Nucleoside analogue mutations and Q151M in HIV-1 subtype A/E infection treated with nucleoside reverse transcriptase inhibitors
Sirivichayakul, Suneea,b; Ruxrungtham, Kiata,b; Ungsedhapand, Chaiwatb; Techasathit, Wichaic; Ubolyam, Sasiwimolb; Chuenyam, Theshineeb; Emery, Seand; Cooper, Davidb,4; Lange, Joepb,e; Phanuphak, Praphana,b
From the aDepartment of Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand, the bHIV Netherlands Australia Thailand (HIV-NAT) Research Collaboration, The Thai Red Cross AIDS Research Centre, cSiriraj Hospital, Mahidol University; Bangkok, Thailand, the dNational Centre in HIV Epidemiology and Clinical Research (NCHECR), Sydney, Australia, and the eNational AIDS Therapy Evaluation Centre (NATEC), Amsterdam, the Netherlands.
Correspondence to K. Ruxrungtham, HIV-NAT, the Thai Red Cross AIDS Research Centre, 104 Rajadamri Road, Bangkok 10330, Thailand.
Received: 3 October 2002; revised: 27 February 2002; accepted: 11 March 2003.
Objectives: To investigate genotypic drug resistance in HIV-1 subtype A/E infection associated with failure of double/triple-nucleoside reverse transcriptase (RT) inhibitor therapy.
Methods: Patients from HIV-NAT 002 [stavudine (d4T)/didanosine (ddI) dose reduction study] and HIV-NAT 003 (zidovudine (ZDV)/lamivudine (3TC) versus ZDV/3TC/ddI) whose HIV-1 RNA was > 1000 copies/ml at week 48 and/or week 96 were tested for genotypic resistance. In both studies, after 48 weeks, patients were switched to the other dual or triple-nucleoside RT inhibitor (NRTI) either according to randomization or to the occurrence of virological failure.
Results: Resistance mutations found in the d4T/ddI, ZDV/3TC, and ZDV/3TC/ddI groups: none at baseline; at week 48, nucleoside analogue mutations (NAM), 2/17 (12%), 2/10 (20%), and 1/8; Q151M complex, 3/17 (18%), 0%, and 0%; M184V, 0%, 10/10 (P < 0.001), 3/8; V75T, 3/17 (18%), 0%, and 0%; L74V, 3/7 (18%), 0%, and 0%, respectively. At week 96, among the switchers, i.e., group A d4T/ddI to ZDV/3TC, group B ZDV/3TC to d4T/ddI, and group C ZDV/3TC/ddI to d4T/3TC/abacavir: NAM, 12/21 (57%), 4/7 and 1/3; Q151M, 4/21 (19%), 0% and 1/3, respectively. Interestingly, four or more NAM were observed in a higher proportion in group A (4/17 versus none in the others).
Conclusions: Multi-NRTI resistance (NAM and Q151M) and M184V (only in 3TC failure) are commonly found in HIV-1 subtype A/E infection associated with NRTI failure. Suboptimal d4T/ddI therapy led to a high incidence of V75T and L74V mutations. Switching from d4T/ddI to ZDV/3TC may be associated with a higher incidence of four or more NAM. Thus, suboptimal and dual NRTI therapy is not recommended for global application.
Combinations of highly potent antiretroviral drugs given to HIV-1 infected immunocompromised patients can restore immunocompetence, improve quality of life and prolong life [1–4]. Such effective combinations usually consist of two nucleoside analogue reverse transcriptase inhibitors (NRTI) plus one non-NRTI or protease inhibitor — highly active antiretroviral therapy. However, HAART is out of reach for the majority of patients in the developing world due to its high cost. Thus, in countries with moderate resources dual-nucleoside therapy has often been used instead. The frequently used dual-nucleoside combinations were zidovudine (ZDV)/didanosine (ddI), ZDV/lamivudine (3TC), stavudine (d4T)/ddI and d4T/3TC. Multi-drug resistance against NRTI – nucleoside analogue mutations (NAM) – Q151M complexes and 69 insertion complexes has been well reported and characterized in subtype B infection . The HIV Netherlands Australia Thailand Research Collaboration (HIV-NAT), an HIV clinical trial centre in Bangkok  has cohorts of Thai patients infected predominantly with HIV-1 subtype A/E (also known as subtype CRF01_AE), who were treated with d4T/ddI or ZDV/3TC for as long as 96 weeks. Approximately half of the patients were switched to the opposite double-nucleoside combination after week 48 [7,8]. Changes in the genotypic resistance patterns of these two cohorts of patients are described in this report.
Materials and methods
Patients from two HIV-NAT cohorts, HIV-NAT 002 and 003 are included in the study. HIV-NAT 002 was a randomized comparative study of low- or high-dose d4T combined with low- or high-dose ddI versus ddI monotherapy in antiretroviral drug-naive patients . All patients in the ddI monotherapy arm were switched to high-dose d4T/ddI at 24 weeks. At week 48, patients without virologic failure (i.e., HIV RNA < 10 000 copies/ml or < 1.0 log10 increase above nadir) were randomized to receive ZDV/3TC for another 48 weeks, (`immediate switch’ group) or to continue on high-dose d4T/ddI until virologic failure, then switched to ZDV/3TC (`deferred switch’ group). HIV-NAT 003 was a randomized comparative study of ZDV/3TC versus ZDV/3TC/ddI, also in antiretroviral-naive patients . At week 48, patients without virologic failure were randomized to continue on ZDV/3TC or on ZDV/3TC/ddI (deferred switch) or changed to d4T/ddI or to d4T/3TC/abacavir, respectively for another 48 weeks (immediate switch).
Patients with HIV RNA > 1000 copies/ml at weeks 48 and/or 96 were selected for genotypic resistance study. Stored (at –70°C) EDTA-treated plasma specimens at baseline, week 48 and/or week 96 were retrieved for genotypic resistance assay.
RNA separation and reverse transcriptase gene amplification
HIV RNA was extracted from EDTA-treated plasma by using the guanidinium isothiocyanate and isopropanol precipitation technique. HIV reverse transcriptase (RT) RNA was reverse transcribed with B887-3 primer (5′-ATAGGCTGGACTGTCCATCTGTCA GG-3′). The first round PCR was carried out with primers A-35 (5′-GGTTGTACTTTAAATTTCCCA ATTAGTCC-3′) and B887-3 and then nested with primers B887-2 (5′-CTGTACCAGTAACATTAAAG CCAGG-3′) and B887-3. The 711 base pair PCR products were purified by Qiagen and then the sequencing reaction was commenced. The sensitivity of the assay was > 95% in plasma samples with HIV-1 RNA > 1000 copies/ml.
RT gene sequencing
The purified PCR products sequencing reaction by using the ABI PRISM dideoxy Dye Terminator Cycle Sequencing Kit (BigDye, Applied Biosystems, Foster City, California, USA), and were analysed on an ABI PRISM 310 automatic sequencing system. Sequence Navigator Software (Applied Biosystems) was used for analysis.
Data for baseline viral load and CD4 T lymphocyte counts were reported as median (range). The differences in percentages of the specific mutations among the arms were analysed using Fisher's Exact test and the Mann–Whitney U test. P values < 0.05 were considered statistically significant.
Overall baseline characteristics are summarized in Table 1. The baseline CD4 cell count and viral load of the patients included for the genotypic resistance study were not significantly different among the three groups.
HIV-NAT 002/002.1 cohort
Twenty-one out of the 71 patients (29.6%) who completed the 48 weeks of HIV-NAT 002 study had HIV RNA > 1000 copies/ml. Only 17 patients had available RT sequence for analysis, in two the RT gene was not amplified and the other two did not have week 48 specimens available. Of these 17 patients, four were from the initial ddI monotherapy group and the other 13 were distributed evenly in all d4T/ddI combination arms.
An additional nine patients were selected at week 96 from HIV-NAT 002.1 study on the basis of having HIV RNA > 1000 copies/ml at week 96. Eight out of these nine patients had HIV RNA < 1000 copies/ml at week 48. The other patient had HIV RNA 4763 copies/ml at week 48 but the RT gene could not be amplified from the week 48 specimen. Of these nine patients, six were in the immediate switch group (switched to ZDV/3TC at week 48), one was switched at week 52 whereas the other two were on d4T/ddI through week 96.
Twenty out of the 101 patients (20%) who completed the 48-week HIV-NAT 003 study had plasma HIV RNA > 1000 copies/ml, 11 from the ZDV/3TC arm and nine from the ZDV/3TC/ddI arm. Eight of the 11 patients in the ZDV/3TC arm who were studied for RT genotypic analysis at week 48 also had plasma HIV RNA > 1000 copies/ml at week 96. Three of nine in the ZDV/3TC/ddI arm also had plasma HIV RNA > 1000 copies/ml at week 96. In addition, eight more patients were identified at week 96 with HIV RNA > 1000 copies/ml, six from the double-nucleoside arm of whom five were still on ZDV/3TC by week 96. The other two patients were still on ZDV/3TC/ddI by week 96.
RT genotypic resistant mutations following d4T/ddI with or without switching to ZDV/3TC treatment (HIV-NAT 002/002.1 study)
At baseline, all nine of the patients who had successful RT genotypic analysis showed wild-type genotype. Two of these patients had valine to leucine substitution at position 75 of RT (V75L) which is different from the known d4T resistance mutations (V75T, V75I). At week 48, 10 of the 17 selected d4T/ddI-treated patients had RT genotypic resistance mutations (Table 2). Of these 10 resistance mutations, three were associated with d4T resistance (V75T), three with ddI resistance (two with L74V and one with K65R), two with NAM (M41L/D67N/T215F, D67N/K70R) and two were associated with multi-drug resistance (MDR; Q151M complex), as shown in Fig. 1a. The d4T resistant (V75T), NAM and MDR-associated mutations were distributed evenly in all d4T/ddI arms whereas all the three ddI resistance mutations (L74V) clustered in the group given ddI monotherapy for the first 24 weeks.
At week 96, 24 of the 25 patients (96%) had RT genotypic resistance mutations including the same 10 patients who had had mutations at week 48. The most frequently found mutation (18/24, 75%) was the 3TC-associated mutation (M184V). The next were NAM (M41L, D67N, K70R, L210W, T215F, K219Q), found in 15 cases (62.5%) (Tables 2 and 3, Fig. 1b). Four patients had an isolated M184V mutation at week 96. All but three of the ZDV/3TC-related mutations had received ZDV/3TC for 24–48 weeks. The three Q151M mutations at week 48 persisted at week 96 and two new Q151M complexes were found at week 96 (patient 12, Table 2; patient 6 in Table 3). Two d4T-associated (V75T) mutations at week 48 disappeared at week 96, one persisted but changed its signature from V75T to V75M (patient 68, Table 2). Three V75I mutations appeared at week 96 (patients 6, 12 and 55, Tables 2 and 3). All three of these patients also had Q151M mutations.
RT genotypic resistance mutations following ZDV/3TC with or without switching to d4T/ddI treatment (HIV-NAT 003/003.1 study)
All HIV-NAT 003 patients who had RT genotypic analysis at baseline had wild-type genotype. In the ZDV/3TC and ZDV/3TC/ddI groups, 10/11 and 8/9 patients’ plasma samples were available for analysis. All 10 patients in the ZDV/3TC arm (100%) showed M184V mutation at week 48. Two of these patients also had NAM (Table 3, Fig. 1a). Every M184V mutation disappeared at repeat genotypic analysis at week 96 if ZDV/3TC had been discontinued between weeks 48 and 60. However, in the two cases with NAM (patients 385 and 390, Table 4) the mutations still persisted at week 96 despite changing to d4T/ddI at week 48.
RT genotypic resistance mutation following sequential triple-nucleoside analogue therapy
For the ZDV/3TC/ddI arm, five of the selected eight samples at week 48 still had wild-type genotype despite virologic failure. The M184V mutation was found in three patients including the one with concurrent NAM (Table 4). By week 96, five of these initial eight patients had HIV RNA < 1000 copies/ml as a result of changing ZDV/3TC to d4T/3TC/abacavir. One patient had wild-type genotype despite continuing on 3TC but with a different triple-nucleoside regimen (patient 357, Table 4). A new M184V appeared (patient 352) and the one patient with NAM continued to have the same mutations plus Q151M despite stopping ZDV at week 48 (patient 370, Table 4). For an additional seven patients who had plasma HIV-1 RNA > 1000 copies/ml at week 96, all had M184V mutation. Three of these patients also had NAM (Table 5). Only one patient was switched from ZDV/3TC to d4T/ddI at week 48 (immediate switch group) but still had the M184V mutation at week 96 (patient 389, Table 5). None of the 25 patients in the HIV-NAT 003/003.1 study who had RT sequenced at week 48 or 96 had any d4T or ddI-associated mutations despite being switched to d4T/ddI for as long as 48 weeks.
Of all evaluable patients who had plasma HIV-1 RNA > 1000 copies/ml, 66% (23/35) and 88% (38/43), had at least one of the known NRTI resistance-associated mutations at week 48 and 96 of treatment, respectively. The most common mutation was M184V (primary mutation of 3TC) – 38% (13/35) at week 48 and 67% (29/43) at week 96. All of these patients had received a 3TC-containing combination. The other common resistance mutations were NAM (14% and 30% at week 48 and 96, respectively) and the Q151M complex (9% and 14%,). There were no 69 insertion complexes seen in this cohort.
Frequency of many codons of NAM
As shown in Fig. 2a, more NAM mutations (≥ 3) were observed in group A (suboptimal d4T/ddI-treated) compared with group B (ZDV/3TC-treated) at 48 weeks (6% versus 0%). At week 96, four or more NAM mutations were also found only in group A (d4T/ddI with switching to ZDV/3TC; 14.3%) whereas none was found in the other groups (Fig. 2b). However, because of the small sample size, the differences were not statistically significant.
We report the first systematic study of genotypic resistance in the RT gene in predominantly HIV-1 subtype A/E-infected individuals receiving double- and triple-nucleoside analogue therapy. HIV-1 subtyping was done by V3 peptide serotyping . Sixty of the 78 patients (77%) from the HIV-NAT 002 study and 62 of the 72 patients (83.3%) from the HIV-NAT 003 study were found to be subtype A/E . Genotyping of selected patients from these cohorts (five from HIV-NAT 002 and 19 from HIV-NAT 003) confirmed subtype A/E . This is the longitudinal analysis of changes in resistance mutations in patients on long-term treatment (96 weeks) with d4T/ddI or ZDV/3TC or ZDV/3TC/ddI with options for randomized switching to the other double- or triple-nucleosides after 48 weeks of treatment or when the defined virologic failure was observed. Both of our cohorts (HIV-NAT 002 and 003) consisted of antiretroviral-naive patients who started their antiretroviral treatment in 1996 and 1997 [7,8]. None of these patients had any known resistance mutations to nucleoside analogues at baseline.
In this study, NRTI-associated resistance mutations were commonly found in patients who had failed NRTI treatment. As shown in Fig. 1a and b, the detection rate was up to 70% and 90% at weeks 48 and 96, respectively. The most common mutation was M184V (primary mutation of 3TC), i.e., up to 40% at week 48 and up to 70% at week 96. All of these patients, however, had received 3TC-containing regimens. The other common resistance mutations were NAM (14% and 30%) and the Q151M complex (9% and 14% at week 48 and 96, respectively). The findings indicate that genotypic resistance was also commonly found in predominantly HIV-1 subtype A/E-infected patients with NRTI treatment failure. There was however no 69 insertion complex seen in this cohort.
NAM are defined as any of the six codon mutations M41L, D67N, K70R, L210W, T215F/Y, K219Q . A greater number of codon mutations, i.e., four or more, has been found to be associated with cross-resistance with other NRTI except 3TC. As shown in Fig. 2a, a greater number of NAM codons NAM (≥ 3) was observed in the group A (suboptimal d4T/ddI-treated group) compared to group B (ZDV/3TC-treated group) at 48 weeks. And at 96 weeks, four or more NAM codons were found 14.3% in group A (d4T/ddI with switching to ZDV/3TC), whereas none was found in the other groups (Fig. 2b). However, the differences were not statistically significant probably due to the limited sample size. The results suggest that suboptimal d4T/ddI treatment may be associated with an increased risk of multiple NAM codon mutations that may lead to a higher level of intra-class cross-resistance .
We found that in the HIV-NAT 002 cohort d4T and ddI genotypic resistance occurred more readily than in patients from developed countries similarly treated with d4T/ddI [11,12]. ddI genotypic resistance (L74V) occurred exclusively in the ddI monotherapy group (during the first 24 weeks). d4T genotypic resistance (V75T/M) occurred only in the low-dose d4T groups and in the ddI monotherapy pretreated group for the first 24 weeks. Thus, suboptimal dosing of d4T/ddI leads to the more rapid development of d4T or ddI resistance. Genotypic resistance to d4T (V75T/M), ddI (L74V) and Q151M was not found in d4T/ddI-treated patients [with the exception of patient 382 (Table 4) who had two NAM mutations] if they were initially pretreated with ZDV/3TC. Our results also confirm that Q151M complex emerged mostly in patients receiving ddI-containing regimens [5,11,13,14].
All 10 patients who started with ZDV/3TC had M184V mutation conferring resistance to 3TC by week 48. In fact, almost all patients studied had this mutation by week 24 (data not shown). Two of these 10 patients also had NAM (Table 4). By contrast, five out of eight patients in the ZDV/3TC/ddI group still had wild-type RT genotype by week 48 even with HIV RNA > 1000 copies/ml. The remaining three of these patients had the 3TC resistance mutation (M184V) including one patient with NAM. This indicates that a triple-nucleoside 3TC-containing regimen (i.e., ZDV/3TC/ddI) could delay the emergence of the 3TC resistance mutation for up to 48 weeks. Of interest, in patients treated with ZDV/3TC, NAM rarely developed for up to 96 weeks of treatment (Tables 4 and 5). This is relatively low as compared with HIV-1 subtype B-infected patients similarly treated with ZDV/3TC . The reasons may be difference in subtype, low baseline viral load, higher ZDV dosage in relation to body weight or better adherence in our patients. It will be of interest if phenotypic resistance assay in these two cohorts of patient could be performed in order to confirm the genotypic resistance data.
This study also offers the opportunity to follow the changes in RT resistance mutations when the RT inhibitors are changed. Most RT resistance mutations disappeared upon changing nucleoside analogues, for example: V75T, L74V, and M184V. One V75T mutation changed its signature to V75M 48 weeks after switching treatment to ZDV/3TC (patient, Table 2). V75M has been described as a d4T-related resistance mutation . MDR mutations i.e., NAM and the Q151M complex, however, always persisted despite nucleoside analogue changes. These resistance mutations predict unresponsiveness to all drugs in the nucleoside analogue family. Our study is limited by the small number of patients eligible for genotypic analysis in each arm. Nevertheless, the results indicate that ddI monotherapy and dose reductions of d4T and ddI are contraindicated, d4T/ddI may be a suitable alternative for patients failing ZDV/3TC but not the reverse and use of genotypic analysis may be helpful in guiding therapy with nucleoside analogues.
In countries where dual nucleoside therapy has been widely used such as Thailand (ZDV/ddI has been widely prescribed in the past), the selection of treatment options after failure is a real challenge. Our unpublished data found that more than 90% of patients who failed ZDV/ddI showed carry-over NAM that are cross-resist to all NRTI except 3TC . Switching from ZDV/ddI to a d4T/3TC/nevirapine fixed dose combination (which is currently available in Thailand as the cheapest triple regimen) may result in an increased risk of rapid virological failure due to the rapid emergence of 3TC/nevirapine resistance. A combination of two new classes such as efavirenz with ritonavir-boosted protease inhibitors may be the most appropriate option in this setting. However, the study of the long-term efficacy of this option is underway and the cost is the major obstacle for its implementation in a resource-limited setting.
In summary, multi-NRTI resistance (NAM and Q151M) and M184V (only in 3TC failure) are also commonly found in subtype A/E infection associated with NRTI failure. Suboptimal d4T/ddI therapy led to a high incidence of V75T and L74V mutations. Switching from d4T/ddI to ZDV/3TC may be associated with a higher incidence of NAM (≥ 4 codons, in particular). Thus, for the global implementation of antiretroviral therapy, suboptimal and dual NRTI therapies should be avoided. Considerable efforts from all sectors are needed to make highly active antiretroviral therapy accessible worldwide for all patients in whom treatment is indicated.
Sponsorship: Supported mainly by the HIV-NAT Intramural Research Fund, The Thai Red Cross AIDS Research Center, and partially supported by the Chulalongkorn Medical Research Center (Chula MRC), Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand.
1. Connick E, Lederman MM, Kotzin BL, Spritzler J, Kuritzkes DR, St Clair M, et al. Immune reconstitution in the first year of potent antiretroviral therapy and its relationship to virologic response. J Infect Dis 2000, 181:358–363.
2. Al-Harthi L, Siegel J, Spritzler J, Pottage J, Agnoli M, Landay A. Maximum suppression of HIV replication leads to the restoration of HIV-specific responses in early HIV disease. AIDS 2000, 14:761–770.
3. Yeni PG, Hammers SM, Carpenter CC, Cooper DA, Fischl MA, Gatell JM, et al. Antiretroviral treatment for adult HIV infection in 2002. Updated recommendations of the International AIDS Society-USA Panel. JAMA 2002, 288:222–235.
4. Autran B, Carcelain G, Li TS, Blanc C, Mathez D, Tubiana R, et al. Positive effects of combined antiretroviral therapy on CD4+ T-cell homeostasis and function in advanced HIV disease. Science 1997, 277:112–116.
5. D’ Aquila RT, Schapiro JM, Brun-Vezinet F, Clotet B, Conway B, Demeter LM, et al. International AIDS Society-USA Drug Resistance Mutations Group. Drug resistance mutations in HIV-1. Topics in HIV Med 2002, 10:21–25.
6. Phanuphak P, Cooper DA, Lange J. Clinical trials in Asia. AIDS 1998, 12(suppl. B):S163–S167.
7. Ruxrungtham K, Kroon EDMB, Ungsedhapand C, Teeratakulpisarn S, Ubolyam S, Buranapraditkun S, et al. A randomized, dose-finding study with didanosine plus stavudine versus didanosine alone in antiviral-naïve, HIV-infected Thai patients. AIDS 2000, 14:1375–1382.
8. Ungsedhapand C, Kroon EDMB, Suwanagool S, Ruxrungtham K, Yimsuan N, Sonjai A, et al. A randomized, open-label, comparative trial of zidovudine plus lamivudine versus zidovudine plus lamivudine plus didanosine in antiretroviral-naïve HIV-infected Thai patients. J Acquir Immune Def Syndr 2001, 27:116–123.
9. Ubolyam S, Ruxrungtham K, Sirivichayakul S, Okuda K, Phanuphak P. Evidence of three HIV-1 subtypes in subgroups of individuals in Thailand. Lancet 1994, 344:485–486.
10. Sirivichayakul S, Chantratita W, Sutthent R, Ruxrungtham K, Phanuphak P, Oelrichs RB. Survey of reverse transcriptase from the heterosexual epidemic of human immunodeficiency virus type 1 CRF01_AE in Thailand from 1990–2000. AIDS Res Hum Retroviruses 2001, 17:1077–1081.
11. 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:F9–F15.
12. Picard V, Angelini E, Maillard A, Race E, Clavel F, Chene G, et al. Comparison of genotypic and phenotypic resistance patterns of human immunodeficiency virus type 1 isolates from patients treated with stavudine and didanosine or zidovudine and lamivudine. J Infect Dis 2001, 184:281–284.
13. Pellegrin I, Izopet J, Reynes J, Denayrolles M, Montes B, Pellegrin JL, et al. Emergence of zidovudine and multidrug resistance mutations in the HIV-1 reverse transcriptase gene in therapy-naïve patients receiving stavudine plus didanosine combination therapy. AIDS 1999, 13:1705–1709.
14. Ross L, Scarsella A, Raffanti S, Henry K, Becker S, Fisher R, et al. Thymidine analog and multinucleoside resistance mutations are associated with decreased phenotypic susceptibility to stavudine in HIV type 1 isolated from zidovudine-naïve patients experiencing viremia on stavudine-containing regimens. AIDS Res Hum Retroviruses 2001, 17:1107–1115.
15. Ait-Khald M, Rakik A, Griffin P, Cutrell A, Fischl MA, Clumeck N, et al. Mutations in HIV-1 reverse transcriptase during therapy with abacavir, lamivudine and zidovudine in HIV-1-infected adults with no prior antiretroviral therapy. Antivir Ther 2002, 7:43–51.
genotypic resistance; stavudine; didanosine; zidovudine; lamivudine; abacavir; HIV-1 subtype A/E; nucleoside reverse transcriptase inhibitors
© 2003 Lippincott Williams & Wilkins, Inc.
Highlight selected keywords in the article text.