Darunavir (DRV) is the latest licensed protease inhibitor (PI) with in vitro activity against both wild-type and PI-resistant HIV-1 isolates, exhibiting excellent clinical efficacy in patients in whom multiple PI-containing regimens have failed.1 The efficacy of ritonavir-boosted DRV (DRV/r) in combination with an optimized background therapy has been demonstrated in highly experienced patients with extensive drug resistance and limited treatment options2,3 and in treatment-naive patients.4,5
Treatment failure of a DRV/r-containing regimen may occur in the presence of high number of resistance mutations, nevertheless, complete DRV/r resistance is uncommon.6 Specific resistance mutations changes are known to show the highest impact on DRV sensitivity (I50V, I54L, L76V, I84V), whereas other mutations, also affecting DRV sensitivity, seem to exert a lower effect (V11L, V32I, L33F, I47V, I54M, T74P, I89V). All these 11 PI codon substitutions are considered associated with DRV/r resistance by International AIDS Society–USA HIV-1 drug resistance panel, 6 of them being labeled as major mutations (I47V, I50V, I54M/L, L76V, I84V). Detection of 3 or more of these substitutions is expected to cause reduced susceptibility to DRV.7
A score for the interpretation of resistance mutations to DRV was derived from the pooled analysis of the POWER 1, 2, and 3 trials, performed in patients with extensive treatment and PI resistance.8 Other predictive algorithms were subsequently derived from observational data by different research groups.9–11 As resistance to DRV/r is associated with multiple genotypic profiles, the interpretation of genotypic resistance to DRV/r remains challenging, lacking a full consensus on which mutations at baseline are better correlated with treatment failure. To further evaluate the ability of the state-of-the-art algorithms to predict DRV/r failure in an independent validation set, pretreated patients histories were retrieved from the Italian Antiretroviral Resistance Cohort Analysis (ARCA) database and analyzed with the major HIV genotype interpretation systems and the International AIDS Society--USA reference mutation list.
The study concerns a retrospective longitudinal analysis for risk of failure from a large observational database.
The ARCA database, containing demographic, clinical, and virological data from patients who undergo genotype resistance testing for any reason throughout most of Italy, was used to perform the retrospective analysis. Patients initiating any DRV/r-containing regimen between 2005 and 2010 after treatment failure were selected.
The protease mutational pattern before DRV/r was analyzed taking into account both specific DRV/r mutations and major protease mutations, after International AIDS Society–USA criteria.12 Baseline protease mutation patterns, within the resistance test closest to the start of the DRV-containing therapy and performed during the previous failing regimen, were compared between patients who subsequently failed versus patients who did not fail.
Viral rebound (2 consecutive >50 copies/mL HIV-RNA or 1 >50 copies/mL before DRV/r interruption, after initial suppression), or detectable viral load at 24 weeks after DRV/r start, without suppression were considered treatment failure.
The interpretation algorithms HIV Drug Resistance Database (HIVdb) (version 6.0.11), Rega (version 8.0.2), and ANRS (version 2010.07), available on line at the Stanford University website,13 were used to assess the level of DRV resistance and compared with each other as predictors of failure. Moreover, the impact of single mutations associated with DRV/r resistance was also analyzed.
The statistical analysis was performed using kappa statistic, Kaplan-Meier survival analysis and Cox proportional hazard model, adjusted by variables of interest, that were included as categorized if fixed values (ie, 0/1—yes/no; 0/1/2—degree of resistance for interpretation algorithms) and as continuous if discrete numeric.
Overall, 1104 patients starting any DRV/r-contanining regimen made of 2 to 5 drugs were included. Treatment failure was observed in 118 patients at a median time of 11 months [interquartile range (IQR): 5–20]. Among the failing patients, 97 (82.2%) had a genotype resistance testing also at failure as follows: V32I and I54L mutations were the only 2 mutations that significantly increased (P < 0.05) at failure.
The main patients characteristics were as follows: males 72.9%, median age 45 (IQR: 41–50), intravenous drug users 35.0%; patients were highly pretreated because their median number of previous regimens was 11 (IQR: 6–16) and the median number of previous PI regimens was 7 (IQR: 3–11). Moreover, median CD4+ count at the baseline was 246/mmc (IQR: 113–412); median CD4+ nadir was 62 (19–166); and median log10 viral load 3.9 (2.6–4.8).
The median number of protease mutations at baseline was 3 (IQR: 0–4), but it was higher in patients who subsequently failed DRV (4, IQR: 2–6) than in patients who responded to DRV (3, IQR: 2–4). Taking into account only the 11 DRV associated mutations, the difference at baseline in median number of mutations was higher in patients who subsequently failed DRV (2, IQR: 0–3) than in patients who responded (0, IQR: 0–2).
At survival analysis, the proportion of 1-year and 3-year failure was 10.0% and 24.6%, respectively. At baseline, the interpretation systems detected different proportions of DRV full resistance (Table 1), in particular, HIVdb identified a far lower number of patients with DRV full resistance compared with Rega and ANRS.
The kappa statistic for failure showed the following results: HIVdb versus Rega 0.287, HIVdb versus ARNS 0.435, Rega versus ARNS 0.335, indicating a general low agreement between the algorithms.
Despite the lower number of full resistance detected, HIVdb results showed a gradient of risk with a progressive prediction of virological failure risk at adjusted Cox from sensitive to intermediate. ANRS also showed progressively increased risk of failure though nonsignificant, possibly due to lack of power in the intermediate group. In contrast, Rega did not show increased risk of failure for intermediate resistance but only for full resistance (Table 2).
HIV subtype (non-B—86 patients, 7.8%—vs. B—1018 patients, 92.2%) was tested in unadjusted analysis and was not found associated with risk of failure (HR: 0.9, 95% CI: 0.4 to 2.0), so that it was not added to adjusted analysis.
When analyzing the single PI mutations, 4 among them, all included in the 11 DRV-associated resistance mutations, were predictive of failure at unadjusted Cox model as follows: V32I (HR: 2.9; 95% CI: 1.9 to 4.6), I50V (HR: 2.3; 95% CI: 1.1 to 5.4), L76V (HR: 2.2; 95% CI: 1.2 to 4.1), I84V (HR: 1.8; 95% CI: 1.3 to 2.7). Association with failure was neither observed for the other 7 DRV-associated mutations (V11L, L33F, I47V, I54L/M, T74P, I89V) nor for the 14 major PI mutation not considered related to DRV resistance (D30N, M46I/L, I47A, G48V, I50L, Q58E, V82A/F/L/S/T, N88S, L90M).
At multivariate Cox analysis, the HR progressively increased by detecting any 1, 2 or 3 of the 4 mutations (Table 2). No patient harbored all 4 mutations.
All the models presented in the tables were adjusted by 4 variables that were found significantly associated with virological failure at unadjusted analysis including: baseline CD4+ cell count; baseline viral load; number of previous treatment regimens; and CD4+ cell nadir. None of them remained significantly associated with response at adjusted analysis.
In our observational data set of heavily experienced patients treated with DRV/r-containing regimens, DRV/r exhibited a low rate of failure at short-medium time of observation. A limitation of the analysis may be related to the absence of adherence data because they are not routinely collected in the ARCA database. However, the consistency of the results makes our findings of interest in clinical practice.
Not surprisingly, DRV resistance was the only factor associated with virological failure, implying that DRV/r is usually the most potent drug in the regimen. Although the 3 most widely used HIV genotype interpretation systems have been often shown to be comparable, in our large DRV/r observational cohort, HIVdb provided the most consistent association between predicted degree of DRV resistance and failure even though it identified a small number of full DRV-resistant patients. This finding suggests that HIVdb may have a low-negative predictive value (ie, less patients at risk of failure identified, falling in the category of intermediate resistance). On the other hand, ARNS algorithm also predicts increasing risk of failure associated with degree of resistance although identifies the lowest number of intermediate resistant patients and the Rega algorithm allows to identify a higher number of patients at risk of failure, without losing statistical significance, altough do not detect increased risk of failure associated with intermediate resistance to DRV.
This information can be useful when accessing the different systems to build a DRV/r-based salvage regimen.
The virological response to DRV/r was not associated with HIV subtype in our study. This finding is consistent with the observation recently reported in the Artemis phase III trial.10
The analysis of the role of single mutations confirmed the predictive value of 4 substitutions included in the DRV resistance reference list (V32I, I50V, L76V, I84V) and, differently from other reports, did not detect any other relevant mutation. Of note, 3 of the 4 DRV resistance mutations scored significantly (I50V, L76V, I84V) are included in the 2 tops of the 4-level categories of the weighted score, whereas none of the lowest category (V11I, I54L, T74P, L89V) was predictive of DRV/r failure. Thus, inferring the activity of DRV based on the reference resistance mutation list also does not seem to be straightforward.
1. De Meyer S, Azijn H, Surleraux D, et al.. TMC114, a novel human immunodeficiency virus type 1 protease inhibitor active against protease inhibitor-resistant viruses, including a broad range of clinical isolates. Antimicrob Agents Chemother. 2005;49:2314–2321.
2. Clotet B, Bellos N, Molina JM, et al.. Efficacy and safety of DRV-rtv at week 48 in treatment-experienced patients with HIV-1 infection in POWER 1 and 2: a pooled subgroup analysis of data from two randomized trials. Lancet. 2007;369:1169–1178.
3. Madruga JV, Berger D, McMurchie M, et al.. Efficacy and safety of DRV-rtv compared with that of lopinavir-rtv at 48 weeks in treatment-experienced, HIV-infected patients in TITAN: a randomized controlled phase III trial. Lancet. 2007;370:49–58.
4. Ortiz R, Dejesus E, Khanlou H, et al.. Efficacy and safety of once-daily DRV/rtv versus lopinavir/rtv in treatment-naive HIV-1-infected patients at week 48. AIDS. 2008;22:1389–1397.
5. Mills AM, Nelson M, Jayaweera D, et al.. Once-daily DRV/rtv vs. lopinavir/rtv in treatment-naive, HIV-1-infected patients: 96-week analysis. AIDS. 2009;23:1679–1688.
6. De Meyer S, Dierynck I, Lathouwers E, et al.. Identification of mutations predictive of a diminished response to darunavir/ritonavir: analysis of data from treatment-experienced patients in POWER 1, 2, 3 and DUET-1 and 2 [abstract 54]. Presented at: 6th European Resistance Workshop; March 26-28, 2008; Budapest, Hungary.
7. Johnson VA, Brun-Vézinet F, Clotet B, et al.. Update of the drug resistance mutations in HIV-1. Top HIV Med. 2008;16:138–145.
8. De Meyer S, Vangeneugden T, van Baelen B, et al.. Resistance profile of DRV: combined 24-week results from the POWER trials. AIDS Res Hum Retroviruses. 2008;24:379–388.
9. Pellegrin I, Wittkop L, Joubert LM, et al.. Virological response to DRV/rtv-based regimens in antiretroviral-experienced patients (PREDIZISTA study). Antivir Ther. 2008;13:271–279.
10. Descamps D, Lambert-Niclot S, Marcelin AG, et al.. Mutations associated with virological response to DRV/rtv in HIV-1-infected protease inhibitor-experienced patients. J Antimicrob Chemother. 2009;63:585–592.
11. Delaugerre C, Pavie J, Palmer P, et al.. Pattern and impact of emerging resistance mutations in treatment experienced patients failing DRV-containing regimen. AIDS. 2008;22:1809–1813.
12. Johnson VA, Brun-Vézinet F, Clotet B, et al.. Update of the drug resistance mutations in HIV-1. Top HIV Med. 2010;18:156–163.
13. Stanford University HIV Drug Resistance Database, Version 6.1.0. 2011. Available at: http://hivdb.stanford.edu
. Accessed November 14, 2011.