of tenofovir + FTC + efavirenz as a single tablet (Atripla™), which is the most widely prescribed initial therapy for HIV-1 infection. Our results may not be generalizable to other treatment regimens, although the level of residual viremia observed in the current study is very similar to that observed in patients on a variety of suppressive treatment regimens [2-4, 29].
Another potential limitation is that persistent viremia may arise from anatomic sites into which neither the standard regimen nor the intensifying agent penetrates sufficiently to suppress HIV-1 replication completely [18-24]. Ongoing studies, notably the CHARTER study, have investigated drug penetration into sanctuary sites [22, 24]. A gradient of penetration of drug into cerebrospinal fluid has been described for a number of antiretrovirals. Similar studies have not yet been completed with raltegravir, although several ongoing trials will shed new light on this subject. Ongoing HIV-1 replication in “sanctuary sites” such as the central nervous system, gut-associated lymphoid tissue, and the genitourinary tract are not absolutely excluded by the current study result, and differential levels of drug transporter activity, cellular activation, and segregation of immune cells could contribute to differences in antiretroviral drug efficacy in an anatomic-specific fashion. It is notable, however, that participants in the current study were suppressed for a median of 9 y or approximately 1700 HIV generations (generation time = 1.5- 2 d). With a mutation rate in the range of c. 3 × 10-5 mutations/site per cycle, even a relatively small population of virus undergoing new cycles of replication would have a reasonable probability to generate drug resistance mutations, resulting in probable drug failure within this period. However, previous genetic analyses of virus in chronically-infected individuals who are suppressed on ART have not identified the emergence of drug resistance mutations [6, 29].
Although our analyses did not detect changes in HIV-1 viremia in this or our prior intensification study, we have detected changes in viremia by single copy assay in individuals undergoing regimen simplification of standard combination therapy to atazanavir/ritonavir monotherapy. Increases in plasma HIV-1 RNA occurred 4-12 weeks before viremia became detectable in commercial assays. These data demonstrate that the single copy assay can detect small changes in residual HIV-1 RNA in patients who simplify to a less potent regimen.
The absence of a detectable effect of antiretroviral drug intensification on HIV-1 viremia on therapy indicates that neither ongoing, complete cycles of infection nor delayed integration of viral DNA formed pretherapy in actively cycling cells is a major source of persistent viremia. Instead, viremia is most likely sustained by long-lived cells containing integrated proviruses. Antiretroviral drugs that only block new cycles of replication, such as currently used inhibitors of virion attachment, fusion, reverse transcription, integration or proteolytic cleavage are not useful for eliminating long-lived virus producing cells. Rather, new approaches targeting infected cells directly, such as with immunotoxins [41, 42]or activation- or immune-based therapies are likely needed to reduce or eliminate important HIV-1 reservoirs.
We thank J. Dinoso, R. Siliciano, J. Kovacs, R. Davey, S. Migueles, I. Sereti, R. Leavitt, H.C. Lane, and H. Masur for helpful discussions, Merck & Co., Inc. for providing raltegravir, and especially the volunteers for participating in this study.
Financial Disclosure This work was supported by NIAID Contract HHSN261200800001E (to D.M.).
Author ContributionsJ. Jones Authored protocol, enrolled patients
D. McMahon Authored protocol, administered protocol, co-authored manuscript
Wiegand, A. Performed laboratory analyses
Gange, S. J. Performed statistical analyses
Kearney, M. Performed laboratory analyses
Palmer, S. Performed laboratory analyses
McNulty, S. Study coordinator for study
Metcalf, J., Provided contract and administrative support
Acosta, E. Performed drug level determinations
Rehm, C. Provided administrative and laboratory support
Coffin, J. M. Planned experiments, analyzed data, and authored manuscript
Mellors, J. W. Planned experiments and analyzed data and authored manuscript
Maldarelli, F. Planned experiments, authored the protocol, analyzed data, and authored manuscript