- Email: [email protected]
Resistance to antiretroviral drugs in patients with primary HIV-1 infection
International Journal of Antimicrobial Agents 16 (2000) 429 – 434 www.ischemo.org
Resistance to antiretroviral drugs in patients with primary HIV-1 infection Bluma Brenner a, Mark A. Wainberg a,*, Horacio Salomon a, Danielle Rouleau b, Andre´ Dascal a, Bonnie Spira a, Rafik-Pierre Sekaly c, Brian Conway d, Jean-Pierre Routy e, Investigators of the Quebec Primary Infection Study a
McGill Uni6ersity AIDS Centre, Lady Da6is Institute-Jewish General Hospital; 3755 Cote Site-Catherine Road, Montreal, Quebec, Canada H3T 1E2 b The Reseau FRSQ-SIDA, Montreal, Quebec, Canada H3T 1E2 c Institut de Recherches Cliniques, Montreal, Quebec, Canada H3T 1E2 d Viridae Clinical Sciences Inc., Vancou6er, BC, Canada H3T 1E2 e Royal Victoria Hospital, Montreal, Quebec, Canada
Abstract The widespread use of antiretroviral agents (ARVs) and the growing occurrence of HIV strains resistant to these drugs have given rise to serious concerns regarding the transmission of resistant viruses to newly infected persons. Plasma viral RNA from 80 individuals newly infected between 1997 and 1999 was genotyped by automated sequencing to analyze the profile of viruses resistant to nucleoside and non-nucleoside reverse transcriptase inhibitors (NRTIs and NNRTIs) and to protease inhibitors (PIs). The prevalence of mutations that conferred primary resistance to PIs (L10I, D30Y, V82A, L90M) was 15% of the cohort. RT genotypic variants, associated with high-level resistance to ARVs, were observed in 21% of individuals, including NRTI, NNRTI and multidrug (MDR) resistance in 6, 5, and 10% of cases, respectively. The phenotypic susceptibility of viral isolates to ARVs was also assayed and showed transmission of high-level resistance to ZDV, 3TC, and PIs in those individuals with MDR. The transmission of drug-resistant HIV genotypic variants is a serious problem that merits further attention by public health officials, virologists, and clinicians. © 2000 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. Keywords: Antiretroviral therapy; HIV drug resistance/resistance mutations; Reverse transcriptase inhibitors; Protease inhibitors
1. Introduction While potent antiretroviral drugs (ARVs) are of proven benefit in individuals and communities with HIV-1 infection, the evolution of ARV drug resistance is a leading factor in treatment failure [1,2]. Drug resistance arises as a direct result of an error-prone viral reverse transcriptase (RT) causing ongoing HIV mutagenesis. ARV drug pressure amplifies resistant variants that carry resistance-associated mutations in the viral RT and protease (PR) enzymes [2 – 5]. Advances in genotypic analysis have established over 150 changes in the RT and PR regions associated with selective resistance to nucleoside and non-nucleoside * Corresponding author. Tel.: + 1-514-3408260; fax: 1-5143407537. E-mail address: [email protected] (M.A. Wainberg).
RT inhibitors (NRTIs and NNRTIs) and PR inhibitors. An important consideration in ARV resistance, under conditions of drug pressure, is the selective growth advantage of mutated viruses [2]. High-level resistance to indinavir (Ind) and zidovudine (ZDV) take long periods (6 months to years) to develop, requiring the sequential accumulation of multiple mutations [6–8]. In contrast, resistance to nevirapine (Nev) and 3TC develops rapidly, since single mutations confer 500–1000fold increases in resistance [9,10]. In contrast, resistance to didanosine (ddI), zalcitabine (ddC) and stavudine (d4T) emerge slowly since pertinent mutations confer only low-level resistance [2]. Particularly problematic is the emergence of PI (L90M), NRTI (Q151M and F77L) and NNRTI (K103N) single mutations that confer cross-resistance to many ARVs of the same class [2,5,11].
0924-8579/00/$ - $20 © 2000 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. PII: S 0 9 2 4 - 8 5 7 9 ( 0 0 ) 0 0 2 7 0 - 3
430
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
Distinct from the obvious question of the relationship between drug resistance and treatment failure, is the problem of transmission of resistant virus among the general HIV-infected population. Identification of HIV resistance in newly infected patients has great clinical and public health significance, especially if available drugs are ineffective against resistant strains present in primary infection. Unless this topic is successfully addressed, the likelihood is that drug-resistant variants of HIV, if fit, will sustain the epidemic and limit the effectiveness of antiviral therapy. Since the initial observation of transmission of ZDVresistant isolates in 1993, many case reports have shown the sexual transmission of ARV-resistant HIV isolates from treated individuals to their drug-native partners [12 – 19]. The frequency of the ZDV resistance mutation, T215Y, in cohorts with primary infection has increased from 1.45 (1988 – 1991) to 7.5 in 1992 and 10.4% in 1993 and 1994 [19]. HIV variants carrying RT mutations at codon-215, (conferring high level ZDV resistance), currently range from 6 to 11% in newly infected cohorts [20 – 22]. However ZDV is unique, insofar as it was the first antiretroviral drug used, often in monotherapy. It is now important to determine how these findings apply to other types of drug-resistant viruses and to evaluate the extent to which transmitted viruses resistant to inhibitors of both RT and PR can cause disease. Two studies have shown the presence of genotypic markers, associated with ARV resistance in 16 – 19% of newly infected cohorts as compared with a study showing genotypic resistance in only 2% of a cohort comprised of patients dating to 1989, when PIs were not in use [20–22]. Multidrug (NNRTI/NRTI/PI) resistance (MDR) was observed in 2 – 4% of cases [20 – 22]. The aim of this investigation was to monitor the prevalence of resistance to ARVs and RT and PR mutations in a large cohort of individuals infected between 1997 and 1999 in the metropolitan area of Montreal, Canada, when universal access to combination chemotherapy was available. Our findings show that despite a high prevalence of resistance-associated mutations in newly infected individuals, these mutations do not confer high-level antiretroviral drug resistance in cell culture.
immune response with reactive EIAs and indeterminant Western blot, and p24 antigenemia with emergent positive HIV serology within 3 months; (c) HIV seroconversion within a 6-month period following negative EIA testing; and/or (d) typical acute retroviral syndrome within the earlier 3 months. The cohort included 80 persons where transmission was established to be sexual (n= 56, 93% males), intravenous (n= 21, 80% male), or unknown (n= 3). CD4 counts and viral loads of the sexual (493 9 29 cells per l and 338 1139110 882 copies per ml) and intravenous (5729 53 cells per l and 183 151 9 71 799 copies per ml) transmission groups were similar. HIV-1 antibody testing was performed at three university sites. Western blot analysis for HIV-1 antibodies and p24 antigen capture assays were performed at the Laboratoire de Sante´ Publique de Que´bec. Plasma viremia was measured using the Roche Amplicor assay (Roche Diagnostics, Mississauga, ON). Flow cytometric analysis determined T-cell subset distributions.
2.2. Phenotypic drug susceptibility testing Drug susceptibility was measured by determining the extent to which ARVs inhibit in vitro HIV replication [23–25]. PBMCs from 62 patients were co-cultured with PHA-pre-stimulated cord blood mononuclear cells (CBMCs) from normal donors. Levels of virus in cocultures were monitored weekly by p24 antigen capture assays until deemed positive after 3–5 weeks (a first p24 value of \ 20 pg/ml followed by a second value of \ 100 pg/ml). High-titer stocks of primary virus isolates were obtained from only 22 patients. These isolates were re-amplified and quantified by RT enzyme assays to generate defined, titrated clinical isolates to minimize inter-inoculum effects [24]. CBMCs infected with patient isolates were then plated in 96-well plates both in the absence and presence of a variety of ARV dosages. After 7 days, RT enzyme assays were used to determine the 50% drug inhibitory concentration (ID50) [23–25]. The observed ID50 values of patient viral isolates were then compared with the ID50 values obtained for HIV-IIIb, the ARV susceptible control [23].
2.3. Sequence analysis 2. Materials and methods
2.1. Study population In January 1997, a clinical research network was established to enrol individuals with clear documentation of recent HIV infection. Enrolment criteria included (a) non-reactive ELISA immunoassays (EIAs) in subjects with p24 antigenemia; (b) evolving humoral
Viral RNA was isolated from plasma samples using the QIAamp viral extraction kit (Qiagen Inc., Chatsworth, California). The TruGene HIV-1 Assay Gene System was used in conjunction with the Open Gene automated DNA sequencer, (Visible Genetics Inc., Toronto, Canada), to sequence the PR and RT regions of HIV-1 cDNA. Testing involved simultaneous clip sequencing of PR and codons 35–244 of RT from amplified cDNA in both the 3% and 5% directions. Se-
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
quences were aligned and compared with a LAV-1 consensus sequence using Visible Genetics Gene Librarian software.
3. Results
3.1. Phenotypic drug susceptibility testing PBMCs from 62 patients with primary HIV infection were co-cultured with PHA-pre-stimulated cord blood mononuclear cells (CBMCs) from normal donors. High-titer viral stocks were obtained from 22 subjects. Titrated viral inocula were used to infect CBMCs, incubated in the presence or absence of ARVs. The drug inhibitory concentrations (ID50) of PBMC-derived isolates were compared with the genotype of cDNA derived from plasma viral RNA. The spectrum of ID50 values for the three NRTIs tested, ZDV, ddI, and 3TC, and the three PIs tested, ritonovir (Rit), Ind, and saquinavir (Saq), are shown in Figs. 1 and 2. Based on numerous studies in our laboratory with clinical isolates, phenotypic resistance was considered to be present when ID50 values exceeded 0.1 mm, 1 mm, 50 mm, and 0.1 mm for ZDV, ddI, 3TC, and all PIs, respectively. These values exceed the laboratory mean ID50 value for HIVIIIb by at least 10-fold. Our findings show transmission of phenotypic resistance to ARVs in 5 of 22 samples (22.7%). Isolates from three (13.6%) of the newly infected individuals showed high level resistance to ZDV and one isolate showed low level ZDV resistance (Fig. 1). These four subjects were the only individuals among the 22 tested who carried either the T215Y or T215F mutations. Similarly, isolates from three subjects (13.6%) carrying the M184V mutation showed high level resistance to 3TC (Fig. 1). ID50 values for ddI for all the clinical isolates from subjects with primary infection were within the range of control values for HIVIIIb (Fig. 1).
431
None of these isolates were from individuals carrying the L74V mutation. Phenotypic cross-resistance to PIs was observed in isolates from three of the subjects, carrying viruses with the L90M mutation, conferring high level PI resistance (Fig. 2). Thus, there was concordance in transmission of resistance to ZDV, 3TC, and PIs in 13.6% of tested clinical isolates, consistent with genotypic data showing the T215Y/F, M184V, and L90M substitutions in plasma viral RNA.
3.2. Sequence analysis of protease and RT regions Viral cDNA derived from plasma viral RNA was sequenced to identify protease and RT mutations from 80 persons. In all, 60 patients (75%) carried at least one mutation in regions associated with PI resistance. However, many mutations may represent polymorphisms in clinical viral isolates rather than true resistance-conferring mutations. These include L63P/S/A/N, M36l, and V77l codon substitutions, associated with low-level resistance to nelfinavir (Nelf), Ind, and/or Rit, present in 50, 30 and 12.5% of our cohort, respectively. Excluding these three codons, protease substitutions associated with resistance to PIs were observed in 18 subjects, 22.5% of our cohort [21]. Similarly, 17 persons (21.5%) carried mutations in regions of RT associated with resistance to NRTIs and NNRTIs. The spectrum of RT and PI resistance mutations in our cohort of 80 subjects with primary infection is shown in Table 1. Eight patients harbored viruses with RT and protease mutations, conferring high level resistance to each of PIs, NRTIs and NNRTIs, i.e. 10% of cases had evidence of multidrug resistance. In three of these patients (Table 1, patients 69, 73, and 80), source partners were identified who carried identical genotypic variants. Although NRTIs were not included in our initial phenotypic drug resistance screening, subsequent testing of these clinical isolates showed low-level pheno-
Fig. 1. Phenotypic drug testing for ZDV, 3TC and ddI resistance in 22 subjects. The ID50 values for each ARV in each subject are indicated. The mean ID50 values observed for drug-sensitive HIVIIIb is indicated by arrows. The laboratory thresholds established for low and high drug resistance are also shown.
432
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
Fig. 2. Phenotypic drug testing for Rit, Ind and Saq resistance in 22 subjects. The ID50 values for each ARV in each subject are indicated. The mean ID50 values observed for drug-sensitive HIVIIIb is indicated by arrows. The laboratory thresholds established for low and high drug resistance are also shown.
typic resistance to Nev (data not shown). The primary L90M and V82A/F/T/S, associated with PI cross-resistance, were present at frequencies of 3.8 and 4%, respectively. The profile of mutations, in relation to resistance to ARVs, is shown in Fig. 3. Mutations conferring resistance to ZDV were present in 8.8% of subjects with 5% of subjects carrying the T215Y or the T215F mutations. The M184V mutation conferring resistance to cytosine analogues, including high-level resistance to 3TC and low-level resistance to ddC and Abacavir (Aba) was present in 5% of our cohort. The K103N mutation, conferring high level cross-resistance to NNRTIs, appeared in 3.8% of subjects. Mutations associated with cross-resistance to different classes of NRTIs, including F116S, F77L, A62V, were observed in three patients (3.8%). One individual carried the H208Y mutation associated with resistance to pyrophosphate (PP) analogues.
4. Discussion There is still uncertainty in the literature about the extent to which ARV resistant strains are transmitted to drug-native partners. In our study, we demonstrated transmission of phenotypic and genotypic resistance to ZDV (T215Y/F) and 3TC (M184V) in 5% of newly infected persons. The multi-NNRTI resistance mutation, K103N was present in 3.8% of newly infected individuals, and the primary V82A and L90M mutations for PIs in 2.5 and 3.8% of patients, respectively. Multidrug resistant variants were present in 10% of our cohort. These variants were present at seroconversion prior to initiation of any antiretroviral therapy. Our findings are consistent with the study of Boden et al. and Yerly et al. who showed transmission of genotypic resistant variants in 16.3 and 11% of individuals infected between 1995 and 1998 [20,21]. This also
compares with a report by Little et al. showing transmission of genotypic resistance in a cohort infected between 1989 and 1998, oftentimes prior to the introduction of PI and NNRTIs [22]. Taken together, the cumulative findings from these studies are consistent with an increase in transmission of drug-resistant variants with the advent of MDR resistance in treated individuals. We observed MDR in 10% of our cohort. Identified source partners of the newly infected individuals had received many NRTI/ NNRTI/PI treatments and harbored viral variants showing a predominant resistance-associated genotype. The high transmission of NNRTI resistant variants to newly infected persons is somewhat surprising, insofar as all of the source partners had been on NNRTI regimens, including Efavirenz and delaviridine for less than 6 months. It should be noted that the K103N mutation occurs in the drug-binding pocket and confers cross-resistance to all NNRTIs [11]. The prevalence of broad-spectrum resistance to all NRTIs is less common than that reported for ZDV. The M184V mutation associated with resistance to 3TC and Aba was observed in four cases, 5% of our cohort. This is despite the widespread use of 3TC in Canada and a prevalence of the M184V mutation in 58% of the HIV-infected population (n= 93, data not shown). The low transmission of 3TC-resistant HIV-1 variants indirectly supports the lower fitness and replicative capacity of HIV-1 variants with the M184V mutation [26,27]. The phenotypic analytic approach used in this study is laborious, requiring six to eight weeks to generate results. Because viruses are isolated from lectin-stimulated PBMCs, there is selective expansion of lymphotropic viruses. This may explain the low recovery of amplified virus in our cohort, since the viral burst in primary infection may be derived from macrophage-tropic virus. However, phenotypic analysis is still essential to establish the cumulative effect of mutations on drug susceptibility. We observed near
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434 Table 1 Prevalence of resistance mutations associated with resistance to nucleoside (NRTI) and non-nucleoside (NNRTI) reverse trancriptase and protease (PI) inhibitors in newly infected individualsa Case numbers
Protease codon substitutions
Reverse transcriptase (RT) codon substitutions
Number of mutations (n =51, 64%) 1–51 None
None
NRTI only (n = 5, 6.3%) 60 None 68, 77 None 76 None 71 None
I50V; A62V(all NRTIs) M184V F77L (all NRTIs) K219Q
NNRTI only (n = 4, 5%) 52, 53, 65 None 66 None
A98S A98E/D; L100Y
PI only (n = 12, 15%) 54 D30N 58 V82A 59, 62, 79 L33I/V 61, 63, 64 K20R 70, 72, 74, 75 L10I
None None None None None
MDR (n = 8, 10%) 55 L90M 56 57 67 69
L10I L10I None G48V; V82A; L90M
73
L10I; K20R; I54I/V; L90M; G73S
78 80
L10L/R None
M41L; L210W; T215Y; D67E*Y A98S; L210F F116S T69A; A98S; V189I M41L; T215Y; K103N; M184V M41L; D67N; T69N; K70R; L74V; L100I; K103N; M184V; T215I/F; K219Q T69D; K219D K103K/N; T215Y
a Frequently observed PI mutations at codons 36, 63, and 77 that were present in 12.5, 50, and 30% of our cohort, are not included.
433
complete concordance between phenotypic and genotypic analysis in detecting the presence of ZDV, 3TC, and PI resistance and the absence of ddI resistance. Four years after the introduction of PIs and novel NNRTIs, the observation of transmission of drug-resistant variants to newly infected individuals is troubling. Although many PI mutations are natural variants that confer low-level resistance, follow-up analysis should be encouraged to establish whether patients with pre-existing mutations are more prone to treatment failure. Further studies in socio-demographics, behaviors, and clinical data are warranted to better determine predictive factors of transmission of drug-resistance mutations and the resultant ramifications for disease management. Acknowledgements This research was sponsored by MRC of Canada (MT-14738), the FRSQ Quebec (969971.42), NIAIDNIH, USA (RO1 A143271–01). The authors acknowledge the participation of Clinique Goldberg LeBlanc and Rosengren; Jean-Guy Baril, Pierre Cote´, Franc¸ois Laplante, Dominique Tessier, Clinique Quartier Latin; Louise Charest, Harold Dion, Eric Lefebvre, Re´jean Thomas, Benoit Trottier, Sylvie Vezina, Clinique l’Actuel; Richard Lalonde, John MacLeod, Graham Smith, Pierre Cholette, Clinic 1851 and MUHC, Normand Lapointe, Hospital St. Justine, Montreal; and Micheline Fauvel, Laboratoire de Sante´ Publique de Que´bec. We thank Yudong Quan, Daniella Moise, Maureen Oliviera, and Merve Detorio for their technical support. Horacio Salomon is presently affiliated with the Department of Microbiology, School of Medicine, University of Buenos Aires, Argentina.
Fig. 3. The frequency of genotypic substitutions associated with resistance to ARV drugs in our cohort of 80 subjects.
434
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
References [1] Palella FJ, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. New Engl J Med 1998;338:853 – 60. [2] Wainberg MA, Friedland G. Public health implications of antiretroviral therapy and HIV drug resistance. J Am Med Assoc 1999;279:1977– 83. [3] Hirsch MS, Conway B, D’Aquila RT, et al. Antiretroviral drug resistance testing in adults with HIV infection. J Am Med Assoc 1998;279:1984– 91. [4] Schinazi R, Larder B, Mellors J. Mutations in retroviral genes associated in drug resistance. Int Antiviral News 1997;5:129 – 42. [5] Condra J. Virologic and clinical implications of resistance to HIV-1 protease inhibitors. Drug Resistance Updates 1998;1:292 – 9. [6] Larder BA, Kemp SD. Multiple mutations in HIV-1 reverse transcriptase confer high-level resistance to zidovudine (AZT). Science 1989;46:1155–8. [7] Doyon L, Croteau G, Thibeault D, et al. Second locus involved in human immunodeficiency virus type 1 resistance to protease inhibitors. J Virol 1996;70:3763–9. [8] Japour AJ, Welles S, D’Aquila RT, et al. Prevalence and clinical significance of zidovudine resistance mutations in human immunodeficiency virus isolated from patients after long-term zidovudine treatment. J Infect Dis 1995;171:1172–9. [9] Tisdale M, Kemp SD, Parry NR, Larder BA. Rapid in vitro selection of human immunodeficiency virus type 1 resistant to 3%-thiacytidine inhibitors due to a mutation in the YMDD region of reverse transcriptase. Proc Natl Acad Sci USA 1993;90:5653 – 6. [10] Nunberg JH, Schleif WA, Boots EJ, et al. Viral resistance to human immunodeficiency virus type 1-specific pyridinone reverse transcriptase inhibitors. J Virol 1991;65:4887–92. [11] Esnouf R, Ren J, Ross C, et al. Mechanism of inhibition of HIV-1 reverse transcriptase by non-nucleoside inhibitors. Nat. Struct. Biol. 199;2:303–08. [12] Erice A, Mayers DL, Strike DG, et al. Brief report: primary infection with zidovudine-resistant human immunodeficiency virus type 1. New Engl J Med 1993;328:1163–5. [13] Hecht FM, Grant RM, Petropoulos CJ, et al. Sexual transmission of an HIV-1 variant resistant to multiple reverse transcriptase and protease inhibitors. New Engl J Med 1999;339:307 – 11. [14] Angarano G, Monno L, Appice A, et al. Transmission of zidovudine-resistant HIV-1 through heterosexual contacts. AIDS
.
1994;8:1013 – 4. [15] Imrie A, Beveridge A, Genn W, Vizzard J, Cooper DA. Transmission of human immunodeficiency virus type 1 resistant to nevirapine and zidovudine. Sydney Primary HIV Infection Study Group. J Infect Dis 1997;175:1502 – 6. [16] Cohen OJ, Fauci AS. Transmission of multidrug-resistant human immunodeficiency virus — the wakeup call. New Engl J Med 1998;339:341 – 3. [17] Quigg M, Rebus S, France AJ, McMenamin J, Darby G, Leigh Brown AJ. Mutations associated with zidovudine resistance in HIV-1 among recent seroconverters. AIDS 1997;11:835–6. [18] de Ronde A, Schuurman R, Goudsmit J, van den Hoek A, Boucher C. First case of new infection with zidovudine-resistant HIV-1 among prospectively studied intravenous drug users and homosexual men in Amsterdam, The Netherlands. AIDS 1996;10:231 – 2. [19] Mayers DL, Yerly S, Perrin L, et al. Prevalence of AZT-Resistant (AZTR) HIV-1 Persons Seroconverting in Switzerland, Australia, and the United States Between 1988 and 1994. Presented at the 34th ICAAC, Orlando, FL, 1994. abstract A/11. [20] Yerly S, Kaiser L, Race E, Bru J-P, Clavel F, Perrin L. Transmission of antiretroviral-drug-resistant HIV-1 variants. Lancet 1999;354:729 – 33. [21] Boden D, Hurley A, Zhang L, et al. HIV-1 drug resistance in newly infected individuals. J Am Med Assoc 1999;282:1135–41. [22] Little SJ, Daar ES, D’Aquila RT, et al. Reduced antiretroviral drug susceptibility among patients with primary infection. J Am Med Assoc 1999;282:1142 – 9. [23] Salomon H, Belmonte A, Nguyen K, Gu Z, Gelfand M, Wainberg MA. Comparison of cord blood and peripheral blood mononuclear cells as targets for viral isolation and drug sensitivity studies involving human immunodeficiency virus type 1. J Clin Microbiol 1994;32:2000 – 2. [24] Boulerice F, Bour S, Geleziunas R, Lvovitch A, Wainberg MA. High frequency of isolation of defective human immunodeficiency virus type 1 and heterogeneity of viral gene expression in clones of infected U-937 cells. J Virol 1990;64:1745 – 55. [25] Gao Q, Gu Z, Parniak MA, Li X, Wainberg MA. In vitro selection of variants of human immunodeficiency virus type 1 resistant to 3%-azido-3%-deoxythymidine and 2%,3%-dideoxyinosine. J Virol 1992;66:12 – 9. [26] Wainberg MA, Parniak MA. Is HIV resistance to 3TC of clinical benefit? Int Antivir News 1997;5:3 – 4. [27] Bock NK, Nijhuis M, Keulen W, et al. Reduced replication of 3TC-resistant HIV-1 variants in primary cells due to a processivity defect of the reverse transcriptase enzyme. EMBO J 1996;15:4040 – 9.
Resistance to antiretroviral drugs in patients with primary HIV-1 infection Bluma Brenner a, Mark A. Wainberg a,*, Horacio Salomon a, Danielle Rouleau b, Andre´ Dascal a, Bonnie Spira a, Rafik-Pierre Sekaly c, Brian Conway d, Jean-Pierre Routy e, Investigators of the Quebec Primary Infection Study a
McGill Uni6ersity AIDS Centre, Lady Da6is Institute-Jewish General Hospital; 3755 Cote Site-Catherine Road, Montreal, Quebec, Canada H3T 1E2 b The Reseau FRSQ-SIDA, Montreal, Quebec, Canada H3T 1E2 c Institut de Recherches Cliniques, Montreal, Quebec, Canada H3T 1E2 d Viridae Clinical Sciences Inc., Vancou6er, BC, Canada H3T 1E2 e Royal Victoria Hospital, Montreal, Quebec, Canada
Abstract The widespread use of antiretroviral agents (ARVs) and the growing occurrence of HIV strains resistant to these drugs have given rise to serious concerns regarding the transmission of resistant viruses to newly infected persons. Plasma viral RNA from 80 individuals newly infected between 1997 and 1999 was genotyped by automated sequencing to analyze the profile of viruses resistant to nucleoside and non-nucleoside reverse transcriptase inhibitors (NRTIs and NNRTIs) and to protease inhibitors (PIs). The prevalence of mutations that conferred primary resistance to PIs (L10I, D30Y, V82A, L90M) was 15% of the cohort. RT genotypic variants, associated with high-level resistance to ARVs, were observed in 21% of individuals, including NRTI, NNRTI and multidrug (MDR) resistance in 6, 5, and 10% of cases, respectively. The phenotypic susceptibility of viral isolates to ARVs was also assayed and showed transmission of high-level resistance to ZDV, 3TC, and PIs in those individuals with MDR. The transmission of drug-resistant HIV genotypic variants is a serious problem that merits further attention by public health officials, virologists, and clinicians. © 2000 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. Keywords: Antiretroviral therapy; HIV drug resistance/resistance mutations; Reverse transcriptase inhibitors; Protease inhibitors
1. Introduction While potent antiretroviral drugs (ARVs) are of proven benefit in individuals and communities with HIV-1 infection, the evolution of ARV drug resistance is a leading factor in treatment failure [1,2]. Drug resistance arises as a direct result of an error-prone viral reverse transcriptase (RT) causing ongoing HIV mutagenesis. ARV drug pressure amplifies resistant variants that carry resistance-associated mutations in the viral RT and protease (PR) enzymes [2 – 5]. Advances in genotypic analysis have established over 150 changes in the RT and PR regions associated with selective resistance to nucleoside and non-nucleoside * Corresponding author. Tel.: + 1-514-3408260; fax: 1-5143407537. E-mail address: [email protected] (M.A. Wainberg).
RT inhibitors (NRTIs and NNRTIs) and PR inhibitors. An important consideration in ARV resistance, under conditions of drug pressure, is the selective growth advantage of mutated viruses [2]. High-level resistance to indinavir (Ind) and zidovudine (ZDV) take long periods (6 months to years) to develop, requiring the sequential accumulation of multiple mutations [6–8]. In contrast, resistance to nevirapine (Nev) and 3TC develops rapidly, since single mutations confer 500–1000fold increases in resistance [9,10]. In contrast, resistance to didanosine (ddI), zalcitabine (ddC) and stavudine (d4T) emerge slowly since pertinent mutations confer only low-level resistance [2]. Particularly problematic is the emergence of PI (L90M), NRTI (Q151M and F77L) and NNRTI (K103N) single mutations that confer cross-resistance to many ARVs of the same class [2,5,11].
0924-8579/00/$ - $20 © 2000 Elsevier Science B.V. and International Society of Chemotherapy. All rights reserved. PII: S 0 9 2 4 - 8 5 7 9 ( 0 0 ) 0 0 2 7 0 - 3
430
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
Distinct from the obvious question of the relationship between drug resistance and treatment failure, is the problem of transmission of resistant virus among the general HIV-infected population. Identification of HIV resistance in newly infected patients has great clinical and public health significance, especially if available drugs are ineffective against resistant strains present in primary infection. Unless this topic is successfully addressed, the likelihood is that drug-resistant variants of HIV, if fit, will sustain the epidemic and limit the effectiveness of antiviral therapy. Since the initial observation of transmission of ZDVresistant isolates in 1993, many case reports have shown the sexual transmission of ARV-resistant HIV isolates from treated individuals to their drug-native partners [12 – 19]. The frequency of the ZDV resistance mutation, T215Y, in cohorts with primary infection has increased from 1.45 (1988 – 1991) to 7.5 in 1992 and 10.4% in 1993 and 1994 [19]. HIV variants carrying RT mutations at codon-215, (conferring high level ZDV resistance), currently range from 6 to 11% in newly infected cohorts [20 – 22]. However ZDV is unique, insofar as it was the first antiretroviral drug used, often in monotherapy. It is now important to determine how these findings apply to other types of drug-resistant viruses and to evaluate the extent to which transmitted viruses resistant to inhibitors of both RT and PR can cause disease. Two studies have shown the presence of genotypic markers, associated with ARV resistance in 16 – 19% of newly infected cohorts as compared with a study showing genotypic resistance in only 2% of a cohort comprised of patients dating to 1989, when PIs were not in use [20–22]. Multidrug (NNRTI/NRTI/PI) resistance (MDR) was observed in 2 – 4% of cases [20 – 22]. The aim of this investigation was to monitor the prevalence of resistance to ARVs and RT and PR mutations in a large cohort of individuals infected between 1997 and 1999 in the metropolitan area of Montreal, Canada, when universal access to combination chemotherapy was available. Our findings show that despite a high prevalence of resistance-associated mutations in newly infected individuals, these mutations do not confer high-level antiretroviral drug resistance in cell culture.
immune response with reactive EIAs and indeterminant Western blot, and p24 antigenemia with emergent positive HIV serology within 3 months; (c) HIV seroconversion within a 6-month period following negative EIA testing; and/or (d) typical acute retroviral syndrome within the earlier 3 months. The cohort included 80 persons where transmission was established to be sexual (n= 56, 93% males), intravenous (n= 21, 80% male), or unknown (n= 3). CD4 counts and viral loads of the sexual (493 9 29 cells per l and 338 1139110 882 copies per ml) and intravenous (5729 53 cells per l and 183 151 9 71 799 copies per ml) transmission groups were similar. HIV-1 antibody testing was performed at three university sites. Western blot analysis for HIV-1 antibodies and p24 antigen capture assays were performed at the Laboratoire de Sante´ Publique de Que´bec. Plasma viremia was measured using the Roche Amplicor assay (Roche Diagnostics, Mississauga, ON). Flow cytometric analysis determined T-cell subset distributions.
2.2. Phenotypic drug susceptibility testing Drug susceptibility was measured by determining the extent to which ARVs inhibit in vitro HIV replication [23–25]. PBMCs from 62 patients were co-cultured with PHA-pre-stimulated cord blood mononuclear cells (CBMCs) from normal donors. Levels of virus in cocultures were monitored weekly by p24 antigen capture assays until deemed positive after 3–5 weeks (a first p24 value of \ 20 pg/ml followed by a second value of \ 100 pg/ml). High-titer stocks of primary virus isolates were obtained from only 22 patients. These isolates were re-amplified and quantified by RT enzyme assays to generate defined, titrated clinical isolates to minimize inter-inoculum effects [24]. CBMCs infected with patient isolates were then plated in 96-well plates both in the absence and presence of a variety of ARV dosages. After 7 days, RT enzyme assays were used to determine the 50% drug inhibitory concentration (ID50) [23–25]. The observed ID50 values of patient viral isolates were then compared with the ID50 values obtained for HIV-IIIb, the ARV susceptible control [23].
2.3. Sequence analysis 2. Materials and methods
2.1. Study population In January 1997, a clinical research network was established to enrol individuals with clear documentation of recent HIV infection. Enrolment criteria included (a) non-reactive ELISA immunoassays (EIAs) in subjects with p24 antigenemia; (b) evolving humoral
Viral RNA was isolated from plasma samples using the QIAamp viral extraction kit (Qiagen Inc., Chatsworth, California). The TruGene HIV-1 Assay Gene System was used in conjunction with the Open Gene automated DNA sequencer, (Visible Genetics Inc., Toronto, Canada), to sequence the PR and RT regions of HIV-1 cDNA. Testing involved simultaneous clip sequencing of PR and codons 35–244 of RT from amplified cDNA in both the 3% and 5% directions. Se-
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
quences were aligned and compared with a LAV-1 consensus sequence using Visible Genetics Gene Librarian software.
3. Results
3.1. Phenotypic drug susceptibility testing PBMCs from 62 patients with primary HIV infection were co-cultured with PHA-pre-stimulated cord blood mononuclear cells (CBMCs) from normal donors. High-titer viral stocks were obtained from 22 subjects. Titrated viral inocula were used to infect CBMCs, incubated in the presence or absence of ARVs. The drug inhibitory concentrations (ID50) of PBMC-derived isolates were compared with the genotype of cDNA derived from plasma viral RNA. The spectrum of ID50 values for the three NRTIs tested, ZDV, ddI, and 3TC, and the three PIs tested, ritonovir (Rit), Ind, and saquinavir (Saq), are shown in Figs. 1 and 2. Based on numerous studies in our laboratory with clinical isolates, phenotypic resistance was considered to be present when ID50 values exceeded 0.1 mm, 1 mm, 50 mm, and 0.1 mm for ZDV, ddI, 3TC, and all PIs, respectively. These values exceed the laboratory mean ID50 value for HIVIIIb by at least 10-fold. Our findings show transmission of phenotypic resistance to ARVs in 5 of 22 samples (22.7%). Isolates from three (13.6%) of the newly infected individuals showed high level resistance to ZDV and one isolate showed low level ZDV resistance (Fig. 1). These four subjects were the only individuals among the 22 tested who carried either the T215Y or T215F mutations. Similarly, isolates from three subjects (13.6%) carrying the M184V mutation showed high level resistance to 3TC (Fig. 1). ID50 values for ddI for all the clinical isolates from subjects with primary infection were within the range of control values for HIVIIIb (Fig. 1).
431
None of these isolates were from individuals carrying the L74V mutation. Phenotypic cross-resistance to PIs was observed in isolates from three of the subjects, carrying viruses with the L90M mutation, conferring high level PI resistance (Fig. 2). Thus, there was concordance in transmission of resistance to ZDV, 3TC, and PIs in 13.6% of tested clinical isolates, consistent with genotypic data showing the T215Y/F, M184V, and L90M substitutions in plasma viral RNA.
3.2. Sequence analysis of protease and RT regions Viral cDNA derived from plasma viral RNA was sequenced to identify protease and RT mutations from 80 persons. In all, 60 patients (75%) carried at least one mutation in regions associated with PI resistance. However, many mutations may represent polymorphisms in clinical viral isolates rather than true resistance-conferring mutations. These include L63P/S/A/N, M36l, and V77l codon substitutions, associated with low-level resistance to nelfinavir (Nelf), Ind, and/or Rit, present in 50, 30 and 12.5% of our cohort, respectively. Excluding these three codons, protease substitutions associated with resistance to PIs were observed in 18 subjects, 22.5% of our cohort [21]. Similarly, 17 persons (21.5%) carried mutations in regions of RT associated with resistance to NRTIs and NNRTIs. The spectrum of RT and PI resistance mutations in our cohort of 80 subjects with primary infection is shown in Table 1. Eight patients harbored viruses with RT and protease mutations, conferring high level resistance to each of PIs, NRTIs and NNRTIs, i.e. 10% of cases had evidence of multidrug resistance. In three of these patients (Table 1, patients 69, 73, and 80), source partners were identified who carried identical genotypic variants. Although NRTIs were not included in our initial phenotypic drug resistance screening, subsequent testing of these clinical isolates showed low-level pheno-
Fig. 1. Phenotypic drug testing for ZDV, 3TC and ddI resistance in 22 subjects. The ID50 values for each ARV in each subject are indicated. The mean ID50 values observed for drug-sensitive HIVIIIb is indicated by arrows. The laboratory thresholds established for low and high drug resistance are also shown.
432
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
Fig. 2. Phenotypic drug testing for Rit, Ind and Saq resistance in 22 subjects. The ID50 values for each ARV in each subject are indicated. The mean ID50 values observed for drug-sensitive HIVIIIb is indicated by arrows. The laboratory thresholds established for low and high drug resistance are also shown.
typic resistance to Nev (data not shown). The primary L90M and V82A/F/T/S, associated with PI cross-resistance, were present at frequencies of 3.8 and 4%, respectively. The profile of mutations, in relation to resistance to ARVs, is shown in Fig. 3. Mutations conferring resistance to ZDV were present in 8.8% of subjects with 5% of subjects carrying the T215Y or the T215F mutations. The M184V mutation conferring resistance to cytosine analogues, including high-level resistance to 3TC and low-level resistance to ddC and Abacavir (Aba) was present in 5% of our cohort. The K103N mutation, conferring high level cross-resistance to NNRTIs, appeared in 3.8% of subjects. Mutations associated with cross-resistance to different classes of NRTIs, including F116S, F77L, A62V, were observed in three patients (3.8%). One individual carried the H208Y mutation associated with resistance to pyrophosphate (PP) analogues.
4. Discussion There is still uncertainty in the literature about the extent to which ARV resistant strains are transmitted to drug-native partners. In our study, we demonstrated transmission of phenotypic and genotypic resistance to ZDV (T215Y/F) and 3TC (M184V) in 5% of newly infected persons. The multi-NNRTI resistance mutation, K103N was present in 3.8% of newly infected individuals, and the primary V82A and L90M mutations for PIs in 2.5 and 3.8% of patients, respectively. Multidrug resistant variants were present in 10% of our cohort. These variants were present at seroconversion prior to initiation of any antiretroviral therapy. Our findings are consistent with the study of Boden et al. and Yerly et al. who showed transmission of genotypic resistant variants in 16.3 and 11% of individuals infected between 1995 and 1998 [20,21]. This also
compares with a report by Little et al. showing transmission of genotypic resistance in a cohort infected between 1989 and 1998, oftentimes prior to the introduction of PI and NNRTIs [22]. Taken together, the cumulative findings from these studies are consistent with an increase in transmission of drug-resistant variants with the advent of MDR resistance in treated individuals. We observed MDR in 10% of our cohort. Identified source partners of the newly infected individuals had received many NRTI/ NNRTI/PI treatments and harbored viral variants showing a predominant resistance-associated genotype. The high transmission of NNRTI resistant variants to newly infected persons is somewhat surprising, insofar as all of the source partners had been on NNRTI regimens, including Efavirenz and delaviridine for less than 6 months. It should be noted that the K103N mutation occurs in the drug-binding pocket and confers cross-resistance to all NNRTIs [11]. The prevalence of broad-spectrum resistance to all NRTIs is less common than that reported for ZDV. The M184V mutation associated with resistance to 3TC and Aba was observed in four cases, 5% of our cohort. This is despite the widespread use of 3TC in Canada and a prevalence of the M184V mutation in 58% of the HIV-infected population (n= 93, data not shown). The low transmission of 3TC-resistant HIV-1 variants indirectly supports the lower fitness and replicative capacity of HIV-1 variants with the M184V mutation [26,27]. The phenotypic analytic approach used in this study is laborious, requiring six to eight weeks to generate results. Because viruses are isolated from lectin-stimulated PBMCs, there is selective expansion of lymphotropic viruses. This may explain the low recovery of amplified virus in our cohort, since the viral burst in primary infection may be derived from macrophage-tropic virus. However, phenotypic analysis is still essential to establish the cumulative effect of mutations on drug susceptibility. We observed near
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434 Table 1 Prevalence of resistance mutations associated with resistance to nucleoside (NRTI) and non-nucleoside (NNRTI) reverse trancriptase and protease (PI) inhibitors in newly infected individualsa Case numbers
Protease codon substitutions
Reverse transcriptase (RT) codon substitutions
Number of mutations (n =51, 64%) 1–51 None
None
NRTI only (n = 5, 6.3%) 60 None 68, 77 None 76 None 71 None
I50V; A62V(all NRTIs) M184V F77L (all NRTIs) K219Q
NNRTI only (n = 4, 5%) 52, 53, 65 None 66 None
A98S A98E/D; L100Y
PI only (n = 12, 15%) 54 D30N 58 V82A 59, 62, 79 L33I/V 61, 63, 64 K20R 70, 72, 74, 75 L10I
None None None None None
MDR (n = 8, 10%) 55 L90M 56 57 67 69
L10I L10I None G48V; V82A; L90M
73
L10I; K20R; I54I/V; L90M; G73S
78 80
L10L/R None
M41L; L210W; T215Y; D67E*Y A98S; L210F F116S T69A; A98S; V189I M41L; T215Y; K103N; M184V M41L; D67N; T69N; K70R; L74V; L100I; K103N; M184V; T215I/F; K219Q T69D; K219D K103K/N; T215Y
a Frequently observed PI mutations at codons 36, 63, and 77 that were present in 12.5, 50, and 30% of our cohort, are not included.
433
complete concordance between phenotypic and genotypic analysis in detecting the presence of ZDV, 3TC, and PI resistance and the absence of ddI resistance. Four years after the introduction of PIs and novel NNRTIs, the observation of transmission of drug-resistant variants to newly infected individuals is troubling. Although many PI mutations are natural variants that confer low-level resistance, follow-up analysis should be encouraged to establish whether patients with pre-existing mutations are more prone to treatment failure. Further studies in socio-demographics, behaviors, and clinical data are warranted to better determine predictive factors of transmission of drug-resistance mutations and the resultant ramifications for disease management. Acknowledgements This research was sponsored by MRC of Canada (MT-14738), the FRSQ Quebec (969971.42), NIAIDNIH, USA (RO1 A143271–01). The authors acknowledge the participation of Clinique Goldberg LeBlanc and Rosengren; Jean-Guy Baril, Pierre Cote´, Franc¸ois Laplante, Dominique Tessier, Clinique Quartier Latin; Louise Charest, Harold Dion, Eric Lefebvre, Re´jean Thomas, Benoit Trottier, Sylvie Vezina, Clinique l’Actuel; Richard Lalonde, John MacLeod, Graham Smith, Pierre Cholette, Clinic 1851 and MUHC, Normand Lapointe, Hospital St. Justine, Montreal; and Micheline Fauvel, Laboratoire de Sante´ Publique de Que´bec. We thank Yudong Quan, Daniella Moise, Maureen Oliviera, and Merve Detorio for their technical support. Horacio Salomon is presently affiliated with the Department of Microbiology, School of Medicine, University of Buenos Aires, Argentina.
Fig. 3. The frequency of genotypic substitutions associated with resistance to ARV drugs in our cohort of 80 subjects.
434
B. Brenner et al. / International Journal of Antimicrobial Agents 16 (2000) 429–434
References [1] Palella FJ, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. New Engl J Med 1998;338:853 – 60. [2] Wainberg MA, Friedland G. Public health implications of antiretroviral therapy and HIV drug resistance. J Am Med Assoc 1999;279:1977– 83. [3] Hirsch MS, Conway B, D’Aquila RT, et al. Antiretroviral drug resistance testing in adults with HIV infection. J Am Med Assoc 1998;279:1984– 91. [4] Schinazi R, Larder B, Mellors J. Mutations in retroviral genes associated in drug resistance. Int Antiviral News 1997;5:129 – 42. [5] Condra J. Virologic and clinical implications of resistance to HIV-1 protease inhibitors. Drug Resistance Updates 1998;1:292 – 9. [6] Larder BA, Kemp SD. Multiple mutations in HIV-1 reverse transcriptase confer high-level resistance to zidovudine (AZT). Science 1989;46:1155–8. [7] Doyon L, Croteau G, Thibeault D, et al. Second locus involved in human immunodeficiency virus type 1 resistance to protease inhibitors. J Virol 1996;70:3763–9. [8] Japour AJ, Welles S, D’Aquila RT, et al. Prevalence and clinical significance of zidovudine resistance mutations in human immunodeficiency virus isolated from patients after long-term zidovudine treatment. J Infect Dis 1995;171:1172–9. [9] Tisdale M, Kemp SD, Parry NR, Larder BA. Rapid in vitro selection of human immunodeficiency virus type 1 resistant to 3%-thiacytidine inhibitors due to a mutation in the YMDD region of reverse transcriptase. Proc Natl Acad Sci USA 1993;90:5653 – 6. [10] Nunberg JH, Schleif WA, Boots EJ, et al. Viral resistance to human immunodeficiency virus type 1-specific pyridinone reverse transcriptase inhibitors. J Virol 1991;65:4887–92. [11] Esnouf R, Ren J, Ross C, et al. Mechanism of inhibition of HIV-1 reverse transcriptase by non-nucleoside inhibitors. Nat. Struct. Biol. 199;2:303–08. [12] Erice A, Mayers DL, Strike DG, et al. Brief report: primary infection with zidovudine-resistant human immunodeficiency virus type 1. New Engl J Med 1993;328:1163–5. [13] Hecht FM, Grant RM, Petropoulos CJ, et al. Sexual transmission of an HIV-1 variant resistant to multiple reverse transcriptase and protease inhibitors. New Engl J Med 1999;339:307 – 11. [14] Angarano G, Monno L, Appice A, et al. Transmission of zidovudine-resistant HIV-1 through heterosexual contacts. AIDS
.
1994;8:1013 – 4. [15] Imrie A, Beveridge A, Genn W, Vizzard J, Cooper DA. Transmission of human immunodeficiency virus type 1 resistant to nevirapine and zidovudine. Sydney Primary HIV Infection Study Group. J Infect Dis 1997;175:1502 – 6. [16] Cohen OJ, Fauci AS. Transmission of multidrug-resistant human immunodeficiency virus — the wakeup call. New Engl J Med 1998;339:341 – 3. [17] Quigg M, Rebus S, France AJ, McMenamin J, Darby G, Leigh Brown AJ. Mutations associated with zidovudine resistance in HIV-1 among recent seroconverters. AIDS 1997;11:835–6. [18] de Ronde A, Schuurman R, Goudsmit J, van den Hoek A, Boucher C. First case of new infection with zidovudine-resistant HIV-1 among prospectively studied intravenous drug users and homosexual men in Amsterdam, The Netherlands. AIDS 1996;10:231 – 2. [19] Mayers DL, Yerly S, Perrin L, et al. Prevalence of AZT-Resistant (AZTR) HIV-1 Persons Seroconverting in Switzerland, Australia, and the United States Between 1988 and 1994. Presented at the 34th ICAAC, Orlando, FL, 1994. abstract A/11. [20] Yerly S, Kaiser L, Race E, Bru J-P, Clavel F, Perrin L. Transmission of antiretroviral-drug-resistant HIV-1 variants. Lancet 1999;354:729 – 33. [21] Boden D, Hurley A, Zhang L, et al. HIV-1 drug resistance in newly infected individuals. J Am Med Assoc 1999;282:1135–41. [22] Little SJ, Daar ES, D’Aquila RT, et al. Reduced antiretroviral drug susceptibility among patients with primary infection. J Am Med Assoc 1999;282:1142 – 9. [23] Salomon H, Belmonte A, Nguyen K, Gu Z, Gelfand M, Wainberg MA. Comparison of cord blood and peripheral blood mononuclear cells as targets for viral isolation and drug sensitivity studies involving human immunodeficiency virus type 1. J Clin Microbiol 1994;32:2000 – 2. [24] Boulerice F, Bour S, Geleziunas R, Lvovitch A, Wainberg MA. High frequency of isolation of defective human immunodeficiency virus type 1 and heterogeneity of viral gene expression in clones of infected U-937 cells. J Virol 1990;64:1745 – 55. [25] Gao Q, Gu Z, Parniak MA, Li X, Wainberg MA. In vitro selection of variants of human immunodeficiency virus type 1 resistant to 3%-azido-3%-deoxythymidine and 2%,3%-dideoxyinosine. J Virol 1992;66:12 – 9. [26] Wainberg MA, Parniak MA. Is HIV resistance to 3TC of clinical benefit? Int Antivir News 1997;5:3 – 4. [27] Bock NK, Nijhuis M, Keulen W, et al. Reduced replication of 3TC-resistant HIV-1 variants in primary cells due to a processivity defect of the reverse transcriptase enzyme. EMBO J 1996;15:4040 – 9.