| |
High rate of virological failure with didanosine and tenofovir in treatment-experienced patients
|
| |
| |
RESEARCH LETTER
AIDS: Volume 19(15) 14 October 2005 p 1695-1697
León, Agathe; Mallolas, Josep; Martinez, Esteban; Lazzari, Elisa De; Pumarola, Tomás; Larrousse, Maria; Milincovic, Ana; Lonca, Montserrat; Blanco, Jose Luis; Laguno, Montserrat; Biglia, Alejandra; Gatell, Josep Maria
Infectious Diseases Unit, Hospital Clinic, University of Barcelona, Spain.
Abstract
We evaluated the virological outcome of tenofovir plus didanosine-based regimens in 67 HIV-suppressed patients. After a median follow-up of 26 months (IQR 10.5-40.5), 12 (18%) discontinued the therapy because of virological failure. At virological failure 'de novo' selected mutations were identified in 11 of the 12 failing patients, including the K65R mutation in seven patients. These results argue against the use of tenofovir-didanosine not only in naive patients, but also in previously suppressed patients.
ARTICLE TEXT
A high rate of early virological failure with the selection of resistance mutations has recently been reported in treatment-naive HIV-infected adults receiving a non-nucleoside reverse-transcriptase inhibitor (NNRTI) plus didanosine and tenofovir [1-3]. This combination has also been associated with a higher rate of pancreatitis [4] and with a lower CD4 cell response [5]. Therefore, the European Agency for the Evaluation of Medicinal Products pointed out that the co-administration of tenofovir and didanosine is not recommended within any antiretroviral combination unless considered strictly necessary [6]. We have assessed the outcome of virologically suppressed patients switching to a didanosine and tenofovir containing regimen.
From our database, we identified all HIV-infected adults (n = 67) with a plasma viral load less than 200 HIV-1-RNA copies/ml for at least 6 months who changed their antiretroviral regimen, following the criteria of the treating physician, to a new antiretroviral combination consisting of tenofovir (300 mg) a day plus didanosine EC (adjusted to body weight) plus a third nucleoside reverse-transcriptase inhibitor (NRTI) (n = 21), a NNRTI (n = 34) or a protease inhibitor (PI) (n = 12). The number of previous antiretroviral regimens, time on antiretroviral therapy and on viral suppression, archived resistance mutations and previous exposure to tenofovir or didanosine was assessed at baseline. Clinical data and routine laboratory parameters were evaluated at baseline and at least every 3 months thereafter.
We evaluated the outcome in an intent-to treat analysis in which death, clinical progression, loss to follow-up, switching study medication or the development of virological failure were considered endpoints, and we also evaluated the rate of virological failure (defined as HIV-1 RNA > 200 copies/ml in two consecutive measurements separated by at least one month) in an on-treatment analysis. Whenever virological failure was detected, a genotypic resistance test (Viroseq HIV-1 genotyping system, Applied Biosystems, Foster City, California, USA) was performed both at virological failure, before any antiretroviral therapy and in previous episodes of virological failure, if any. The time to study endpoint and to virological failure was estimated using the Kaplan-Meier method, and survival functions were compared among groups using the Wilcoxon-Breslow test and adjusting the significance level for the number of pairwise comparisons. Cox proportional hazards models were used to identify the univariate and multivariate factors associated with virological failure.
We identified 67 HIV-infected adults with a viral load less than 200 HIV-1-RNA copies/ml for at least 6 months who began a tenofovir plus didanosine-based regimen. The reasons for switching antiretroviral therapy were lipodistrophy in 24 patients (35.8%), toxicity to the previous regimen in 21 patients (31.3%), patient's desire to simplify the regimen in 16 (23.9%), and other reasons in the remaining six (9%). Regimens were based on an NRTI plus tenofovir plus didanosine in 21 patients (lamivudine, 14; abacavir, 6; zidovudine, 1), an NNRTI plus tenofovir plus didanosine in 34 patients (efavirenz, 20; nevirapine, 14), or a PI plus tenofovir plus didanosine in 12 patients (lopinavir/ritonavir, 7; atazanavir, 2; nelfinavir, 3). The median number of previous antiretroviral regimens were four and the median time on antiretroviral treatment was 63 months [interquartile range (IQR) 44-96 months].
After a median follow-up of 26 months (IQR 10.5-40.5 months), 41 of the 67 patients (61.2%) remained with the scheduled therapy. The main reasons for interrupting the medication were virological failure in 12 (18%), side-effects in 10 (15%), and loss to follow-up in four patients (6%).
In the multivariate analysis the body mass index was the only factor significantly associated with the time to study endpoint (crude hazard ratio (CHR) 0.8, 0.7-0.9, P = 0.04). The overall time to virological failure (P = 0.01) and to the interruption of medication (P = 0.021) were significantly different between the antiretroviral groups, being shorter when the third component was an NRTI compared with an NNRTI. Conversely, the CD4 cell response was similar in all three groups. The median time on treatment until virological failure was 8 months (IQR 6-15). The characteristics of the 12 patients developing virological failure are summarized in Table 1 [7]. The genotypic analysis at virological failure detected resistance mutations in all 12 patients, which were never detected at baseline before any antiretroviral therapy. Selected resistance mutations at virological failure were 'de novo' in nine patients (75%), both 'de novo' and emerging from archived mutations in two (16.7%) and only emerging from archived mutations in one (8.3%). The M184V mutation was selected in all failing NRTI-treated patients (seven out of seven). The L100I, K103N or Y181C mutations were selected in all failing NNRTI-treated patients (four out of four), and the K65R mutation was selected in seven of the 12 patients (58.3%). Previous exposure to tenofovir was the only factor significantly (odds ratio 5.6, 95% confidence interval 1.4-22.5, P = 0.02) associated with virological failure both in the univariate and multivariate analysis.
Our results suggest that in previously HIV-suppressed patients the combination of didanosine plus tenofovir may also lead to an early virological failure similar to that reported in antiretroviral-naive patients, and this is particularly true when the third member of the combination is an NRTI, but may also occur when it is an NNRTI or a PI although less frequently. A similar suboptimal response with the selection of resistance mutations has also been reported in two retrospective switching studies with tenofovir plus two NRTI [8,9]; however, no data were available for the combination of tenofovir-didanosine plus an NNRTI or a PI. Conversely, in previous switching studies with triple NRTI including a thymidine analogue viral suppression was very well maintained [10,11].
Moreover, our results showed a sustained virological response when a PI was the third component of a didanosine-tenofovir backbone. Only one case of virological failure (from 12 PI-treated patients) was detected without the 'de novo' selection of mutations, and we cannot ascribe this virological failure to the current therapy, but maybe to a lack of adherence. These results accord with the BMS Study AI424-0457 [12] of treatment-experienced HIV patients with an antiretroviral regimen containing a ritonavir-boosted PI, tenofovir and didanosine EC or another NRTI, in which good virological control in the didanosine arm has been reported.
|
|
| |
| |
|
|
|
