Answers are grounded in this paper and cite their sources.
To the Editor: Initial antiretroviral therapy of HIV-1 infection frequently contains lamivudine (3TC) as one of the components. Individuals failing 3TC-containing therapy frequently have HIV-1 variants with a reverse transcriptase (RT) gene mutation (methionine-to-valine substitution at codon 184 or M184V) conveying high-level resistance to 3TC and modest in vitro cross-resistance to didanosine (ddI).1 Despite this observation, the M184V mutation does not seem to decrease the antiretroviral response to ddI in vivo markedly,2,3 although isolating the specific impact of M184V on the ddI response is difficult. ddI added to existing therapy provided better short-term virologic responses compared with placebo, despite the fact that most patients had baseline (BL) M184V in addition to many other nucleoside reverse transcriptase inhibitor (NRTI) mutations.3 Few patients in this study lacked the M184V mutation, and, paradoxically, these individuals had poor responses. In addition, the relation between HIV-1 phenotypic susceptibility to ddI and the antiretroviral effect of ddI requires further clarification. In our study,4 ddI was the only active antiretroviral given to patients with and without M184V and no or modest numbers of other NRTI mutations. Phenotypic susceptibility to ddI was assessed retrospectively, allowing us to dissect the effects of these factors on ddI activity in relative isolation and providing important information for clinicians assessing relative activity of ddI when used as a component of multidrug therapy in treatment-experienced patients. Adult Clinical Trials Group (ACTG) 307 was a randomized, multicenter, double-blind study in which patients received ddI alone or in combination with hydroxyurea (HU) for the first 12 weeks of study.4 Randomization was stratified based on prior antiretroviral treatment experience. Subjects on antiretroviral medications discontinued therapy for 14 days before enrollment. The study population for this retrospective analysis was defined as those subjects who were treated initially with ddI plus HU placebo or ddI plus active HU, had a week 8 HIV-1 RNA result on therapy, and had a BL HIV-1 RNA level greater than 400 copies/mL. Using entry or pre-entry samples (27% and 73%, respectively), the HIV-1 RT region corresponding to codons 17 through 237 was sequenced using (Applied Biosystems, Foster City, CA) methods. Viral RNA was isolated from 140 μL of patient plasma with a QIAamp viral RNA kit (Qiagen, Valencia, CA). The RNA was eluted with 50 μL of RNAse-free water. Reverse transcription and polymerase chain reaction (PCR) amplification of 10 μL of the viral RNA solution were carried out using oligonucleotides derived from the HIV-1 subtype B pol gene consensus sequence. The sequence of the purified 665-base pair (bp) RT-PCR product was determined by Applied Biosystems PRISM BigDye Terminator cycle sequencing with AmpliTaq DNA polymerase, FS (Applied BioSystems) on an Applied Biosystems 377XL-96 DNA sequencer. The nucleotide sequences were aligned using Vector NTI (Informax, Carlsbad, CA) and compared with the sequence of the reference strain, HXB2. Sequences were manually confirmed and reviewed. Changes in HIV-1 RNA from BL to week 8 were examined in subjects grouped by 3TC experience and by the presence of M184V at BL. Mean change was estimated by parametric censored-data methods adjusted for BL viral load (VL) centered at 4.5 log10 copies/mL.5 Analyses were also stratified by the ddI and ddI plus HU arms. Proportions with a week 8 HIV-1 RNA level <400 copies/mL were compared using the Fisher exact test. Phenotype assays for ddI susceptibility at BL were performed by Monogram Biosciences (South San Francisco, CA) with results reported as a fold-change (FC), which is the ratio of the 50% inhibitory concentration (IC50) to ddI for the patient virus over the IC50 of the assay control wild-type HIV-1. Of the 134 subjects enrolled in the randomized study, 75 received initial ddI-containing therapy (26 received ddI alone and 49 received ddI with HU) and had a BL HIV-1 RNA level >400 copies/mL and a week 8 result on initial therapy. Forty-three (57%) were antiretroviral experienced, and 24 (32%) had previous 3TC. The BL median HIV RNA level (VL) was 4.4 log10 copies/mL, and the BL median CD4 count was 388 cells/mm3 and both were similar between 3TC-naive and -experienced patients. The BL RT sequence was available on 71 subjects. Nineteen had an M184V mutation; 10 of these had additional NRTI resistance mutations. Three had 2 accompanying thymidine analog mutations; the remainder had 1 or none.6 Six BL samples had RT nucleoside resistance mutations but lacked the M184V. No samples had L74V or I, K65R, 151 complex, or 69 insertion mutations. Sixty-six patients had BL ddI phenotype results, with a median FC of 1.0 and a narrow range (0.7-1.9). The relations between treatment history (Fig. 1A), BL genotype (Fig. 1B), and ddI phenotype are shown in Figure 1.FIGURE 1: ddI phenotypic FC at BL in relation to 3TC treatment (A) and the presence of M184V or other non-M184V NRTI mutations (B). Horizontal bars denote medians.In the week 8 analysis, including all patients initially treated with ddI or ddI plus HU, patients naive to 3TC had an approximately 0.4 log10 copies/mL greater mean decline in HIV-1 RNA than 3TC-experienced subjects (1.0 and 0.6 copies/mL, respectively, ddI arm; 2.0 and 1.6 copies/mL, respectively, ddI + HU arms; stratified P = 0.06). The mean decline in HIV-1 RNA level at week 8 in patients with M184V was less than in those without M184V regardless of previous 3TC experience but not significantly so (stratified P = 0.24). For subjects with a BL genotype, 4 (21%) of 19 with M184V achieved an 8-week VL <400 copies/mL, whereas 22 (42%) of 52 of those without M184V had a VL load <400 copies/mL (P = 0.16). VL response models accounting for censoring and including BL ddI FC and VL as covariates demonstrated a consistent relation between VL change and increasing FC, with a 0.7 log10 lesser decrease in VL for every unit increase in FC (P = 0.06). Results were similar when broken down by ddI and ddI plus HU treatment groups and when maximum VL response was the endpoint. The likelihood of attaining an 8-week VL <400 copies/mL diminished with increasing FC. The ddI FC values most strongly associated with an inflection in VL response using a maximum likelihood model were consistently in the range of 1.5 to 1.7 for pooled and subgroup analyses and using maximum VL response as the viral endpoint. Twenty-four (44%) of 55 subjects with a BL FC <1.5 achieved an 8-week VL <400 copies/mL, whereas only 1 (9%) of 11 subjects with a FC ≥1.5 had the same response (P = 0.04). Our study is limited by the relatively small sample size, the censoring of VL data by the assay lower limit of 400 copies/mL, the relatively narrow IC50 range, and the fact that resistance tests were obtained a median of 7 days before entry; therefore, individuals previously on therapy were off therapy for a brief period before resistance testing. The fact that ddI was the only antiretroviral used in our study enhances our ability to interpret the results, however. Our results show that the M184V mutation clearly has an impact on ddI phenotype (see Fig. 1B). Consistent with the results of others,2,3 we show that ddI has activity in the face of 3TC experience and in the presence of the M184V mutation, but these factors seem to diminish the antiviral effect of ddI. We also show that there is a consistent but narrow or compressed relation between ddI phenotype and antiretroviral response, with small increments in FC resulting in substantial reductions in the activity of ddI. Consistent with published lower and upper clinical cutoffs established for ddI in this assay (1.3- and 2.2-fold greater than controls),7 we found an inflection point in ddI activity between an FC from 1.5 to 1.7 and only 1 subject with an FC ≥1.5 achieved a VL <400 copies/mL at week 8. Similar to other antiretrovirals, ddI has its greatest activity in patients with wild-type virus and low FC on phenotype. Although diminished, ddI has persistent activity in the presence of the M184V mutation whether in the setting of modest treatment experience such as subjects in our study4 or in the face of more substantial treatment experience and nucleoside resistance mutations.3 ACKNOWLEDGMENTS The authors thank the subjects and ACTG investigators and staff who participated in ACTG 307 as well as Monogram Biosciences (South San Francisco CA), particularly Michael Bates, MD. This study was supported by the ACTG A1-27666; Statistical and Data Analysis Center, Harvard School of Public Health, A138855; AI38858 (University of North Carolina Virology Support Laboratory); RR0046; AI50410; AI25868; AI45008; and Bristol-Myers Squibb, which provided study medications and funding for phenotypic resistance testing. Joseph J. Eron, Jr., MD* Ronald J. Bosch, PhD† Daniel Bettendorf, MS† Leslie Petch, PhD‡ Susan Fiscus, PhD‡ Ian Frank, MD§ for the Adult Clinical Trials Group 307 Protocol Team *Department of Medicine University of North Carolina School of Medicine Chapel Hill, NC †Center for Biostatistics in AIDS Research Harvard School of Public Health Boston, MA ‡Department of Microbiology University of North Carolina School of Medicine Chapel Hill, NC §Department of Medicine University of Pennsylvania School of Medicine Philadelphia, PA
No takes yet. Share an insight, caveat, or question.
Eron et al. (2007) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: