HIV Articles  
Pattern and impact of emerging resistance mutations in treatment experienced patients failing darunavir-containing regimen
AIDS:Volume 22(14)12 September 2008p 1809-1813
Delaugerre, Constancea; Pavie, Juliettea; Palmer, Pierrea; Ghosn, Jadeb; Blanche, Stephanec; Roudiere, Laurentc; Dominguez, Stephanied; Mortier, Emmanuele; Molina, Jean-Michela; de Truchis, Pierref aSaint Louis Hospital-APHP, Paris, France bKremlin Bicetre Hospital-APHP, Bicetre, France cNecker Hospital-APHP, Paris, France dMondor Hospital-APHP, Creteil, France eLouis Mourier Hospital-APHP, Colombes, France fRaymond Poincare-APHP, Garches, France.
Background: Ritonavir-boosted darunavir (DRV/r) has proven potent efficacy when used in heavily pretreated patients, harboring protease inhibitor-associated resistance mutations. Limited data are available on resistance pattern emerging in patients failing DRV/r and on subsequent remaining protease inhibitor options.
Methods: Analysis of baseline and failure resistance genotypes were performed in patients experiencing virologic failure (>200 copies/ml) after at least 3 months on a DRV/r (600/100 mg twice daily)-containing regimen.
Results: Twenty-five highly protease inhibitor-experienced patients were included. Baseline median human immunodeficiency virus type 1 RNA was 5 log10 copies/ml and number of total-protease inhibitor, major-protease inhibitor and DRV-associated-resistance mutations was 13, 4 and 3, respectively. Median viral replication duration on DRV/r selective pressure was 34 weeks. Emergence of DRV-associated-resistance mutations was observed in 72% (18/25) of patients, at codons L89I/M/V (32%), V32I (28%), V11I (20%), I47V/A (20%), I54L/M (20%), L33F/I (16%) and I50V (16%). A high risk of DRV resistance was observed in patients with 2 and 3 baseline DRV-associated-resistance mutations and in patients with more than 24 weeks of ongoing viral replication. According to 2007 ANRS algorithm, isolates classified as susceptible to ritonavir-boosted tipranavir decreased from baseline to failure from 76 to 60% and susceptible to DRV/r from 32 to 12%.
Conclusion: Emerging mutations observed after DRV/r failure were those already described to impact the DRV efficacy. Our study provided recommendations to firstly, reconsider lowering the cutoff number of DRV mutations to two; secondly, avoid keeping patients on a DRV-failing regimen for more than 24 weeks and thirdly, to examine the efficacy of using tipranavir after DRV failure.
Development of protease inhibitor resistance occurs via the accumulation of primary and secondary mutations that are, respectively located within and outside the enzyme active site [1]. Resistance protease inhibitor pathways lead to decreased susceptibility to a specific protease inhibitor but mostly conferred high level of protease inhibitor cross resistance, and consequently limits their sequential use. In this setting, it is important to collect cross-resistance data in order to avoid jeopardizing therapeutic options.
Darunavir (DRV) is a new protease inhibitor with a strong binding affinity for the human immunodeficiency virus type 1 (HIV-1) protease that demonstrates high efficacy in vitro against wild-type and multidrug-resistant HIV-1 strains [2,3]. The in-vitro selection experiments starting from wild-type HIV-1 showed that DRV has a very high genetic barrier to the development of resistance and necessitates the simultaneous occurrence of multiple mutations before the virus becomes resistant [3]. The potent activity of ritonavir-boosted DRV (DRV/r) was confirmed in patients failing a protease inhibitor-containing regimen in Power 1, 2 and 3 studies [4,5], as in moderately pretreated patients [6] and also in antiretroviral naive patients [7]. In an adjunctive analysis, De Meyer et al. [8] identified a series of 11 mutations, V11I, V32I, L33F, I47V, I50V, I54L/M, G73S, L76V, I84V, L89V, associated with a reduced in-vitro susceptibility to DRV and with a diminished response in the Power 1, 2 and 3 studies. According to the number of baseline DRV-associated resistance mutations, 75% patients with 0-2 DRV mutations attained more than 1 log10 viral load reduction at week 24, 45% with three mutations and less than 30% with four or more [8]. The different cutoff (susceptible, possible resistant, resistant) of genotypic susceptibility of the 2007 ANRS genotypic algorithm for DRV is based on these data.
Actually, only resistance scores (number of DRV-associated resistance mutations) were available for interpretation of DRV susceptibility (2007 ANRS V16, Stanford HIVDB, Rega V7.1.1 available on ), without taking into consideration the weight of mutations that might have more pronounced impact on virologic response [8]. Observational study of resistant genotypic tests performed in protease inhibitor-experienced patients failing DRV/r-containing regimen is an opportunity to give information concerning DRV resistance profile. The objective of this retrospective study was first, to analyze the emergence of new protease inhibitor-resistance mutations in DRV/r-failing patients and, then, to assess remaining protease inhibitor options.
The 25 patients included in the study were highly experienced with a median of five prior protease inhibitor-containing regimens (Table 1). Baseline resistance was extensive and concerned the three classical therapeutic classes, with a median of 13 total protease inhibitor, four major protease inhibitor and three DRV-associated resistance mutations. As showed in Fig. 1a, mutations at codons V32I (9/25), L33F (19/25), A47V (8/25), I54L/M (21/25) and I84V (15/25) were each present at baseline in more than 30% of failing patients. According to 2007 ANRS algorithm, 44 and 24% of patients harbored baseline viruses with possible resistance and resistance to DRV/r, respectively (Fig. 1b).


Fig. 1. Protease inhibitor genotypic resistance. (a) Protease inhibitors mutations-associated resistance. DRV: darunavir-associated resistance mutations according to the 2007 International AIDS Society list and 2007 ANRS algorithm. (b) Protease inhibitors genotypic susceptibility (2007 ANRS algorithm). ATV/r, atazanavir/ritonavir; BL, baseline; DRV/r, darunavir/ritonavir; F, failure; FPV/r, fosamprenavir/ritonavir; IDV, indinavir; LPV/r, lopinavir/ritonavir; NFV, nelfinavir; SQV/r, saquinavir/ritonavir; TPV/r, tipranavir/ritonavir.


In the new regimen, DRV/r was associated with a median of two NRTi and, for 56% (14/25) of patients with 1-3 new drugs (enfuvirtide 11, etravirine 6 and raltegravir 3). Six patients experienced a viral rebound and 19 had a viral load never suppressed on DRV/r regimen. Median (range) duration of viral replication between DRV/r regimen initiation and failure sampling was 34 weeks (12-104). At failure, HIV-1 RNA, number of total protease inhibitor, major protease inhibitor and DRV-associated resistance mutations were 4.2 (2.6-5.6) log10 copies/ml, 15 (12-21), 5 (3-7) and 5 (1-8), respectively. At least one additional DRV-associated resistance mutation (0-5) was observed in 18 out of the 25 (72%) patients. Mutations that developed in more than 15% of patients were the L89I/M/V in 32% (8/25), the V32I in 28% (7/25), the V11I in 20% (5/25), the I47V/A in 20% (5/25), the I54L/M in 20% (5/25), the L33F/I in 16% (4/25) and the I50V in 16% (4/25) (Fig. 1a). Two mutations switched between baseline and failure in two distinct patients, I50L to I50V and I47V to I47A. The I50V and/or the combination V32I and I47V were present at baseline in 28% (7/25) of patients and were present at failure in 48% (12/25) of patients.
Addition of DRV resistance-associated mutations at failure was significantly different according to the number of baseline DRV mutations (P = 0.009, Fisher exact test) and to the duration of viral replication (P = 0.0068, Fisher exact test). In the three patients with no or one baseline DRV mutation, no additional mutation was observed in two patients and one new mutation was observed in the latest. In the 16 patients with two or three baseline DRV mutations, no change was observed in two patients, one additional mutation was observed in four patients and between two and five new mutations were observed in 10 patients. In the six patients with more than four baseline DRV mutations, no change was observed in three patients and one additional mutation was observed in three patients. Among patients with a duration of viral replication 24 weeks or less (median 22, 12-24 weeks) on DRV/r, additional DRV-associated resistance mutations occurred in four out of 10 (40%) compared with 14 out of the 15 (93%) patients with an ongoing viral replication more than 24 weeks (median 52, 28-104 weeks). No viral HIV-1 subtype, or number of baseline total or major protease inhibitor mutations, or level of baseline and failure plasma HIV-1 RNA or use of new active drug in the regimen was associated with DRV resistance but this is probably limited by the size of the study.
According to 2007 ANRS algorithm, baseline viruses were highly protease inhibitor resistant and only second generation protease inhibitor (such as tipranavir and DRV) retained partial antiretroviral activity (Fig. 1b). Isolates classified as susceptible to ritonavir-boosted tipranavir (tipranavir/r) decreased from baseline to failure from 76 to 60% and susceptible to DRV/r from 32 to 12%.
The aim of this study was to characterize the development of resistance in protease inhibitor-experienced patients with virologic failure to DRV/r-containing regimen. These patients had high levels of reduced protease inhibitor susceptibility at baseline and additional DRV resistance was observed in 18 out of the 25 (72%) patients. As described for others protease inhibitor, barrier to resistance development to DRV/r is substantially compromised in patients with preexisting protease mutations from prior virologic failure [7].
Emergence of mutations L89I/M/V, V32I, V11I, I47V, I54L/M, L33F/I and I50V at failure suggests that these mutations are important for conferring high-level DRV resistance. Importantly, all these mutations were already described to impact DRV susceptibility in vitro and in vivo in experienced patients failing DRV/r-containing regimen [8,10] and even in protease inhibitor moderately experienced patients (Titan trial) in which the V32I mutation was the only one mutation selected in 11% of failing patients [11].
Both (fos)amprenavir and DRV are structurally related molecules [12]. In DRV-naive patients, previous (fos)amprenavir treatment was associated with an increased number of DRV mutations, whereas nelfinavir and atazanavir were not [13,14]. We reported previously that (fos)amprenavir-specific pathways, V32I+I47V and/or I50V, counteract dramatically the efficacy of DRV/r-containing regimen [15]. Our present results confirm that these mutations, emerging at failure to DRV/r regimen, may constitute a significant pathway for DRV resistance. Moreover, a switch from an I47V to an I47A mutation, described only to confer a high level of specific lopinavir resistance [16,17], was observed in one patient. Despite DRV has evidence of a high genetic barrier to resistance, DRV shares probably the same resistance mutations with (fos)amprenavir and lopinavir. Studies of phenotypic susceptibility correlation from Power clinical isolates showed that the highest correlation was observed between DRV and (fos)amprenavir followed by DRV and lopinavir [18]. However, overall data from the Power studies showed only a minimal impact of prior (fos)amprenavir use on the response to DRV/r [18]. But among patients with a high level of phenotypic resistance to (fos)amprenavir [fold change (FC) > 11.4] viruses at baseline, response to DRV/r at week 48 was lower (31%) than those with a high level of phenotypic resistance to (fos)amprenavir (FC 11.4 or less) (66%) [19].
In our study, the relatively high number of additional DRV-associated resistance mutations may be related to the prolonged period of viral replication under DRV exposure. In patients with isolates with zero to one baseline DRV mutation, no or one additional mutation was selected suggesting more adherence problems than a high genetic barrier to DRV resistance development. If baseline isolates harbored four or more DRV-associated resistance mutations, the emergence of additional resistance was also uncommon, probably because DRV susceptibility was already highly reduced. In contrast, selection of DRV resistance was most common in patients whose baseline isolates displayed two or three DRV-associated resistance mutations. Importantly, genotypic susceptibility of these isolates was considered susceptible or possible resistant to DRV/r with the 2007 ANRS genotypic interpretation. If DRV baseline susceptibility is already partially compromised, use of a single new active agent may not be sufficient to avoid the accumulation of resistance mutations. These results suggest to probably reconsidering the cutoff number of mutations leading to possible resistant (=3) and susceptible (0-2) interpretation of DRV score. Indeed, in Power studies, the percentage of patients with plasma HIV-1 RNA levels less than 50 copies/ml at week 24 (and no decrease of >1 log10 viral load) was 64, 50 42, 22 and 10% according to number of 0, 1, 2, 3 and at least 4 baseline DRV mutations, respectively [8]. Recently, Lambert-Niclot et al. [10] showed that a small number of associated nucleoside reverse transcriptase inhibitors (NRTi) were associated with a higher risk of DRV resistance mutation selection and in contrast the L76V mutation was found to decrease the risk of selecting another DRV resistance mutation, suggesting the strong impact of this mutation on DRV susceptibility. All these results should be retained to avoid the risk of failure to a DRV/r-containing regimen.
Tipranavir/r-containing regimen seems to be the only one protease inhibitor option with partial susceptibility after DRV failure according to recent data [8,10]. These data confirmed the in-vitro results showing different resistance profiles between DRV and tipranavir [3]. Further analysis should be necessary to study efficacy of subsequent tipranavir/r-containing regimen in these DRV-failing patients.
To summarize, our study provided recommendations to firstly, reconsider lowering the cutoff number of DRV mutations to two; secondly, avoid keeping patients on a DRV-failing regimen for more than 24 weeks and thirdly, to examine the efficacy of using tipranavir after DRV failure.
  iconpaperstack View older Articles   Back to Top