Abstract
Introduction
Multiclass antiretroviral (ARV) medication–resistant HIV infection, medication toxicity, and medication intolerance limit antiretroviral options for many HIV-infected individuals. There is a growing need for effective, novel antiretroviral combinations. One such combination currently being evaluated is a dual-class, dual-drug regimen containing inhibitors of HIV-1 viral protease and integrase. There is some evidence that this combination is effective in the treatment of HIV infection. 1 -6 In addition, a conceptually similar dual-class, dual-drug combination of a protease inhibitor (PI) with a nonnucleoside reverse transcriptase inhibitor (NNRTI) was as effective as standard 3-drug antiretroviral therapy (ART) in antiretroviral-naive HIV-infected individuals in a large randomized controlled trial. 7
Over the last several years at our institution, we have increasingly utilized combination ART with inhibitors of HIV-1 viral protease and integrase. The specific goals of this study were (1) to evaluate the reasons for initiation of this antiretroviral combination, (2) to assess virologic outcomes during therapy, and (3) to assess factors associated with success or failure of combination therapy with inhibitors of HIV-1 viral protease and integrase.
Methods
Denver Health is an integrated public health care system serving approximately 1600 HIV-infected individuals. 8 We retrospectively identified all individuals treated with the integrase strand transfer inhibitor (INSTI) raltegravir prior to September 2010, utilizing administrative pharmacy refill data and an electronic medication reconciliation system. Approximately 90% of Denver Health patients use our pharmacy system for prescription medications. 9 Further record review was undertaken to identify individuals treated with a dual-class regimen containing a PI (with or without ritonavir [RTV] boosting) and the integrase inhibitor raltegravir. Inclusion required confirmed HIV seropositivity and age ≥18. Individuals were excluded if they had less than 60 days of follow-up. This project was part of a larger ART quality assurance initiative at our institution.
Electronic medical records were reviewed to identify regimen initiation and termination dates along with the clinical scenarios influencing the decision to initiate, change, or terminate therapy. Prior antiretroviral exposure history was obtained through detailed review of records, including records from prior HIV care providers when available. Results from all prior genotypic resistance assays were utilized to generate a cumulative baseline genotype. Resistance was determined based on the International AIDS Society USA (IAS) guidelines. 10 Duration of follow-up spanned from the initiation date of the included regimen until loss to follow-up, death, or December 2010, whichever occurred first. Laboratory data including all CD4 counts and HIV-RNA levels were reviewed. HIV-RNA was measured using the Roche CobasAmpliprep/Cobas (TaqMan) HIV-1 assay.
Adherence was determined based on pharmacy refill data and calculated for the 6-month period prior to the last evaluable HIV-RNA level on the included regimen. For regimens lasting less than 6 months, we evaluated adherence for the entire duration of therapy. Individuals who utilized an outside pharmacy at any time during therapy were excluded from adherence analyses. Virologic success was defined by achieving or maintaining a plasma HIV-RNA level ≤200 copies/mL. Virologic failure was defined by an HIV-RNA level >200 copies/mL at regimen termination or at last follow-up.
Basic descriptive statistics were utilized to describe the included population, prior ARV exposure, the presence of ARV resistance, the reasons for initiation of the included regimen, regimen characteristics, and surrogate treatment outcomes including rates of virologic suppression and changes in CD4 counts. In subgroup analyses, comparisons between continuous variables were performed using the Wilcoxon rank sum test and comparisons of categorical variables utilized the chi-square test. Bivariate logistic regression was utilized to examine the relationship between patient-specific variables, including adherence to therapy, with virologic failure. All analyses were performed using SAS statistical software version 9.2 (SAS Institute Inc, Cary, North Carolina).
Results
A total of 183 HIV-infected individuals were prescribed raltegravir as a component of combination ART prior to September 2010, and 40 individuals were prescribed a dual-class, dual-drug regimen containing inhibitors of HIV-1 viral protease and integrase. One individual was excluded because of limited follow-up of less than 1 month at our institution. Table 1 describes the 39 included individuals; 90% were male, 49% were white, 41% were Hispanic, and 85% were men who have sex with other men.
Baseline Characteristics of 39 Individuals Treated with Dual-Class, Dual-Drug Therapy Containing Inhibitors of HIV-1 Viral Protease and Integrase
Abbreviations: IQR, interquartile range; MSM, men who have sex with other men; IDU, injection drug use.
The median duration of prior ART was 11 years (IQR 7-12 years); 27 individuals (69%) were 3-class ARV medication experienced (Table 2). Four individuals used raltegravir in their prior regimen; none had virologic failure on that regimen. Of all, 29 individuals had resistance testing prior to initiating the dual-class regimen; 93% had nucleoside/nucleotide reverse transcriptase inhibitor (NRTI) resistance mutation/mutations, 62% had NNRTI resistance mutation/mutations, and 24% had at least 1 major PI resistance mutation. Dual-class resistance to NRTIs and NNRTIs was documented in 18 (62%) of 29 individuals.
Details of Prior Antiretroviral Use and Resistance in 39 Included Individuals a
Abbreviations: IQR, interquartile range; NRTI, nucleoside/nucleotide reverse transcriptase inhibitor; NNRTI, nonnucleoside reverse transcriptase inhibitor; PI, protease inhibitor; 3-Class = NRTI + NNRTI + PI.
a Of all, 29 individuals had prior resistance testing, whose results are available.
Of all the 39 individuals, 18 (46%) were switched from a suppressive ARV regimen with HIV-RNA ≤200 copies/mL. Of the 18 individuals, 16 switched regimens because of drug toxicity, side effects, or drug–drug interactions. Of these individuals, 8 also had resistance documented in prior genotypic resistance assays. Of the 39 individuals, 21 (54%) had HIV-RNA >200 copies/mL (median 4.3 copies/mL, IQR 3.4-4.9), when the included dual-class regimen was initiated. Of these 21 individuals, 19 initiated dual-class therapy because of documented ARV medication resistance with or without prior medication toxicity or intolerance. The most common complications in the 22 individuals in this study with prior toxicity or intolerance included renal toxicity (n = 9), gastrointestinal side effects (n = 5), hypercholesterolemia (n = 5), and abnormal liver function tests (n = 3).
Raltegravir was used at 400 mg orally twice daily (BID) in all individuals. Protease inhibitors used by the 39 individuals included the following: Darunavir 600 mg + RTV 100 mg BID (n = 17), atazanavir (ATV) 300 mg + RTV 100 mg once daily ([QD] n = 11), lopinavir (LPV) 400 mg + RTV 100 mg BID (n = 3), ATV 400 mg QD (n = 3), fos-amprenavir (FPV) 700 mg + RTV 100 mg BID (n = 2), saquinavir (SQV) 1000 mg + RTV 100 mg BID (n = 2), and darunavir 800 mg + RTV 100 mg QD (n = 1).
Overall, 36 (92%) of 39 individuals achieved an HIV-RNA level ≤200 copies/mL at least once while on treatment with the dual-class regimen (Table 3). After a median 328 days (IQR 190-737 days) of follow-up, 29 (74%) of 39 individuals maintained an HIV-RNA level ≤200 copies/mL; however, only 44% had an HIV-RNA level <50 copies/mL. Median duration of follow-up was the same for individuals with and without successful HIV-RNA suppression at the end of follow-up, that is, 311 and 328 days, respectively. Median CD4 count change was +93 cells/mm3 (IQR-3 to +190 cells/mm3). This value was the same in individuals with and without successful HIV-RNA suppression ≤200 copies/mL at the end of follow-up, that is, +84 and +93 cells, respectively.
Virologic and Immunologic Outcomes during Treatment with a Dual-Class, Dual-Drug Combination of Inhibitors of HIV-1 Viral Protease and Integrase a
Abbreviations: IQR, interquartile range; cp, copies; PI, protease inhibitor; RAM, resistance-associated mutation; ATV, atazanavir; RTV, ritonavir; r, ritonavir.
aCD4 count changes were measured from baseline to virologic failure or the end of follow-up.
b Ever refers to achieving that HIV-RNA level at any point during treatment.
c Last refers to the last recorded HIV-RNA value while on study treatment.
Pharmacy refill data were available for 34 individuals, including all 10 individuals with virologic failure. Median adherence for the 6 months prior to the end of follow-up or virologic failure was 96.4% (IQR 83.3%-100%). Individuals who maintained virologic suppression had a median adherence level of 98.5% (IQR 91.1%-100%), while those experiencing virologic failure had a median adherence level of 82.0% (IQR 66.7%-85.6%; P < .01). Figure 1 shows the proportion of individuals who maintained successful virologic suppression at the end of follow-up by tercile of adherence. In the top tercile, 92% maintained virologic suppression compared with 82% in the middle tercile and 36% in the bottom tercile. In bivariate logistic regression, for every 10% decrease in adherence, the odds of virologic failure increased by 90% (OR 1.9, 95% CI 1.1-3.3; P = .03).
Proportion of individuals with successful viral suppression at the end of follow-up or at the time of regimen termination stratified by tercile of antiretroviral medication adherence in the 6 months prior to HIV-RNA measurement.
Several subgroup analyses were undertaken to evaluate other potential factors associated with successful viral suppression at the end of follow-up (Table 3). No baseline demographic or disease-specific variables were associated with virologic response. There were no significant differences comparing individuals initiating therapy with a baseline HIV-RNA level ≤200 copies/mL versus those with higher viral loads. Similarly, no differences were noted comparing individuals with or without prior PI exposure or specific PI used (ATV/r vs darunavir/r). However, in the subgroup of individuals with prior resistance testing available, none of the 4 individuals with 2 or more major PI mutations at baseline maintained viral suppression at the end of follow-up compared with 21 (84%) of 25 with 1 or no major PI resistance mutations (P < .001). All 4 of these individuals with ≥2 major PI resistance mutations also had poor adherence to therapy, with adherence ranging from 58% to 83%.
In addition to the 10 individuals with virologic failure, 3 individuals discontinued their regimen in the setting of having an HIV-RNA level <200 copies/mL. One individual added tenofovir (TDF) and emtricitabine in the setting of acute hepatitis B virus infection. Another individual changed regimens in the setting of jaundice from medication-associated hyperbilirubinemia. The third individual discontinued treatment and entered hospice for preexisting non-HIV-related comorbidities.
Five individuals had genotypic resistance testing of HIV-1 viral protease and reverse transcriptase performed at the time of virologic failure of their dual-class regimen. There were no new major protease resistance mutations in the 4 individuals who had baseline resistance testing available. In the fifth individual, 2 major protease resistance mutations were present at virologic failure of the dual-class regimen, but this individual had no prior resistance testing at our institution and had 16 years of prior antiretroviral exposure. Integrase resistance testing was not performed for any individuals failing therapy.
Discussion
A total of 39 individuals received a dual-class, dual-drug combination of inhibitors of HIV-1 protease and integrase at our institution. Preexisting ARV medication resistance and prior ARV medication toxicity or intolerance were the most common reasons for initiating the dual-class regimen. Most individuals (92%) achieved at least 1 HIV-RNA level ≤200 copies/mL during follow-up. Over time, 10 (26%) individuals experienced virologic failure with an HIV-RNA level >200 copies/mL. The factor most associated with virologic failure was poor adherence to therapy; the odds of virologic failure increased by 90% for every 10% decline in adherence. Preexisting major PI resistance mutations may also predispose to virologic failure. However, this analysis was complicated by the fact that all 4 individuals with 2 or more major PI resistance mutations at baseline had poor adherence to therapy.
Novel ARV treatment combinations are necessary for HIV-infected individuals with ARV medication resistance, medication toxicity, or intolerance and for individuals with combinations of these factors. One novel combination undergoing evaluation is the dual-class, dual-drug combination of inhibitors of HIV-1 viral protease and integrase. The rationale for this regimen stems from at least 3 observations. First, a conceptually similar dual-class combination of a PI and NNRTI was as effective as standard 3-drug NNRTI-based ART in antiretroviral-naive HIV-infected individuals in a large randomized controlled trial. 7 However, these regimens were generally less tolerable than standard NNRTI- or PI-based ART. Second, in the pivotal phase III clinical trials of raltegravir in ARV treatment-experienced patients, individuals randomized to raltegravir who had a genotypic sensitivity score of 1 in the optimized background therapy (functional dual therapy) had durable HIV suppression with 115 (72%) of 160 individuals having an HIV-RNA level <50 copies/mL at 96 weeks. 11 Third, this combination does not include NRTIs or NNRTIs, which are commonly associated with toxicity and/or intolerance.
We are aware of 6 reports on the use of dual-class, dual-drug therapy with inhibitors of HIV-1 viral protease and integrase in antiretroviral-experienced individuals, which have been published or presented in abstract form. 1 -6 These studies have evaluated 121 HIV-infected individuals. Most individuals (71%) were switching to the dual-class regimen while on an effective regimen as demonstrated by having an undetectable HIV-RNA at the time of the switch. In these 6 studies, 83% to 100% of antiretroviral-experienced individuals achieved and maintained an undetectable HIV-RNA on the dual-class regimen. We are aware of 2 studies of this dual-class strategy in antiretroviral-naive individuals. 12,13 In a randomized controlled trial, raltegravir was compared with coformulated TDF/emtricitabine, each in conjunction with LPV/r. 12 At 48 weeks, virologic efficacy was similar with 83% and 85% of raltegravir and TDF/emtricitabine patients having an undetectable HIV-RNA. 12 In a single-arm study that utilized darunavir/r with raltegravir, 26% of individuals met criteria for virologic failure at 48 weeks. 13 Our study is the largest to date of this dual-class strategy in antiretroviral-experienced patients and is the first to evaluate predictors of success of the dual-class regimen in HIV-treatment-experienced patients.
The first goal of this study was to evaluate the reasons for initiation of this novel combination of ARV medications. Not surprisingly, all included individuals had preexisting ARV medication resistance, prior medication toxicity or intolerance, or a combination of these reasons for initiating their regimen. Drug–drug interactions or potential overlapping toxicity with non-HIV medications were other reasons observed. One individual was previously antiretroviral-naive but had extensive dual-class resistance to NRTIs and NNRTIs. The most common combination of factors leading to regimen selection was combined NRTI (usually deoxycytidine analog) and NNRTI resistance, with or without prior toxicity or intolerance.
The second goal of this study was to evaluate virologic outcomes of therapy. There was excellent initial success with 92% of individuals achieving an HIV-RNA level ≤200 copies/mL. At the end of follow-up, 74% continued to have HIV-RNA levels ≤200 copies/mL; however, only 44% had HIV-RNA levels <50 copies/mL. It is known that raltegravir decreases the plasma level of both ATV and darunavir, the main PIs utilized in our population, although these are not believed to be clinically significant interactions. 14,15 If the durability of this dual-class strategy proves to be limited, this drug interaction is one potential contributing factor.
The third goal of this study was to assess factors associated with poor virologic response to the dual-class, dual-drug regimen. Adherence to therapy was the main predictor of virologic response. Adherence to ART is critical to successful treatment outcomes and may be more important in treatment-experienced populations where regimen potency may be undermined in the setting of measured or unmeasured resistance mutations. 16 In prior studies of this dual-class strategy, most individuals switched therapy while on a suppressive ARV regimen. 1 -4 This approach selects individuals with a higher likelihood of adequate adherence and may account for the high rates of success of this dual-class strategy in the published literature. Against this theory is a single study 5 of this dual-class strategy in which 16 of 18 treatment-experienced individuals initiating therapy with detectable HIV viremia were able to achieve an HIV-RNA level <50 copies/mL. The 2 failures in that study were, however, attributed to poor adherence.
It is concerning that we found a potential association between preexisting major PI-resistance–associated mutations and virologic failure. This is a potential rather than likely association because all of these individuals also had suboptimal adherence. Preexisting major PI-resistance mutations and poor adherence are not independent events. Major protease mutations are much more likely to be present in individuals with suboptimal adherence and failure of prior PI-containing therapy. 16 In the 4 patients who failed therapy with 2 or more preexisting major PI-resistance mutations, all individuals had predicted success with the specific combination of mutations and the PI selected based on the Stanford algorithm. 17 We believe this finding causes lingering concern over the use of this novel combination in individuals with preexisting major PI-resistance mutations. This concern is magnified by the low genetic barrier to resistance of first-generation integrase inhibitors and the high likelihood of developing resistance to these agents during virologic failure. 18
This study has several important limitations. It is retrospective in nature and includes a small population of individuals. Due to these limitations, the findings are hypothesis generating and not intended to represent the definitive evaluation of this novel ARV medication combination. However, it is the largest study to date evaluating the combination of inhibitors of HIV-1 viral protease and integrase in treatment-experienced patients. Furthermore, this study represents a real-world snapshot of the effectiveness of this combination. This project was designed to evaluate practices at Denver Health Medical Center. The results may not be generalizable outside of this institution. Another limitation related to the small sample size and low rate of virologic failure was that it restricted the ability to perform multivariate analysis to assess factors associated with virologic failure. Finally, the duration of follow-up was limited. Further follow-up of this treatment combination will be vital, particularly when considering the proportion of individuals with low-level HIV viremia (50-200 copies/mL) at the end of this study and the uncertainty of the long-term implications of this finding.
In conclusion, there was a high rate of initial virologic success of dual-class, dual-drug antiretroviral regimens containing inhibitors of HIV-1 viral protease and integrase. The major factor associated with virologic failure was suboptimal adherence to therapy. There was also a concerning potential association between cumulative major PI-resistance–associated mutations and an increased risk of virologic failure. In addition, the fact that many individuals had low-level viremia at the end of follow-up raises some concern regarding the durability of success with these regimens. In the right setting, particularly with resistance and/or intolerance to NRTIs and NNRTIs, it appears from the cumulative published data and our study that this novel combination will be effective in the majority of individuals. While this regimen may become a viable treatment option in antiretroviral-naive and antiretroviral-experienced populations in the future, the data presented here suggest cautious use of this regimen until further prospective and long-term follow-up data are available.
Footnotes
This research was presented at the Sixth International Conference on HIV Treatment and Prevention Adherence, May 2011, Miami, FL, USA.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The author(s) received no financial support for the research, authorship, and/or publication of this article.