AAC Accepts, published online ahead of print on 14 July 2008 Antimicrob. Agents Chemother. doi:10.1128/AAC.00133-08 Copyright © 2008, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.

Tipranavir/ritonavir genotypic resistance score in protease inhibitor-experienced patients Running title: tipranavir genotypic resistance Anne-Genevieve Marcelin,1* Bernard Masquelier2, Diane Descamps3, Jacques Izopet4, Charlotte Charpentier5, Chakib Alloui6, Magali Bouvier-Alias7, Anne Signori-Schmuck8, Brigitte Montes9, MarieLaure Chaix10, Corinne Amiel11, Georges Dos Santos12, Annick Ruffault13, Francis Barin14, Gilles Peytavin15, Marc Lavignon16, Philippe Flandre17, Vincent Calvez1

D E

1 Laboratoire de Virologie, AP-HP, Groupe Hospitalier Pitié-Salpêtrière, F 75013, Paris, France ; UPMC Univ Paris 06, EA2387, F 75005, Paris, France 2 Laboratoire de Virologie, CHU de Bordeaux, France ; EA 2968, Université Victor Segalen, Bordeaux, France

T P

3 Laboratoire de Virologie, AP-HP, Groupe Hospitalier Bichat-Claude Bernard, Paris, F 75018 ; Université Denis Diderot-Paris 7 ; INSERM U552, Paris, France 4 Laboratoire de Virologie, CHU de Toulouse, France

E C

5 Laboratoire de Virologie, HEGP, Paris, AP-HP, France

6 Laboratoire de Virologie, CHU Avicenne, AP-HP, France

7 Laboratoire de Virologie, CHU Henri Mondor, AP-HP Créteil, France 8 Laboratoire de Virologie, CHU de Grenoble, France

9 Laboratoire de Virologie, CHU de Montpellier, France

C A

10 Laboratoire de Virologie, Hôpital Necker, AP-HP, Paris, France 11 Laboratoire de Virologie, Hôpital Tenon, AP-HP, Paris, France 12 Laboratoire de Virologie, CHU de Fort de France

13 Laboratoire de Virologie, CHU de Rennes, France 14 Laboratoire de Virologie, CHU de Tours, France 15 Pharmacie, AP-HP, Groupe Hospitalier Bichat-Claude Bernard, Paris, F 75018, France 16 Boehringer-Ingelheim, France

17 INSERM U720 ; AP-HP, Groupe Hospitalier Pitié-Salpêtrière, F 75013, Paris, France ; UPMC Univ Paris 06, F 75005, Paris, France

* Corresponding author: AG Marcelin, PharmD, Ph.D. Mailing address: Department of Virology, Pitié-Salpêtrière Hospital, 83 Boulevard de l'Hôpital, 75013 Paris, France Phone: 33142177401, Fax: 33142177411 e mail: [email protected] th

This work was presented at the 14 Conference on Retroviruses and Opportunistic Infections, February 2007, Los Angeles, USA (abstract 612) and at the XVI International HIV Drug Resistance Workshop, June 12-16, 2007, Barbados, Barbados (abstract 74).

Key

words:

tipranavir,

protease,

mutations,

resistance,

genotype.

1

ABSTRACT To identify mutations associated with the virological response (VR) to a tipranavir/ritonavir (TPV/r)-based regimen, 143 patients previously treated with protease inhibitor (PI) were studied. VR was defined by a decrease of at least 1 log10 or undetectable HIV-RNA at month 3. The effect of each mutation in the protease, considering all variants at a codon as a single variable, on the VR to TPV/r was investigated. Mutations at 6 codons were associated with a lower VR

D E

(E35D/G/K/N, M36I/L/V, Q58E, Q61D/E/G/H/N/R, H69I/K/N/Q/R/Y, L89I/M/R/T/V) and one

mutation with a higher VR (F53L/W/Y). The following genotypic score: M36I/L/V – F53L/W/Y + Q58E + H69I/K/N/Q/R/Y + L89I/M/R/T/V, was selected as providing a strong association with

T P

VR. In the 7 patients with a genotypic score of -1 (viruses with only mutation at codon 53), the percentage of responders was 100% and it was 79%, 56%, 33%, 21% and 0% in those with a

score of 0, 1, 2, 3 and 4, respectively. The percentage of patients showing a response to TPV/r

E C

was lower in patients infected with non-clade B (n =16, all non-B subtypes considered together) than clade B (n = 127) viruses (25% and 59%, respectively; p=0.015). Most mutations associated with VR to TPV/r had not previously been associated with PI resistance. This is consistent with phenotypic analysis showing that TPV has a unique resistance profile. Mutations

C A

at 5 positions (35, 36, 61, 69 and 89) were observed significantly more frequently in patients with non-B than B subtype, probably explaining the lower VR observed in these patients.

2

Introduction Tipranavir (TPV) is a recently approved nonpeptidic protease inhibitor (PI) with antiviral activity against multi-PI resistant clinical HIV-1 isolates. Its average 50% effective concentration (EC50) value for these isolates is 240 nmol/l (range 50 to 380 nmol/l) (9, 13, 14). TPV-resistant viruses were selected in vitro in a previous study, from serial passages of wild-type HIV-1NL4-3 in the presence of increasing concentrations of TPV in cell culture. HIV-1 variants with 70-fold decreased susceptibility to TPV were selected after 9 months in passage. Ten mutations were

D E

identified, arising in the following order: L33F, I84V, K45I, I13V, V32I, V82L, M36I, A71V, L10F, and I54V (4).

The efficacy of ritonavir-boosted tipranavir (TPV/r) was examined in HIV-infected patients in two

T P

Phase III trials. These patients were highly treatment-experienced and displayed stronger virological and immunological responses to TPV/r than to other ritonavir-boosted PI (2, 7). Previous analyses of phase II and III clinical trials with TPV/r in PI-experienced patients were conducted to determine the association of protease mutations with reduced susceptibility and

E C

virological response to TPV (1). A TPV mutation score was generated from these analyses, incorporating a set of 16 protease amino-acid positions and 21 mutations (10V, 13V, 20M/R/V, 33F, 35G, 36I, 43T, 46L, 47V, 54A/M/V, 58E, 69K, 74P, 82L/T, 83D and 84V). HIV-1 isolates

C A

with a greater number of these TPV resistance-associated mutations had reduced phenotypic susceptibility and virological response to TPV. Parkin et al. proposed revisions to the TPV mutation score based on the analysis of 1411 clinical samples from the Monogram database. They added new mutations and weighted other specific mutations that were associated with lower than expected (10I, 11L, 32I, 36L, 46I, 47V, 54A, 55R, 60E, 71L, 73T, 82T, 84V, 89V, and 90M) or higher than expected (10F/V, 13V, 20R, 24I, 30N, 36I, 46L, 50L/V, 54L, 76V, 82I, and 88D) susceptibility to TPV (N. T. Parkin, and C. Chappey. Abstr. 13th Conf. Retrovir. Opportunistic Infect., abstr. 637, 2006). The US Food and Drug Administration determined in Boehringer Ingelheim studies that at least 5 of 8 mutations present at baseline (I13, V32, M36, I47, Q58, D60, V82, and I84) were associated with a poorer virological response rate at week 24 and that the most common substitutions emerging in patients with virological failure were L33V/I/F, V82T, and I84V (12). These studies showed that resistance to tipranavir is complex, involving mutations that have not previously been associated with resistance to other PIs. The aim of this study was to determine, using a simple and previously described methods, mutations associated with virological response to TPV/r in a population of PI-experienced patients.

3

Patients and Methods Patients and antiretroviral regimens One hundred and forty-three PI-experienced patients were recruited to the study. All patients were treated with RTV (200 mg bid) plus TPV (500 mg bid) with a background regimen comprising nucleoside reverse transcriptase inhibitors (NRTI) and/or non nucleoside reverse transcriptase inhibitors (NNRTI) and/or enfuvirtide (ENF). TPV and RTV were the only PIs used

D E

in the antiretroviral combinations. Sociodemographic data, clinical data and treatment histories

were collected for all patients recruited. Inclusion criteria and all data were checked by the study monitor. The main characteristics of the study population are shown in Table 1. At baseline, the

T P

median number (interquartile range, IQR) of major and minor PI resistance mutations using the International AIDS Society-USA (IAS-USA) panel list was 3 (0-7) and 9 (1-14), respectively (8).

Participating laboratories belonged to the Agence Nationale de Recherches sur le SIDA (ANRS)

AC11 network and participated in the ANRS quality control assessment of HIV-1 drug

E C

resistance sequencing (3).

Genotypic resistance testing

C A

Sequences of the protease and reverse transcriptase (RT) genes were determined at baseline in each laboratory using the ANRS consensus technique (http://www.hivfrenchresistance.org), the Bayer TrueGene kit, the Abbott ViroSeq kit or an in-house method. All protease and RT gene mutations were identified from the International AIDS Society-USA resistance testing panel (September 2006) (8). The lower limit of quantification (LOQ) was 200 copies/mL or 50 copies/mL according to the virological center.

TPV plasma concentration measurements TPV plasma concentrations were measured by a specific and validated high-performance liquid chromatographic assay (HPLC) coupled to ultraviolet detection at 240 nm as previously described (G. Peytavin, A. G. Marcelin, A. Rouault, M. Bonmarchand, H. Ait-Mohand, B. Cassard, A. Simon, L. Schneider, D. Costagliola, V. Calvez, and C. Katlama. Abstr. 13th Conf. Retrovir. Opportunistic Infect., abstr. 591, 2006). All patients had detectable TPV plasma concentrations (> 100 ng/ml) and were retained for the determination of the resistance score.

4

Statistical methods The end-point for the analysis was the percentage of responders at month 3. Virological response (VR) was defined by a decrease of at least 1 log10 of HIV-RNA from baseline or HIVRNA < the LOQ at month 3. These criteria were also used in the RESIST studies and the FDA TPV resistance analysis (7, 12). Potential associations between each protease mutation (codons 1 to 99) and VR were determined using Fisher’s exact test. Multiple mutations arising at a given position were grouped together as a single variable. Mutations present in at least 10%

D E

of patients giving a p-value lower than 0.10 in the above-mentioned univariate analysis were

retained. They were then analyzed using the removing procedure with a non-parametric test to

select the combination of mutations most strongly associated with VR (5, 10, 11, 15). Cochran-

T P

Armitage (CA) test was used. The removing procedure begins with all k mutations retained from

the univariate analysis. The first step is to calculate the p value with the CA test to give a score

incorporating these initial k mutations. All combinations of k-1 mutations are investigated oneby-one: combination providing the lower p value with the CA test is retained if lower than the p-

E C

value obtained with the k mutations. In the second step, mutations are again removed one-byone to compare the combinations of k-2 mutations, the combination providing the lower p value, and lower than the p value obtained with k-1 mutations, is again retained, and so on. The procedure ends when removing a mutation does not provide a lower p-value than the previous

C A

one. The removing procedure was selected because scores obtained using this technique tends to be more strongly associated to the VR than those obtained using the adding procedure (7). A series of univariate logistic regression was fitted to the data to retain, in the multivariate analysis, variables associated with the VR (p10) were responders (p=0.0002). At last, a modified version of the score developed by Boehringer Ingelheim was proposed, based on 22 mutations (I47V, I54A, V82T, I84V, L10V, I13V, K20R, M46L, V11L, V32I, A71L, G73T, L89V, L10I, M46I, L90M, L24I, D30N, I50L/V, I54L, L76V, V82I) (N. T. Parkin, and C. Chappey. Abstr. 13th Conf. Retrovir. Opportunistic Infect., abstr. 637, 2006). Some mutations increased the resistance score (by 1, 2 or 0.5), whereas other mutations decreased the resistance score (reducing the score by 1 or 0.5). In our sample, 17/25 (68%) patients with a score ≤2, 45/73 (62%) patients with a score >2 and ≤5, 17/41 (42%) patients with a score >5 and ≤8, and 1/5 (20%) of patients with a score >8 were responders (p=0.003). In univariate analysis, mutations at 6 codons (all variants of each codon considered as a single variable) were associated with a lower VR to TPV/r (p