Early virologic suppression with three-class experienced patients: 24-week effectiveness in the darunavir outcomes study

Abstract

Objective: 

To evaluate 24-week virologic effectiveness of novel antiretroviral regimens for treatment of three-class experienced adult patients in a clinical practice setting following the US Food and Drug Administration approval of darunavir (DRV) for this population.

Design: 

A prospective cohort study.

Setting: 

A single-center, academic HIV clinic.

Participants: 

Three-class antiretroviral-experienced patients changing regimens between July 2006 and May 2008. Sociodemographic, psychosocial, and clinical characteristics were collected at baseline and during prospective follow-up.

Outcome measures: 

Plasma HIV viral load below 50 copies/ml and change in CD4 cell count at 24 weeks following regimen change. The Stanford Genotype Database was used to analyze HIV genotype resistance results and determine the number of active drugs in each regimen. Multivariate models and propensity score methods were employed to assess outcome measures.

Results: 

Among 109 three-class experienced patients, who previously received an average of 10.5 prior antiretrovirals, 55% achieved viral load below 50 copies/ml at 24 weeks. Treatment strategy was classified as nonprotease inhibitor (n = 25), DRV/ritonavir (DRV/r) (n = 51), or other protease inhibitor (n = 33). The number of active drugs was not significantly different across strategies (P = 0.24). In multivariate analysis, patients treated with DRV/r (65%, odds ratio = 4.24 vs. nonprotease inhibitor strategy, 95% confidence interval = 1.28–14.06), raltegravir (65%, odds ratio = 3.10, 95% confidence interval = 1.12–8.62), or both were more likely to achieve viral load below 50 copies/ml.

Conclusion: 

Among antiretroviral-experienced patients changing regimens, those treated with DRV/r, raltegravir, or both were more likely to achieve a viral load below 50 copies/ml at 24 weeks. The effectiveness of these agents in routine clinical care mirrors their efficacy in clinical trials and has ushered in a new era in the therapy of three-class experienced patients.

Introduction

Since 2006, the arrival of several novel antiretroviral agents, namely darunavir (DRV), raltegravir (RAL), maraviroc (MRV), and etravirine (ETR), has transformed the landscape of therapeutic options for treatment-experienced patients with HIV infection. In randomized controlled clinical trials (RCTs) ^[1–5], these newer medications have not only been well tolerated but also have been associated with exceptional levels of virologic success. The Performance Of TMC114/r When evaluated in treatment-Experienced patients with protease inhibitor Resistance (POWER)-1 and POWER-2 trials enrolled treatment-experienced patients previously exposed to protease inhibitors and documented superior virologic suppression with DRV/ritonavir (DRV/r) over comparator protease inhibitors (CPIs); at week 24 [HIV-1 RNA < 50 copies/ml, DRV/r 53%, CPI 18% (POWER 1); DRV/r 39%, CPI 7% (POWER 2)] and week 48 (combined POWER 1 + 2 analysis: DRV/r 45%, CPI 10%) ^[1,6,7]. In a trial ^[2] of three-class experienced patients, the integrase inhibitor RAL controlled viral replication more effectively than placebo in combination with an optimized background regimen (OBR) at 16 weeks (HIV-1 RNA < 50 copies/ml: RAL, 62%; placebo, 35%) and 48 weeks (RAL, 62%; placebo, 33%). In other studies ^[3–5] of treatment-experienced patients, the chemokine (C–C motif) receptor 5 (CCR5) antagonist, MRV, and the second-generation nonnucleoside reverse transcriptase inhibitor (NNRTI), ETR, have both demonstrated superior efficacy when used as part of a regimen for patients with limited treatment options.

Partly on the basis of the findings of these RCTs and because of the availability of several novel, highly efficacious antiretroviral drugs, treatment guidelines were revised in 2006 endorsing, for the first time, complete virologic suppression (<50 copies/ml) as the goal of therapy for all patients, including those with extensive treatment experience ^[8,9]. Beyond clinical trials, a paucity of data exists on the performance of these new antiretroviral agents in routine clinical practice settings (effectiveness). Therefore, we designed a prospective cohort study, the Darunavir Outcomes Study (DOS), to evaluate the virologic and immunologic outcomes of highly antiretroviral-experienced patients changing regimens after June 2006, when DRV was approved for use in the clinical management of treatment-experienced adults in the United States.

Methods

The University of Alabama at Birmingham (UAB) 1917 Clinic Cohort is an ongoing, institutional review board (IRB)-approved, prospective cohort study that collects psychosocial, sociodemographic, and clinical information from patients receiving ambulatory services at the UAB outpatient HIV clinic (www.uab1917cliniccohort.org). The cohort was established in 1988 and has been described in detail previously ^[10]. Since 2004, the UAB 1917 Clinic Cohort has captured data at the point-of-care through a locally programmed electronic medical record that imports all laboratory values directly from the central UAB laboratory, includes electronic prescription for all medications, and contains detailed encounter notes. The electronic medical record and database are 100% quality assured, whereby all provider notes and prescriptions are reviewed within 72 h of entry into the system to ensure accurate data capture. Providers have complete autonomy in antiretroviral regimen selection as no clinic-based therapeutic recommendations or treatment algorithms exist.

The current study is nested within the UAB 1917 Clinic Cohort and was sponsored by Tibotec Therapeutics. The study was approved by the UAB IRB, and all participants provided written informed consent. According to the study protocol, participants completed a questionnaire at each visit and had extra blood drawn at each routine clinic visit for future studies. The scheduling of appointments and all treatment decisions were made at the discretion of clinic providers with no influence from research personnel. The study was observational in design and did not provide antiretroviral medications, laboratory tests, or remuneration to participants. Representatives from Tibotec Therapeutics participated in the study design, data interpretation, and manuscript preparation but did not contribute to data collection or analysis. Academic authors made final determination of manuscript content.

Study sample and procedures

Three-class antiretroviral-experienced patients (patients who had been prescribed medications from at least three of the following drug classes: NRTI, protease inhibitor, entry inhibitor, and NNRTI for at least 15 days) changing regimens between 1 July 2006 and 30 April 2008 were prospectively enrolled in this study. Providers had access to all available patient-level clinical data, ordered resistance testing at their discretion, and chose the subsequent antiretroviral regimen without the involvement of study personnel. Study inclusion criteria required a plasma viral load above 1000 copies/ml. There were no CD4 cell count exclusion criteria. Patients were excluded from the study if they were enrolled in a clinical trial evaluating antiretroviral medications.

Study variables

At study enrollment, the following sociodemographic and clinical data were obtained: age, sex, race, health insurance, baseline plasma HIV viral load (copies/ml), and baseline CD4⁺ T-lymphocyte count. Antiretroviral treatment history was determined by electronic database search and confirmed by medical record abstraction. If an HIV resistance assay was obtained prior to regimen change, genotypic and phenotypic data were recorded by study personnel in an electronic database. Genotypic mutations were analyzed with the Stanford Genotype Database to determine the number of active drugs in the study regimen. An antiretroviral agent was classified as active if the Stanford Genotype Database score was below 30 and inactive if the score was at least 30 ^[11]. At the time of data analysis, the Stanford Genotype Database did not have a model to evaluate RAL resistance. RAL was considered an active agent for all study participants as none of the patients had previously been exposed to an integrase inhibitor. Antiretroviral regimen composition variables used in the primary analysis were the following: treatment strategy, defined as protease inhibitor-sparing, DRV/r-containing, or other protease inhibitor-containing; use of other novel antiretroviral agents; and the number of active drugs in the regimen. Because MRV and ETR were prescribed in less than 20 participants, these agents were not evaluated in analyses.

The primary outcome variables were plasma HIV viral load below 50 copies/ml and change in CD4 cell count at 24 weeks (−8 weeks and +24 weeks) following initiation of a new antiretroviral regimen. Plasma HIV-1 RNA was quantified (lower limit ≤50 copies/ml) using the ultrasensitive Roche Amplicor Monitor (Roche Molecular Systems, Inc., Pleasanton, California, USA). The CD4⁺ T-lymphocyte count was calculated using a dual platform method with relative CD4⁺ percentage determined by FACScaliber acquisition using MULTIset software (BD Biosciences, San Jose, California, USA). The UAB central laboratory conducted all laboratory studies.

Statistical analysis

Descriptive statistics were employed to evaluate distributions of study variables. All analyses were completed with an intent to continue initial treatment approach. Sociodemographic and regimen characteristics were compared between patients who started a nonprotease inhibitor-based, DRV/r-containing, and other protease inhibitor-containing regimen using analysis of variance and chi-squared methods. Because of the modest sample size and emphasis on clinical and regimen characteristics, propensity score methods were employed to determine a composite (propensity) score for DRV/r receipt on the basis of age, sex, race, and insurance provider ^[12]. Multivariate models controlling for propensity score allowed for adjustment for these sociodemographic characteristics while using fewer degrees of freedom in multivariate models, thereby avoiding model overfitting.

Unadjusted analyses using chi-squared logistic regression and a multivariable logistic regression model were used to evaluate factors associated with achieving 24-week virologic suppression (viral load ≤50 copies/ml). Student t-tests and multivariable linear regression were used to assess change in CD4 cell count at 24 weeks. For the primary analyses, missing data were not imputed and as such were not used in analyses. In addition, sensitivity analyses for worst-case scenarios, that is, missing is equal to failure, were performed to assess the impact of missing data on study findings. Statistical tests were run using SAS software version 9.0 (SAS Institute Inc., Cary, North Carolina, USA). All P values were two-sided, with values less than 0.05 considered statistically significant.

Results

Among 109 study participants, 83% were men, 50% were minority race/ethnicity, and 61% lacked private health insurance (Table 1). On average, patients had received 10.5 prior antiretroviral medications.

Table 1: Baseline characteristics of 109 three-class antiretroviral-experienced patients changing regimens at the University of Alabama at Birmingham 1917 HIV/AIDS Clinic, July 2006–May 2008.

With regard to the new regimens, 39% contained one or two active drugs and 41% contained three or four active drugs according to genotypic sensitivity score (no resistance data available for 20% of patients). Among the new regimens, 25 patients (23%) received a nonprotease inhibitor-based regimen, 51 (47%) received a DRV/r-based regimen, and 33 (30%) received another protease inhibitor-based regimen (atazanavir/r n = 15, atazanavir n = 4, lopinavir/r n = 12, fosamprenavir/r n = 2, or nelfinavir n = 1), including one patient who received a dual protease inhibitor regimen containing atazanavir and lopinavir/r (Table 1). Patients treated with DRV/r were more likely to be men and had higher baseline plasma HIV viral loads relative to the other groups (Table 1). Significant differences in prescribing patterns of RAL, enfuvirtide (T-20), and ETR were observed between treatment arms, with RAL and ETR prescribed less frequently in the other protease inhibitor group. MRV was used too infrequently to be included in the analysis of prescribing patterns (n = 1). Of note, the number of active drugs in the study regimen was not significantly different between the treatment strategies (P = 0.24). From the original population, 10 patients had missing data at 24 weeks, but continued to receive care in the clinic (total patients used in subsequent analyses, n = 99).

Virologic success (<50 copies/ml) was achieved by 54 of 99 patients (55%) with available 24-week viral load measures. Roughly two-third of patients with DRV/r-based regimens (65%) and RAL-containing regimens (65%) each achieved plasma viral load below 50 copies/ml at 24 weeks (Table 2 and Fig. 1), whereas 71% of patients treated with both DRV/r and RAL as components of their regimen (n = 35) achieved viral suppression (data not shown). In unadjusted analyses, regimens containing three or four active agents and RAL-containing regimens were significantly associated with 24-week viral load suppression; lower baseline viral load and DRV/r-containing regimens were of borderline statistical significance. In multivariable logistic regression analysis, DRV/r-based regimens [odds ratio (OR) = 4.24 vs. protease inhibitor-sparing regimens, 95% confidence interval (CI) = 1.28–14.06] and RAL-containing regimens (OR = 3.10 vs. protease inhibitor-sparing regimens, 95% CI = 1.12–8.62) were significantly associated with achieving an undetectable plasma HIV viral load at 24 weeks. Sensitivity analysis of 24-week viral load suppression using a missing is equal to failure approach yielded consistent findings to primary analyses, with parameter estimates generally of comparable magnitude (Table 3). However, in the sensitivity analysis, lower baseline viral load achieved statistical significance with undetectable viral load at 24 weeks in both unadjusted and adjusted analyses.

Table 2: Factors associated with plasma HIV viral load below 50 copies/ml at week 24 among 99 patients with available 24-week viral load measures in unadjusted and multivariable logistic regression analyses (missing = missing).

Fig. 1: Antiretroviral regimen factors associated with plasma HIV viral load below 50 copies/ml at 24 weeks. PI, protease inhibitor.

Table 3: Sensitivity analysis of factors associated with plasma HIV viral load below 50 copies/ml at week 24 among 109 patients in unadjusted and multivariable logistic regression analyses (missing = failure).

For the entire group, the mean increase in CD4 cell count at 24 weeks was 48.1 cells/μl. Patients with baseline plasma viral load above 100 000 copies/ml had more robust increases in CD4 cell count relative to those with viral loads below 100 000 copies/ml. Regimen characteristics did not appear to be significantly associated with a change in CD4 cell count at 24 weeks (Table 4). Sensitivity analysis evaluating CD4 cell count as an increase in more than 50 cells yielded similar findings to primary analyses (data not shown).

Table 4: Factors associated with change in CD4 cell count at 24 weeks among 109 study patients in unadjusted and multivariable linear regression analyses.

Antiretroviral medication adherence at week 24 captured by a standardized self-reported adherence measure (AIDS Clinical Trial Unit 4, ACTU-4) was available for 68 study participants. Patients reporting the last missed dose of antiretroviral medication occurring within 2 weeks from the questionnaire date were recorded as being nonadherent. Antiretroviral medication adherence varied by treatment strategy, with 69% of patients whose regimen included DRV/r being adherent (n = 39), 50% of those receiving other protease inhibitors being adherent (n = 16), and 62% of patients on protease inhibitor-sparing regimens being adherent (n = 13) (P = NS). Among participants whose regimens included RAL (n = 46), 72% were adherent compared with 46% on non-RAL regimens (n = 22) (P < 0.05).

Discussion

Over half of highly treatment-experienced patients changing antiretroviral regimens at the UAB 1917 Clinic since July 2006 achieved undetectable plasma viral loads (<50 copies/ml) at 24 weeks, including roughly two-third of patients whose regimens contained DRV/r, RAL, or three to four active agents. This study demonstrated a high level of effectiveness of these antiretroviral agents in the setting of routine clinical practice, complimenting recently published clinical trial results on the efficacy of these medications. A recent analysis from the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) ^[13] identified a temporal trend towards increased virologic suppression among patients treated with second-line regimens between 2000 and 2005. The current study extends this work by evaluating more heavily treatment-experienced patients changing regimens in more recent years and examines the impact of several new antiretroviral agents that were not approved during the NA-ACCORD study period. We suggest that the availability of these novel antiretroviral agents since 2006 has ushered in a new era for treatment-experienced HIV-infected patients, whereby achieving an undetectable plasma viral load, as advocated by recent treatment guidelines, is a considerably more attainable goal ^[8,9].

Virologic suppression in the clinical practice setting of this study, remarkably, mirrors or surpasses that observed in randomized clinical trials. Outcomes observed with RAL-containing regimens in our study (DOS) closely approximate those seen from data in the BENCHMRK trial (BENCHMRK, 62%; DOS, 65%) ^[2]. Similarly, in patients on a darunavir/-containing regimen in DOS, the number of patients with a viral load below 50 copies/ml eclipsed that observed at 24 weeks in the POWER-1 and POWER-2 trials (POWER-1, 43–53%; POWER-2, 18–39%; and DOS, 65%) ^[6,7]. Because of fundamental differences in study design and the different response rates observed in the control groups, it is difficult to make direct comparisons between studies. For example, RAL and ETR were used frequently with DRV/r in the current study but were not available during the POWER clinical trials. With more treatment options, providers in the current study were able to construct regimens with more active agents and achieved better virologic suppression in both the DRV/r and comparator populations. Although caution must be exercised when comparing results across trials, the findings from this clinical study indicate that DRV/r and RAL are effective in the routine practice setting for the management of HIV in patients with limited treatment options.

The concept of using at least two active agents in every antiretroviral regimen, as proposed in current HIV treatment guidelines ^[8,9], is supported firmly by the findings in this study. Indeed, the current study suggests that newer agents, owing to their inherent antiretroviral activity in the presence of resistance to previously existing agents, are effective in this setting simply because they are more likely fully active. In our cohort, over 70% of patients treated with regimens, including both DRV/r and RAL had 24-week viral loads below 50 copies/ml. A similar synergy between novel agents was seen at 24 weeks in the DUET clinical trials in which coadministration of DRV/r and ETR showed improved outcomes (ETR/DRV/OBR, 56–62%; placebo/DRV/OBR, 39–42%) and at 48 weeks in the BENCHMRK clinical trials in which coadministration of RAL and DRV, enfuvurtide, or tipranavir (TPV) showed improved outcomes (RAL/DRV/OBR, 69%; DRV/OBR, 47%; RAL/T-20/OBR, 80%; T-20/OBR, 57%; RAL/TPV/OBR, 73%; and TPV/OBR, 40%) ^[14–16]. Findings from both clinical trial and cohort studies support meaningful clinical differences between newer more ‘fully active’ drugs vs. older ‘partially active’ drugs and suggest that treatment regimens for experienced patients should target optimally effective combinations. To help clinicians predict which agents should be used in the next regimen, future studies are needed to further define the precise degree of activity for specific drugs on the basis of virologic susceptibility, achievable plasma and tissue levels of drug, and overall tolerability.

In contrast to the widespread uptake of DRV/r and RAL, ETR (n = 18) and MRV (n = 1) were prescribed less frequently in our study population. This finding may be related to later availability of these agents in our clinic. However, in the case of MRV, it is likely that logistic challenges to utilizing this antiretroviral agent limited its use ^[17–20]. For example, providers ordered the assay to assess for the presence of CCR5 tropic virus very infrequently, perhaps because of concerns about the cost or the availability of the test ^[20]. As time elapses, future studies should evaluate the utilization patterns and effectiveness of ETR, MRV, and other novel antiretroviral agents in clinical practice settings.

The results of this study should be interpreted within the context of its limitations. As a single, academic HIV clinic in the southeastern USA, results may not be generalizable to other locations. As an observational study, unmeasured confounders related to antiretroviral regimen selection may have influenced the findings of the study. Additionally, as with all observational studies, the results identify associations but cannot attribute causality. Data were collected through routine practice patterns and accordingly, a fair number of patients (n = 21) changed regimens without resistance assays being performed. A post-hoc chart review indicated that these patients had been nonadherent, recently stopped antiretroviral agents because of side effects, or were returning to care after a long absence. None of the patients had resistance testing performed prior to changing regimens, as they were not on therapy. Although certainly a limitation of the study, initiating antiretrovirals without the help of resistance testing is one of the realities of routine HIV care not captured in a clinical trial. Missing resistance data were included in all analyses as an unknown number of active agents. Last, although adherence data were not uniformly collected, the data we did collect suggested that some of the therapeutic effectiveness of the newer agents was due to better tolerability of the individual antiretroviral agents.

In summary, the results from this observational study of highly antiretroviral-experienced patients suggest the effectiveness of DRV/r and RAL in routine clinical practice mirror findings from clinical trials. We suggest the use of novel antiretrovirals approved since July 2006 have ushered in a new era for treatment-experienced patients in which virologic suppression is a realistic and attainable therapeutic goal. The availability of several potent agents with improved tolerability profiles increases therapeutic options for highly treatment-experienced patients to construct regimens containing multiple active agents. Future studies should assess the durability of virologic suppression among heavily treatment-experienced patients and the utilization and impact of other novel antiretroviral agents in routine clinical practice settings as more time elapses to allow for increased uptake of these agents.

Acknowledgements

This study was sponsored by Tibotec Therapeutics. Representatives from Tibotec Therapeutics participated in study design, data interpretation, and manuscript preparation but did not contribute to data collection or data analysis. Academic authors made final determination of manuscript content.

The UAB 1917 Clinic Cohort is supported by the UAB CFAR (P30-AI27767), CNICS (1R24-AI067039-1), and the Mary Fisher CARE Fund. We would also like to thank Tommy Liu and the Stanford Genotype Database team for their assistance with the genotype analysis.

We thank the University of Alabama at Birmingham 1917 Clinic HIV/AIDS Clinic Cohort management team for their assistance with this project.

The 1917 Clinic Cohort Team members are listed below:

Steering committee: Michael S. Saag, Michael J. Mugavero, James H. Willig, James L. Raper, Paul Goepfert, Jeroan J. Allison, Mirjam-Colette Kempf, Joseph E. Schumacher, Inmaculada B. Aban.

Faculty investigators: Hui-Yi Lin, Maria Pisu, Linda Moneyham, David Vance, Susan L. Davies, Eta Berner, Edward Acosta, Jennifer King, Richard A. Kaslow, Eric Chamot, Andrew O. Westfall.

Research support team: Karen Savage, Christa Nevin, Frances B. Walton, Malcolm L. Marler, Sarah Lawrence, Barbara Files-Kennedy, D. Scott Batey.

Informatics team: Manoj A. Patil, Ujavala Patil, Mohit Varshney, Eugene Gibson, Suneetha Thogaripally, Alfredo Guzman, Dustin Rinehart, Ridha T. Bagana.

Current trainees: James McKinnell, Paula Seal, Jessica Pullins, David Jackson, Rebecca Wylie, Cynthia Baffi, Noah Godwin

H.-Y.L. had full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.

This study was approved by the University of Alabama at Birmingham IRB.

J.A.McK. and J.H.W. contributed to this work by assisting with conception and design, acquisition of data, analysis of data, and drafting of the manuscript. H.-Y.L. contributed to this work by assisting with conception and design, analysis of data, and revision of the manuscript. The remaining authors provided substantial contributions to the conception and design, data acquisition or interpretation of data, and critically revised the intellectual content of the manuscript. All authors approved the final version of the manuscript.

J.A.McK. has received research funding from the Bristol-Myers Squibb Virology Fellows Research Program for the 2008–2009 academic years. M.S.S. has received grants and research support from Achillion Pharmaceuticals, Boehringer Ingelheim, Gilead, GlaxoSmithKline, Merck, Panacos, Pfizer, Progenics, Roche, Serono, Tibotec, Trimeris, and Vertex. He served as a scientific advisor to Achillon, Avexa, Boehringer Ingelheim, Bristol-Myers Squibb, Gilead, GlaxoSmithKline, Monogram Biosciences, Panacos, Pfizer, Progenics, Roche, Tanox, Tibotec/Virco, Trimeris, and Vertex. J.H.W. has received research funding from the Bristol-Myers Squibb Virology Fellows Research Program for the 2006–2008 academic years. M.J.M. has consulted and received recent research funding from Tibotec Therapeutics and Bristol-Myers Squibb. J.M.M., L.L.DeL. and W.K. were employees of Tibotec Therapeutics during the time of this analysis.