Portal de Periódicos da CAPES

Prevalence and Conditions of Selection of the K65R Mutation in the Reverse Transcriptase Gene of HIV-1

2004; Lippincott Williams & Wilkins; Volume: 38; Issue: 1 Linguagem: Inglês

10.1097/00126334-200501010-00019

1944-7884

Michel Segondy, Brigitte Montès,

HIV/AIDS Research and Interventions

To the Editor: It has been reported that the K65R mutation in the reverse transcriptase (RT) gene of HIV-1 can be selected through in vitro selection experiments by zalcitabine, abacavir, and tenofovir.1-3 This mutation is also known to confer reduced susceptibility to zalcitabine, didanosine, lamivudine, abacavir, and tenofovir.2-4 Therefore, this mutation could be considered as a major resistance mutation to nucleoside and nucleotide RT inhibitors (NRTIs). Moreover, it has been reported that the K65R mutation in combination with the M184V mutation is involved in failure of antiretroviral therapies based on the combination of 3 NRTIs including tenofovir and abacavir or didanosine.5,6 Although the K65R mutation remains relatively uncommon, an increase in its prevalence over the past few years has been observed.7 This trend probably results from the increased use of drugs that select for this mutation. We searched our HIV-1 drug resistance database for the frequency of the K65R mutation and analyzed NRTI treatments and other RT mutations associated with its selection. A total of 1898 HIV-1 RT sequences were obtained from 1274 patients between March 1999 and March 2004 with 1786 patient-years of follow-up. During the course of the study, 34 patients (2.67%) developed a K65R mutation. The overall prevalence of this mutation was 2.18 cases per 100 patient-years: 0 (0/199 patients) between March 1999 and March 2000, 1.45% (6/414 patients) between March 2000 and March 2001, 1.79% (8/448 patients) between March 2001 and March 2002, 1.32% (5/380 patients) between March 2002 and March 2003, and 5.80% (20/345 patients) between March 2003 and March 2004 (P < 0.0001). It is noteworthy that tenofovir became available to our patients in March 2001; the other NRTIs were available before 1999. The treatment history was documented for 33 of the 34 patients harboring a K65R mutation. Each of these 33 patients was matched with 2 patients harboring a wild-type codon at position 65. Matching criteria were the number of treatment regimens and the number of NRTIs received since the beginning of therapy. As shown in Table 1, the 2 groups of patients were similar in terms of clinical, immunologic, and virologic characteristics. HIV-1 subtype diversity was not taken into account for matching. Indeed, in a previous study, no difference was observed in the K65R mutation prevalence between B or non-B HIV-1 variants.8 In the present study, non-B HIV-1 variants were identified in 4 patients harboring the K65R mutation (2 CRF02_AG, 1 subtype A1, and 1 subtype C) and in 3 matched patients (1 CRF01_AE, 1 CRF06_cpx, and 1 subtype D).TABLE 1: Association of the K65R Mutation With Patient Characteristics and With the Other RT MutationsBy univariate analysis, tenofovir (P < 0.0001) and the combination abacavir/didanosine (P = 0.01) in the failing regimen were associated with the K65R mutation. By multivariate analysis, the K65R mutation was independently associated with tenofovir (P < 0.0001) and didanosine (P = 0.045) in the failing regimen. A negative association between the K65R mutation and stavudine in the failing regimen was also observed by univariate (P = 0.007) or multivariate (P = 0.03) analysis. Of the 33 patients harboring the K65R mutation, 16 (48%) were tenofovir treatment naive. For these tenofovir treatment-naive patients, the NRTI combinations zidovudine/didanosine (P = 0.001) and abacavir/didanosine (P = 0.02) were significantly associated with the K65R mutation by multivariate analysis. As shown in Table 1, the K65R mutation was significantly associated with the A62V, S68G, V75I, F77L, Y115F, F116Y, and Q151M RT mutations, whereas a strong negative association was observed between the K65R mutation and the M41L, E44D/A, D67N, V118I, L210W, T215Y/F, and K219Q/E RT mutations. The Q151M mutation is a multidrug resistance mutation nearly always associated with the RT mutations A62V, V75I, F77L, and F116Y. This Q151M-associated mutational pattern has been mainly observed in patients who received a zidovudine/didanosine combination regimen.9,10 The S68G and Y115F mutations can be associated with this mutational pattern,9-11 and the K65R mutation has also been found in patients who developed the Q151M-associated mutational pattern.9,10 In our database, the K65R mutation was found in 33% of the RT sequences harboring the Q151M mutation. The K65R mutation was associated with the Q151M mutation in 12 (75%) of 16 tenofovir treatment-naive patients and only 1 (6%) of 17 tenofovir treatment-experienced patients (P < 0.0001). We observed a strong association between the mutations K65R and S68G; both mutations were found in combination in 11 (69%) of 16 tenofovir treatment-naive patients and 8 (47%) of 17 tenofovir treatment-experienced patients (P = 0.21). The S68G mutation is not nowadays considered as a resistance mutation. However, its high prevalence among patients harboring the K65R mutation suggests that this mutation could play a role in either the resistance of HIV-1 to NRTIs or the fitness of the K65R-mutated HIV-1 strains. It can be noted that the overall prevalence of the S68G mutation in our database was 7.3%. Our results are consistent with those of previous studies showing a close association between the K65R mutation and the Q151M-associated mutational pattern12 and between the K65R mutation and the S68G mutation,13 as well as a mutual exclusion between the K65R mutation and the thymidine analog mutations M41L, D67N, L210W, T215Y/F, and K219Q/E.7 The results of this study suggest that the K65R mutation is essentially selected by tenofovir-containing regimens but also by didanosine-containing regimens, in particular by the zidovudine/didanosine and abacavir/didanosine combinations. This is in agreement with the results reported by Winston et al14 showing that the K65R mutation is associated with the dual use of tenofovir and didanosine and the triple combination of tenofovir, didanosine, and abacavir. Our results also suggest that in the tenofovir treatment-naive patients, the K65R mutation and the Q151M-associated mutational pattern are selected within the same resistance pathway, mainly by didanosine-containing regimens. Michel Segondy Brigitte Montes Department of Virology; Montpellier University Hospital; Montpellier, France