Q151M-Mediated Multinucleoside Resistance: Prevalence, Risk Factors, and Response to Salvage Therapy

Transcription

1 BRIEF REPORT HIV/AIDS Q151M-Mediated Multinucleoside Resistance: Prevalence, Risk Factors, and Response to Salvage Therapy Mauro Zaccarelli, a Carlo Federico Perno, a Federica Forbici, Fabio Soldani, Sandro Bonfigli, Caterina Gori, Maria Paola Trotta, Maria Concetta Bellocchi, Giuseppina Liuzzi, Roberta D Arrigo, Patrizio De Longis, Evangelo Boumis, Rita Bellagamba, Valerio Tozzi, Pasquale Narciso, and Andrea Antinori Istituto Nazionale per le Malattie Infettive Lazzaro Spallanzani, Istituto di Ricovero e Cura a Carattere Scientifico, Rome, Italy Among 470 patients with acquired immune deficiency syndrome and/or human immunodeficiency virus infection (HIV/AIDS) who underwent genotype resistance testing (GRT) after the failure of therapy, 17 (3.6%) harbored the Q151M mutation. The Q151M mutation was associated with younger age, lower CD4 + lymphocyte count, higher HIV RNA level, and treatment with 12 pre-grt regimens. By contrast, the Q151M mutation was inversely associated with lamivudine administration. A full reversion of the Q151M mutation was observed in 5 of 5 patients who underwent treatment interruption after GRT. The reversion was followed by a response to salvage therapy in 4 (80%) of 5 patients. Cross-resistance among nucleoside reverse-transcriptase inhibitors (NRTIs) of HIV replication has been associated with the selection and appearance of a complex of mutations, including A62V/V75I/F77L/F116Y/Q151M. This complex, located around the catalytic site of the enzyme, is commonly referred to as Q151M-mediated multinucleoside resistance (Q151M- MNR). The substitution of glutamine by methionine at the amino acid 151 (Q151M) represents the primary mutation of this complex; this mutation leads to various levels of reversetranscriptase resistance toward all nucleoside analogues currently in clinical use. The first reported in vivo occurrences of Received 16 July 2003; accepted 23 September 2003; electronically published 6 January Financial support: This work was supported by Programma Nazionale AIDS, IstitutoSuperiore de Sanità, Italy, and Ricerca Corrente e Finalizzata degli Istituto di Ricovero e Cura a Carattere Scientifico, Ministero della Salute, Italy. a Mauro Zaccarelli and Carlo Federico Perno contributed to the paper equally. Reprints or correspondence: Dr. Mauro Zaccarelli, Istituto Nazionale per le Malattie Infettive Lazzaro Spallanzani, Via Portuense 292, Rome, Italy (zaccarelli@inmi.it). Clinical Infectious Diseases 2004; 38: by the Infectious Diseases Society of America. All rights reserved /2004/ $15.00 Q151M-MNR were in patients treated with long-term combination therapy with 2 NRTIs who had undergone genotypic resistance testing (GRT) after virological failure [1 3]. Its appearance is, indeed, reported to be associated with long-term antiretroviral treatment; by contrast, it is thought to be nonexistent in NRTI-naive patients [4 6], with the exception of a single patient with a case of primary infection [7]. Several European studies have evaluated the prevalence of the Q151M mutation in patients for whom an NRTI-containing therapy has failed and have reported a prevalence of 1.6% 3.5% [4 6, 8 11], but risk factors related to the appearance of the Q151M mutation have not been defined. The characterization of such risk factors may help in selecting therapeutic approaches that are able to decrease the frequency of its appearance and, therefore, potentially influence the rate of therapeutic failure. For this reason, we analyzed a database of 470 patients undergoing GRT after experiencing failure of HAART [12] who were attending a reference center for HIV treatment in Rome. We also evaluated the dynamics of the appearance and/or disappearance of the Q151M mutation in response to treatment changes. Methods. A database of all consecutive patients who underwent GRT at our health care center (National Institute of Infectious Diseases, Lazzaro Spallanzani ) was assessed. For each patient, a complete clinical history was recorded at the time of GRT, including data on all antiretroviral drugs the patient had ever used and the length of treatment for each drug, CD4 + lymphocyte counts, HIV RNA levels, and disease history. All patients had experienced virological failure after treatment with HAART regimens that included 3 or 4 drugs. A patient was considered to have experienced failure of therapy if they had received 2 consecutive HIV RNA levels of 150 copies/ml while receiving treatment, although GRT was usually performed for patients with an HIV RNA level of 1500 copies/ ml. Direct sequencing of reverse-transcriptase and protease genes was performed using the Viroseq HIV-1 genotyping kit, versions 1 and 2 (PE Biosystems) and the 3100 ABI sequencer (Applied Biosystems). After genotyping was performed, a therapeutic strategy was determined on the basis of GRT results interpreted by expert virologists and clinicians. All treatments were begun after GRT, and responses to treatment in patients carrying the Q151M mutation were recorded. In cases in which therapy was interrupted, the results of GRT performed after the interruption in treatment were compared with the results of GRT performed HIV/AIDS CID 2004:38 (1 February) 433

2 Table 1. Demographic characteristics and laboratory values for 470 patients with HIV/AIDS who underwent genotype resistance testing (GRT) after the failure of therapy. Variable Value Sex Male 325 (69.4) Female 143 (30.6) Age, median years (interquartile range) 38 (35 43) HIV exposure Sexual 248 (52.8) Parenteral a 222 (47.2) Pre-GRT AIDS diagnosis 154 (36.9) No. of previous treatments, (range) 3 (2 5) CD4 + lymphocyte count at GRT, median lymphocytes/mm 3 (interquartile range) 300 ( ) HIV RNA level at GRT, median copies/ml (interquartile range) 30,000 ( ,000) Duration in months of last pre-grt treatment, median (interquartile range) 9.4 ( ) NOTE. Data are no. (%) of patients, unless otherwise indicated. a Includes injection drug use. before the interruption to evaluate the modification of the Q151M complex. Fisher s exact test and the Anova test were used to assess the correlation between the Q151M mutation and other covariates. Factors that showed significant associations with the presence of the Q151M mutation were included in the multivariate analysis (performed using a logistic regression model), together with the duration of treatment with each NRTI. CD4 + lymphocyte counts and HIV RNA levels at the time of GRT were categorized into 2 groups, according to the highest significant associations with the detection of the Q151M mutation found during the univariate analysis. The statistical analysis was performed using SPSS software, version (SPSS). Results. Overall, 470 patients who experienced failure of HAART were analyzed during a 42-month period (June 1999 December 2002). Demographic characteristics and laboratory values for the patients are reported in table 1, and data on treatment with NRTIs before GRT is reported in table 2. The proportion of patients carrying the Q151M mutation was 17 (3.6%) of 470; the prevalence of the other MNR mutations is reported in table 3. The detection of the Q151M mutation increased with the overall duration of NRTI therapy; indeed, no Q151M mutation was found among patients treated with NRTIs for!6 months. Among those treated with NRTIs for 6 12 months, 2.6% had the mutation; among those treated for 1 2 years, 3.7% had the mutation; and among those treated for 2 3 years, 4.1% had the mutation. The prevalence of the Q151M mutation remained steady at 3.5% of patients among those receiving 13 years of NRTI treatment. The results of multivariate analysis are shown in table 4. The presence of the Q151M mutation was significantly associated with lower CD4 + lymphocyte count (significant in patients with a CD4 + lymphocyte count of!150 lymphocytes/ml at the time of GRT), higher HIV RNA level (i.e., 170,000 copies/ml), and receipt of 12 treatment regimens before GRT. In contrast, presence of the Q151M mutation was inversely associated with older age and longer treatment with lamivudine before GRT. Only 3 (1.3%) of 235 patients receiving therapy with lamivudine developed the Q151M mutation, compared with 14 (6%) of 235 patients not receiving lamivudine at the time of GRT (P p.01). No association with duration of treatment was found for any other single NRTI. Moreover, no association was found between the presence of the Q151M mutation and receipt of any NRTI at the time of GRT (data not shown). Single NRTI-associated mutations (other than the Q151M- MNR complex) that were significantly associated with the detection of the Q151M mutation were K65R (OR, 32.1; 95% CI, ; ) and K219E (OR, 14.3; 95% CI, ; P p.007). By contrast, NRTI mutations inversely associated with the detection of the Q151M mutation were M41L (OR, 0.02; 95% CI, ; ), K70R (OR, 0.1; 95% CI, ; ), and M184V (OR, 0.03; 95% CI, ; ). After GRT, the decision was made to interrupt treatment for 5 patients for whom a valid treatment option could not be found. GRT was repeated for these patients after a median of 6 months of interruption (range, 4 11 months). A full reversion of Q151M-MMR was observed in all 5 patients. All 17 patients carrying the Q151M-MMR complex (including those undergoing treatment interruption) were treated with GRT-based therapy. Seven patients (38.9%) achieved undetectable HIV RNA levels (i.e.,!500 copies/ml); 6 patients responded to a 2-class regimen (2 NRTIs plus a protease inhibitor), and 1 patient responded to a 3-class regimen (table 5). It is remarkable that 4 (80%) of 5 patients undergoing treatment interruption and reversion of the Q151M mutation achieved undetectable HIV RNA levels. Only 3 (25.0%) of 12 Table 2. Nucleoside reverse-transcriptase inhibitor (NRTI) treatments for 470 patients with HIV/AIDS before genotype resistance testing. NRTI No. (%) of patients Duration of therapy, median months (interquartile range) Zidovudine 364 (77.5) 24 (10 41) Stavudine 370 (78.7) 23 (14 31) Zalcitabine 81 (17.2) 8 (4.5 14) Didanosine 278 (59.2) 13 (7 23) Abacavir 34 (7.2) 6.5 (3 13) Lamivudine 420 (89.4) 25 (16 35) 434 CID 2004:38 (1 February) HIV/AIDS

3 Table 3. Prevalence of multinucleosideresistance mutations among 470 patients with HIV/AIDS who underwent genotype resistance testing after the failure of therapy. Mutation No. (%) of patients Q151M only 7 (3.6) A62V With Q151M 5 (1.1) Without Q151M 9 (1.9) Total 14 (3.0) V75I With Q151M 6 (1.3) Without Q151M 5 (1.1) Total 11 (2.3) F77L With Q151M 6 (1.3) Without Q151M 2 (0.4) Total 8 (1.7) F116Y With Q151M 12 (2.6) Without Q151M 2 (0.4) Total 14 (3.0) patients who were immediately treated with a salvage regimen without interruption achieved undetectable RNA levels. Discussion. The prevalence found in our group of patients was similar to that found in previous cohorts of Italian patients [10, 11]. It is of interest that, on multivariate analysis, the Q151M mutation was shown to have a relationship with many different factors, suggesting that the emergence of the Q151M mutation is attributable to multiple factors. Thus, the differences in the prevalence of the Q151M mutation found in different studies can be easily related to differences in the sample of patients studied, particularly differences in the duration of treatment the patients received before GRT was performed and in the status of their HIV infection. An overview of the main studies available also shows that the Q151M mutation was found in patients treated with all NRTIs, whether administered alone or in double-combination [2, 3, 9, 10]. All of the most widely used NRTIs, such as zidovudine, stavudine, didanosine, and zalcitabine, were found to be associated with the Q151M mutation in studies involving dual therapies, but very few data are actually available on HAART failure. Our group was the first to analyze a large cohort of patients experiencing HAART failure, and our data showed no evidence of a clear effect of HAART failure on the onset of the Q151M mutation, taking into account both the duration of treatment with each NRTI and those NRTIs received as part of the treatment current at the time of GRT. More recently, it was hypothesized that stavudine and didanosine play a major role in the emergence of the Q151M mutation [13, 14]. Our data showed only univariate association between the Q151M mutation and protracted didanosine treatment, an association that was not confirmed by multivariate analysis, and our data showed no association with exposure to stavudine. Our data also confirm and extend a previous observation showing that treatment with lamivudine could be significantly associated with a lower rate of detection of the Q151M mutation [15]. This result is strengthened by observations of inverse association between the Q151M and M184V mutations. The M184V mutation was already found to be associated with a low incidence of thymidine analogue mutations, consistent with a low phenotypic resistance to zidovudine and stavudine and associated with a low cross-resistance to NRTIs [16, 17]. In a recent article from our group [18], among patients who experienced treatment failure with regimens containing lamivudine, those who harbored the M184V mutation had higher CD4 + lymphocyte counts and lower HIV RNA levels than did those without the M184V mutation. The results from Table 4. Multivariate analysis of factors associated with the Q151M mutation. Variable AOR 95% CI P Age, per year CD4 + lymphocyte count of!150 lymphocytes/mm 3 at GRT HIV RNA level of copies/ml at GRT treatment regimens before GRT Duration of therapy, per year Zidovudine NS Stavudine NS Zalcitabine NS Didanosine NS Abacavir NS Lamivudine NOTE. AOR, adjusted OR; GRT, genotype resistance testing; NS, not significant (i.e., P 1.05). HIV/AIDS CID 2004:38 (1 February) 435

4 Table 5. Characteristics of 7 patients with HIV/AIDS who experienced treatment failure, underwent genotype resistance testing (GRT), and achieved undetectable HIV RNA levels. Patient Pre-GRT regimen Reverse-transcriptase mutations Protease inhibitor mutations Treatment interruption, duration in months Post-GRT regimen Response to post-grt regimen 1 AZT, 3TC, NVP M41L, K103N, Q151M, M184V, K219E D30N, L63P, v77i, N88D None d4t, 3TC, RTV, IDV Yes 2 d4t, ddi, IDV V75I, F77L, F116Y, Q151M L63P, V77I None d4t, ddi, EFV, IDV Yes 3 d4t, ddi, NFV D67N, K70R, F116Y, Q151M, K219Q L10I, M36I, G48V, I54V, V82A 8 3TC, ABC, RTV, IDV Yes 4 d4t, ddi, NFV D67N, K70R, Q151M, K219Q L10I, K20R, M36I, G48V, 4 AZT, 3TC, NFV Yes L63P, V82A, L90M 5 d4t, 3TC, NVP L10I F116Y, Q151M, Y181C, None d4t, 3TC, LPV/r Yes M184V, K219E 6 NVP, RTV, SQV A62V, V75I, F116Y, Q151M, Y181C L10I, L33F, L63P, V77I, I84V, 11 d4t, ddi, RTV, IDV Yes L90M 7 d4t, ddi, RTV, SQV D67N, K70R, F77L, Q151M, Y181C, M184V, G190A, T215F, K219E M36I, M46I, G48V, L90M 4 3TC, ddi, LPV/r Yes NOTE. 3TC, lamivudine; ABC, abacavir; AOR, adjusted OR; AZT, zidovudine; d4t, stavudine; ddc, zalcitabine; ddi, didanosine; EFV, efavirenz; IDV, indinavir; LPV/r, lopinavir/ritonavir; NFV, nelfinavir; NVP, nevirapine; RTV, ritonavir; SQV, saquinavir. the present study extend this observation by showing the protective effect of the M184V mutation versus Q151M-associated multidrug-resistance. As a final point, our data suggest that the presence of the Q151M mutation is not detrimental to future treatment attempts, because 40% of patients with the Q151M mutation responded to salvage therapies involving treatment with 3 or 4 drugs; in this context, the response to treatment of patients harboring the Q151M mutation seems to be strongly related to complete drug interruption. Two conclusions can be drawn from these results: first, that resistance to the Q151M complex is not absolute and can be overcome without resorting to drug therapies involving 14 agents; second, that the 2-base mutation required to achieve the Q151M mutation is not stable, conceivably because of the limited fitness that it induces. Indeed, in our sample, we observed a 10% reversion of the Q151M mutation after treatment interruption, followed by an 80% response to post-grt treatment. Additional studies are also necessary to assess the real utility of treatment interruption in patients infected with drugresistant viruses. This subject is obviously now quite controversial, and this important point must be better defined in the future [19, 20]. In conclusion, our results suggest that the evaluation of the dynamics of the appearance and disappearance of key mutations (such as Q151M) may help in understanding when and how to recycle drugs successfully, particularly in heavily pretreated patients for whom the choice of an effective regimen is limited. Our data also suggest that an effective salvage regimen can be achieved in the presence of the Q151M mutation, even with limited drug options. Acknowledgments The Collaborative Group for Clinical Use of HIV-1 Genotype Resistance Test at the National Institute for Infectious Diseases Lazzaro Spallanzani in Rome, Italy: A.A. (co-chair), Gianfranco Anzidei, Francesco Baldini, R.B., M.C.B., Ada Bertoli, S.B., E.B., Chiara Carvelli, Francesca Ceccherini-Silberstein, Bruno Christian Ciancio, Carmelo D Amato, R.D., P.D.L., Gianpiero D Offizi, F.F., Sara Giannella, Enrico Girardi, C.G., G.L., Patrizia Lorenzini, Patrizia Marconi, P.N., Emanuele Nicastri, Pasquale Noto, C.F.P. (co-chair), Pietro Sette, F.S., M.P.T., V.T., Ubaldo Visco-Comandini, and M.Z. References 1. Ueno T, Shirasaka T, Mitsuya H. Enzymatic characterization of human immunodeficiency virus type 1 reverse transcriptase resistant to multiple 2,3 -dideoxynucleoside 5 -triphosphates. J Biol Chem 1995; 270: Shafer RW, Kozal MJ, Winters MA, et al. Combination therapy with zidovudine and didanosine selects for drug-resistant human immunodeficiency virus type 1 strains with unique patterns of pol gene mutations. J Infect Dis 1994; 169: Shafer RW, Iversen AK, Winters MA, Aguiniga E, Katzenstein DA, Merigan TC. Drug resistance and heterogeneous long-term virologic responses of human immunodeficiency virus type 1-infected subjects to zidovudine and didanosine combination therapy. The AIDS Clinical Trials Group 143 Virology Team. J Infect Dis 1995; 172: Van Vaerenbergh K, Debaisieux L, De Cabooter N, et al. Prevalence of genotypic resistance among antiretroviral drug-naive HIV-1-infected patients in Belgium. Antivir Ther 2001; 6: Gomez-Cano M, Rubio A, Puig T, et al. Prevalence of genotypic resistance to nucleoside analogues in antiretroviral-naive and antiretroviral-experienced HIV-infected patients in Spain. AIDS 1998; 12: Iversen AK, Shafer RW, Wehrly K, et al. Multidrug-resistant human 436 CID 2004:38 (1 February) HIV/AIDS

5 immunodeficiency virus type 1 strains resulting from combination antiretroviral therapy. J Virol 1996; 70: Brenner BG, Routy JP, Petrella M, et al. Persistence and fitness of multidrug-resistant human immunodeficiency virus type 1 acquired in primary infection. J Virol 2002; 76: Schmit JC, Van Laethem K, Ruiz L, et al. Multiple dideoxynucleoside analogue-resistant (MddNR) HIV-1 strains isolated from patients from different European countries. AIDS 1998; 12: Van Vaerenbergh K, Van Laethem K, Albert J, et al. Prevalence and characteristics of multinucleoside-resistant human immunodeficiency virus type 1 among European patients receiving combinations of nucleoside analogues. Antimicrob Agents Chemother 2000; 44: Re MC, Borderi M, Monari P, et al. Prevalence of multiple dideoxynucleoside analogue resistance (MddNR) in a cohort of Italian HIV- 1 seropositive patients extensively treated with antiretroviral drugs. Int J Antimicrob Agents 2001; 18: Balotta C, Violin M, Monno L, et al. Prevalence of multiple dideoxynucleoside analogue resistance (MddNR) in a multicenter cohort of HIV-1-infected Italian patients with virologic failure. J Acquir Immune Defic Syndr 2000; 24: Hirsch MS, Brun-Vezinet F, Clotet B, et al. Antiretroviral drug resistance testing in adults infected with human immunodeficiency virus type 1: 2003 recommendations of an International AIDS Society USA Panel. Clin Infect Dis 2003; 37: Pellegrin I, Izopet J, Reynes J, et al. Emergence of zidovudine and multidrug-resistance mutations in the HIV-1 reverse transcriptase gene in therapy-naive patients receiving stavudine plus didanosine combination therapy. STADI Group. AIDS 1999; 13: Coakley EP, Gillis JM, Hammer SM. Phenotypic and genotypic resistance patterns of HIV-1 isolates derived from individuals treated with didanosine and stavudine. AIDS 2000; 14:F Mouroux M, Descamps D, Izopet J, et al. Low-rate emergence of thymidine analogue mutations and multi-drug resistance mutations in the HIV-1 reverse transcriptase gene in therapy-naive patients receiving stavudine plus lamivudine combination therapy. Antivir Ther 2001; 6: Ait-Khaled M, Stone C, Amphlett G, et al. M184V is associated with a low incidence of thymidine analogue mutations and low phenotypic resistance to zidovudine and stavudine. AIDS 2002; 16: Miller V, Sturmer M, Staszewski S, et al. The M184V mutation in HIV- 1 reverse transcriptase (RT) conferring lamivudine resistance does not result in broad cross-resistance to nucleoside analogue RT inhibitors. AIDS 1998; 12: Zaccarelli M, Perno CF, Forbici F, et al. Using a database of HIV patients undergoing genotypic resistance test after HAART failure to understand the dynamics of M184V mutation. Antivir Ther 2003; 8: Deeks SG, Grant RM, Wrin T, et al. Persistence of drug-resistant HIV- 1 after a structured treatment interruption and its impact on treatment response. AIDS 2003; 17: Walter H, Low P, Harrer T, et al. No evidence for persistence of multidrug-resistant viral strains after a 7-month treatment interruption in an HIV-1 infected individual. J Acquir Immune Defic Syndr 2002; 31: HIV/AIDS CID 2004:38 (1 February) 437