Welles, Seth L ScD, PhD*; Bauer, Greta R PhD†; LaRussa, Philip S MD‡§; Colgrove, Robert C MD∥; Pitt, Jane MD‡§; for the Women and Infants Transmission Study
Antiretroviral therapy (ART) serves a dual purpose for the pregnant HIV-infected patient: maternal treatment and prophylaxis for mother-to-child transmission. For prophylactic purposes, some form of ART is almost universally initiated by late pregnancy regardless of maternal indicators for treatment. When pregnancy occurs in a woman already on antiretroviral treatment, treatment is continued.
The response to ART may be tempered by ART resistance, reducing the efficacy of viral suppression and possibly reducing prevention of transmission. The prevalence of genotypic resistance among pregnant women on ART has been estimated to range from 2.3% to 25%, depending on whether women were treatment naive or experienced and the treatment regimen used to prevent transmission.1-6 These prevalences probably represent resistance acquired through primary infection or superinfection and through selection of resistant virus as a result of treatment.
Although previous research has identified significant levels of genotypic ART resistance over the course of treatment during pregnancy, little is known about the prevalence of ART resistance among treatment-naive HIV-positive women. Recent reports from Europe7,8 and North America9 indicate a possible downturn in the prevalence of primary infection with resistant strains in the past few years. These studies used heterogeneous patient populations of acutely infected individuals, however, and did not focus on pregnant women. Moreover, previous ART resistance research has been conducted cross-sectionally or has reported period prevalences of resistance over short periods.
We offer new findings that chart time-related changes in ART genotypic resistance prevalence among pregnant women newly enrolled in the Women and Infants Transmission Study (WITS) New York City sites over 1 decade. Large proportions of HIV-infected pregnant women at these sites are ART naive at the onset of pregnancy, offering an opportunity to identify prevalences of HIV ART resistance not acquired through treatment. As such, we determined the prevalence of ART resistance mutations in ART-naive and ART-experienced New York City area pregnant women at enrollment in the WITS from 1991 to early 2001. We present estimates of the prevalence of ART resistance mutations at the onset of pregnancy, time trends, and predictors of genotypic ART resistance.
Selection of Isolates
The WITS is a prospective cohort study of HIV-1 mother-to-child transmission and the natural history of HIV-1 infection in pregnant women and their infants. From all women enrolled at New York City State University of New York (SUNY) Downstate and Columbia-Presbyterian Medical Center sites from 1991 to early 2001, culture supernatants were selected using the following criteria: (1) availability of peripheral blood mononuclear cell (PBMC) coculture supernatant specimens and (2) availability of information on the history of ART use. The sample of pregnant HIV-positive women in the WITS is reflective of infected women in the catchment area seeking prenatal care.
Of 709 women enrolled in the WITS New York City sites from January 1991 to February 2001, women were selected for inclusion in this study if they had a positive HIV viral coculture at study enrollment, there was adequate information about their antiretroviral drug history, and stored coculture isolates were available for their enrollment visit. Using these criteria, 503 women had positive cultures, and 397 of these women had an evaluable antiretroviral drug history. Of the 397 women with positive cultures and evaluable antiretroviral drug histories, 300 isolates (75%) were available for sequencing, from which >95% yielded an evaluable sequence.
Extraction of Viral RNA and Sequencing
Supernatants from primary viral isolates from study entry were used for sequencing using ABI/Perkin Elmer kits (Foster City, CA). To obtain virus isolates from coculture with fresh phytohemagglutinin (PHA)-stimulated donor cells,10 maternal PBMCs were isolated from heparinized blood using Ficoll-Hypaque centrifugation. Supernatants from these cocultures were used as sources of virus isolates for extraction and sequencing. HIV RNA was isolated, reverse transcribed, and amplified from tissue culture supernatants using the HIV Genotyping System kit version 2 from Applied Biosystems (Foster City, CA) under recommended conditions. When first-round amplification was insufficient for sequencing, a second nested polymerase chain reaction (PCR) was used. Big-dye terminator dideoxy sequencing reactions were carried out using reagents provided in the HIV Genotyping System kit. Products of the sequencing reaction were then resolved electrophoretically on an ABI 7700 Automated Sequencer and analyzed using Sequence Analysis 3.2 software from Applied Biosystems. Nucleotide sequences from these reactions were derived using the basecalling program in the Sequence Analysis software and aligned as reverse complement using SeqPup program in the Sequence Analysis software.
Variables and Analysis
Mutations in the HIV-1 reverse transcriptase and protease regions were scored according to the International AIDS Society (IAS)-USA panel11 for mutations associated with nucleoside reverse transcriptase inhibitors (NRTIs) and nonnucleoside reverse transcriptase inhibitors (NNRTIs) and for major mutations to protease inhibitors (PIs). In addition, isolates were scored as being resistant to stavudine if V75T mutations were present.12,13 When a mixed population of mutant and wild-type virus was detected for a particular codon, the mixture was reported as mutant.
Demographics, history of illicit drug use (heroin, cocaine, or marijuana), and standard indicators of stage of HIV-1 infection (CD4 cell counts and plasma HIV-1 RNA levels) are presented for each group of ART-experienced and ART-naive women. Similarly, prevalences of mutations to NRTIs, NNRTIs, and PIs are reported for each group of ART-experienced and ART-naive women. Differences in distributions or proportions of background characteristics between these groups of women were tested using appropriate parametric or nonparametric methods. Statistically significant results are identified by level of significance. P < 0.05 was used as the criterion to assess statistical significance. For contingency table analyses of ART treatment history (being ART experienced or ART naive) and having ART mutations, prevalence ratios and associated 95% confidence intervals (CIs) are given as estimates of the increased risk for having ART resistance mutations attributable to ART treatment history.
For analyses of changes in the prevalence of ART resistance mutations over time, enrollment years were divided into 4 periods of time based on the introduction and established use of PIs. The period from 1991 to 1994 preceded the availability of PIs, 1995 to 1996 was the period of access to PIs through clinical trial protocols, 1997 to 1998 represents the period of US Food and Drug Administration (FDA) approval and introduction of PIs as components of multidrug ART regimens, and 1999 to early 2001 represents a period when the use of PIs was widespread and established as part of combination therapy. These periods were specified before analysis. Trends for changes in prevalence over these 4 periods were evaluated using Mantel-Haenszel χ2 tests for trend in 2 × 4 contingency tables.
Finally, variables shown by crude analysis to be associated with having any ART resistance mutation (significant at the level of P < 0.05 or by having a strong magnitude of association irrespective of significance testing) were entered into unconditional multiple logistic regression models to evaluate the adjusted magnitudes of association of independent and dependent variables. Prevalence ratios and associated 95% CIs are presented for variables associated with having ART genotypic resistance at the time of study enrollment.
Of 300 women included in this study, 39.0% were enrolled at the Columbia University site (Washington Heights and surrounding neighborhoods) and 61.0% were enrolled at the SUNY Downstate Brooklyn site for the WITS. Overall, women included for analysis tended to be young (median age = 28 years), mostly black (60%), of low income (>64% reported a household income <$10,000), and with a high school education or less (77.0%). When comparing ART-naive with ART-experienced women, there were no differences in most demographics and HIV disease indicators. ART-naive women tended to be slightly younger (27 vs. 29 years; P < 0.01) and less likely to be using illicit drugs at study enrollment (16% vs. 28% in use of heroin, cocaine, or marijuana; P < 0.05) when compared with ART-experienced women (Table 1).
When screened for ART resistance, 55 (18.3% [95% CI: 14.0% to 22.7%]) of all women had major PI mutations or NRTI- or NNRTI-associated resistance mutations. Patterns of ART resistance mutations evolved over time and are consistent with the introduction of ART agents and changes in combination therapy prescribing patterns (Tables 2, 3). For ART-naive mothers (n = 128; see Table 2), only zidovudine (ZDV)-associated mutations were observed from 1991 to 1996. Beginning in 1997, ART resistance mutations to NRTIs other than ZDV were seen: the V118I mutation to lamivudine (3TC) and the M184V mutation that is cross-resistant to 3TC, zalcitabine (ddC), and abacavir. Subsequently, mutations to all 3 classes of ART were observed in mothers beginning in 2000. Of 4 women with ART resistance mutations from 2000 to early 2001, 3 had mutations to NRTIs (with 2 having the M184V mutation), 2 had NNRTI mutations (with both having 1 or more mutations to nevirapine or efavirenz, including the K103N, V108I, or Y181C mutation), and 2 had virus with major PI mutations (V82T and V82T/A). In total, 11 of 128 (8.6% [95% CI: 3.7% to 13.4%]) of ART-naive mothers had reverse transcriptase or PI mutations consistent with resistance.
Similar patterns of ART mutations over time were observed for ART drug-experienced women (n = 172; see Table 3). Mutations to only the NRTI class of ART were seen among women who were earliest to be enrolled in the WITS, with all 14 women having resistance to ZDV or 3TC during 1991 through 1996. Although 1 woman had PI-resistant virus at enrollment in 1997, the number of women with genotypic resistance to PIs increased substantially in 1998, with 4 of 11 women with mutations having major PI resistance mutations; PI-associated mutations are seen in these ART-experienced women through early 2001. Similarly, resistance mutations to NNRTIs were first seen in this group of ART-experienced women in 1997 and were particularly seen in 1999 to early 2001.
Summary data are presented that chart the prevalence of ART resistance mutations by period of enrollment for the overall sample and for the subsets of ART-naive and ART-experienced women (Fig. 1). In contrast to resistance mutations to NRTIs, which were of significant prevalence and showed no clear trend over time among ART-naive and ART-experienced women, mutations to NNRTIs and PIs were absent in the early years and appeared in 1997 and later. Among ART-naive women, resistance mutations to NNRTIs were absent through 1998, with a prevalence of 7.1% (95% CI: 2.4% to 16.7%) during 1999 to early 2001. The same trend was seen among ART-experienced women, with 5.9% (95% CI: 0% to 12.3%) of women having NNRTI resistance mutations in 1997 to 1998 and 21.9% (95% CI: 7.6% to 36.2%) having mutations in 1999 to early 2001 (P = 0.08, test for trend). Similar trends in the prevalence of PI-associated major mutations were seen in both groups of women: no resistance mutations to PIs were seen in ART-naive women until 1999 to early 2001, when 7.1% (95% CI: 2.4% to 16.7%) of women had this class of resistance mutation. Similarly, there were no major PI resistance mutations seen among ART-experienced women until 1997 to 1998, when 9.8% (95% CI: 1.6% to 18.0%) of women had such mutations. By 1999 to early 2001, 15.6% (95% CI: 3.0% to 28.2%) of ART-experienced women had PI-associated resistance mutations (P = 0.001, test for trend).
When considered across the 11-year period, ART-experienced women had greater levels of genotypic resistance across all classes of ART when compared with ART-naive women: 25.6% (95% CI: 19.1% to 32.1%) of drug-experienced women had at least 1 mutation to NRTIs, NNRTIs, or PIs across time compared with 8.6% (95% CI: 3.7% to 13.4%) of women who were ART-naive, reflecting a nearly 3-fold increase in the prevalence of resistance mutations to any class of ART (prevalence ratio = 2.98, 95% CI: 1.60 to 5.53; P = 0.0003). When broken down by class of ART, drug-experienced women had between a 2.8-fold (NRTI-associated prevalence ratio = 2.75, 95% CI: 1.42 to 5.33; P = 0.001) and a 3.7-fold (PI-associated prevalence ratio = 3.72, 95% CI: 0.83 to 16.69; P = 0.06) increase in the prevalence of resistance mutations when compared with ART-naive women. Similarly, women who were ART experienced had a 3.7-fold increase in the prevalence of NNRTI resistance when compared with ART-naive women (NNRTI-associated prevalence ratio = 3.72, 95% CI: 0.83 to 16.69).
When women with genotypic ART resistance were compared on demographics and other background characteristics with those who did not have resistance mutations, only age at enrollment and CD4 cell count and percentage differed between mothers with genotypic resistance and mothers with wild-type virus (data not shown). Women with ART resistance mutations tended to be older (median age = 31 years) compared with women not having resistance mutations (median age = 28 years; P = 0.004). In addition, there were differences in HIV disease indicators, because women with ART resistance mutations tended to have lower CD4 cell counts (293 vs. 372 cells/mm3; P < 0.03) and CD4 percentages (19% vs. 25%; P < 0.002) when compared with women without HIV-1 ART resistance mutations.
Final predictive models were constructed to evaluate the adjusted magnitudes of association of factors shown in unadjusted analysis to be associated with having ART genotypic resistance. Thus, having any ART resistance mutation (major PI mutation or any NRTI or NNRTI mutation) was associated with having ever received ART (odds ratio [OR] = 3.41, 95% CI: 1.63 to 7.14; P = 0.001), lower CD4 percentage (OR = 1.05 or a 5.0% reduction of the likelihood of having any ART resistance mutations for each 1% decrease in CD4 percentage; 95% CI: 0.71 to >1.0; P = 0.06), and period of time when compared with the earliest period of observation of 1991 to 1994 (1995-1996: OR = 0.82, 95% CI: 0.29 to 2.30; 1997-1998: OR = 2.39, 95% CI: 0.94 to 6.06; and 1999 to early 2001: OR = 3.52, 95% CI: 1.32 to 9.40).
Our findings suggest that increasing rates of HIV-1 ART resistance among HIV-positive pregnant women could become a major consideration in selecting maternal ART treatment and prevention of mother-to-child transmission of HIV-1 in US population centers. Women who have received ART at some point in the course of their infection before pregnancy have especially high rates of genotypic resistance, and this rise in levels of resistance has occurred since the late 1990s. Although the prevalence of ART resistance was 15% among ART-experienced women from 1991 to 1994, alarmingly high rates of ART resistance (34%) are seen among these drug-experienced women enrolled during 1999 to early 2001. Moreover, resistance patterns included mutations to multiple classes of ART, with 19% of ART-experienced women having ART resistance mutations to NNRTIs and 16% of all ART-experienced women being resistant to PIs in the later years of follow-up.
These findings are of great concern because not only could multiclass resistance to multiple ART potentially reduce the efficacy of treatment for prevention of mother-to-child transmission but the wide spectrum of ART resistance could limit treatment options for women as they progress through their infection. Although it is clear that CD4 cell count and history of ART treatment independently predict resistance in individual women, there may be other factors that drive the temporal trends for increases in ART resistance. In addition, it may well be that the potency of current treatments is not only capable of reducing rates of resistance but of offering effective treatment in face of preexisting resistance mutations. A particular emphasis should be placed on educating patients about how even short discontinuations of ART, or treatment holidays, can increase the likelihood of developing resistance.14-19
Of equal concern are the rising rates of ART resistance among drug-naive women because such resistance indicates acquisition by means of primary infection with resistant strains. Among our study population of mostly women of color, who had a low education level and income, overall rates of ART genotypic resistance among ART-naive women rose from 6% in the early 1990s to 18% by 1999 to early 2001. Examination of resistance by ART class indicates that levels of resistance to NRTIs, NNRTIs, and PIs have all increased over time in untreated women, presumably by means of infection from sexual or drug-using partners. Although our numbers are relatively sparse for women enrolling in the WITS for each period and caution needs to be exercised in making inferences, statistically significant results for tests of trend lend support to the finding that rates of resistance to NNRTIs and PIs are increasing over time among ART-naive women, with 7.4% having genotypic resistance to both classes of ART.
Additionally, we provide multiple logistic regression models that identify factors to predict having genotypic ART resistance among our study population; women who have a lower CD4 percentage or who are treatment experienced should be tested for multiclass ART resistance. Our findings suggest that each of these factors contributes risk independently for having resistance mutations. Rapidly rising rates of drug resistance in ART-naive and ART-experienced pregnant women may necessitate resistance testing to guide treatment to achieve suppression of the mother's virus and prevent transmission and to provide effective treatment of infection in women after delivery.
One potential limitation of our study is that by including culture supernatants from women who were HIV culture-positive, we were not including specimens in which HIV replication was effectively suppressed. In turn, culture negativity could be related to virus isolates being sensitive to ART or wild-type virus. We tested this hypothesis by classifying HIV culture results by mothers' history of ART use, because women who were treatment naive at enrollment should be likely to have wild-type virus. Results indicated that there were no differences in the treatment histories between women with HIV-positive (64% ART naive, 15% ART experienced, and 21% without ART treatment data) versus HIV-negative (60% ART naive, 16% ART experienced, and 24% without ART treatment data) cultures. Therefore, based on comparable treatment history data, it is unlikely that our estimates of ART resistance are biased.
Our present project expands findings of the few earlier studies that identified the presence and rates of ART resistance mutations among pregnant women in the United States: few recent studies by US investigators3,4,6,20-23 and scientists working internationally24-26 have focused on pregnant or nonpregnant HIV-positive women. Although many of these studies focused on resistance to ZDV, NRTIs,4,5,20,23,25,26 or other single classes of drugs20 or on measured rates of resistance in samples of women within short spans of time,3,6,23,26 the present findings represent rates of ART resistance for multiple classes of drugs over a broad sweep of time-1991 to early 2001-by history of treatment among pregnant women. Additionally, our results support a more general trend of increases of ART resistance in HIV high-risk populations; our results parallel increases of ART resistance over time that have been observed among recently or newly infected gay or bisexual men.27,28 Our results add important new information concerning significant rates of ART resistance among mostly ART-naive HIV-positive women enrolling in programs to prevent transmission to their offspring. Additionally, we were able to chart the changing rates over time.
Informed consent was obtained from all research subjects at the time of enrollment. All protocols for this study were approved by Institutional Review Boards at Boston University, Columbia University, Beth Israel Deaconess Medical Center, and the University of Minnesota, where Drs. Bauer and Welles were located before their current affiliations. Dr. Pitt is deceased.
WITS Principal Investigators, Study Coordinators, Program Officers, and financial support include the following: Clemente Diaz and Edna Pacheco-Acosta (University of Puerto Rico, San Juan, PR; U01 AI 34858); Ruth Tuomala, Ellen Cooper, and Donna Mesthene (Boston/Worcester Site, Boston, MA; U01 AI 34856); Philip LaRussa and Alice Higgins (Columbia Presbyterian Hospital, New York, NY; U01 AI 34842); Sheldon Landesman, Edward Handelsman, and Gail Moroso (SUNY, Brooklyn, NY; HD-3-6117 and RO-1-IID-25714); Kenneth Rich and Delmyra Turpin (University of Illinois at Chicago, Chicago, IL; U01 AI 34841); William Shearer, Susan Pacheco, and Norma Cooper (Baylor College of Medicine, Houston, TX; U01 AI 34840); Joana Rosario (National Institute of Allergy and Infectious Diseases, Bethesda, MD); Robert Nugent, (National Institute of Child Health and Human Development, Bethesda, MD); Vincent Smeriglio and Katherine Davenny (National Institute on Drug Abuse, Rockville, MD); and Bruce Thompson (Clinical Trials and Surveys Corporation, Baltimore, MD; N01 AI 85339). The Scientific Leadership Core comprises Kenneth Rich (Principal Investigator) and Delmyra Turpin (Study Coordinator) (1 U01 A150274-01).
1. Eastman PS, Shapiro DE, Coombs RW, et al. Maternal viral genotypic zidovudine resistance and infrequent failure of zidovudine therapy to prevent perinatal transmission of human immunodeficiency virus type 1 in pediatric AIDS Clinical Trials Group Protocol 076. J Infect Dis
2. Kully C, Yerly S, Erb P, et al. Codon 215 mutations in human immunodeficiency virus-infected pregnant women. Swiss Collaborative ‘HIV and Pregnancy’ Study. J Infect Dis
3. Cunningham CK, Chaix ML, Rekacewicz C, et al. Development of resistance mutations in women receiving standard antiretroviral therapy who received intrapartum nevirapine to prevent perinatal human immunodeficiency virus type 1 transmission: a substudy of pediatric AIDS clinical trials group protocol 316. J Infect Dis
4. Palumbo P, Holland B, Dobbs T, et al. Antiretroviral resistance mutations among pregnant human immunodeficiency virus type 1-infected women and their newborns in the United States: vertical transmission and clades. J Infect Dis
5. Welles SL, Pitt J, Colgrove R, et al. HIV-1 genotypic zidovudine drug resistance and the risk of maternal-infant transmission in the Women and Infants Transmission Study. The Women and Infants Transmission Study Group. AIDS
6. Shah SS, Crane M, Monaghan K, et al. Genotypic resistance testing in HIV-infected pregnant women in an urban setting. Int J STD AIDS
7. Bezemer D, Jurriaans S, Prins M, et al. Declining trend in transmission of drug-resistant HIV-1 in Amsterdam. AIDS
8. Chaix ML, Descamps D, Harzic M, et al. Stable prevalence of genotypic drug resistance mutations but increase in non-B virus among patients with primary HIV-1 infection in France. AIDS
9. Turner D, Brenner B, Routy JP, et al. Diminished representation of HIV-1 variants containing select drug resistance-conferring mutations in primary HIV-1 infection. J Acquir Immune Defic Syndr
10. Hollinger FB, Bremer JW, Myers LE, et al. Standardization of sensitive human immunodeficiency virus coculture procedures and establishment of a multicenter quality assurance program for the AIDS Clinical Trials Group. The NIH/NIAID/DAIDS/ACTG Virology Laboratories. J Clin Microbiol
11. Johnson VA, Brun-Vezinet F, Clotet B, et al. Update of the drug resistance mutations in HIV-1: 2005. Top HIV Med
12. Pellegrin I, Segondy M, Garrigue I, et al. Phenotypic resistance pattern of HIV-1 isolates with zidovudine and/or multidrug resistance mutations after didanosine-stavudine combination therapy. J Acquir Immune Defic Syndr
13. Coakley E, Gillis J, Hammer SM. Phenotypic and genotypic resistance patterns of HIV-1 isolates derived from individuals treated with didanosine and stavudine. AIDS
. 2000;14 (Suppl):F9-F15.
14. Metzner KJ, Bonhoeffer S, Fischer M, et al. Emergence of minor populations of human immunodeficiency virus type 1 carrying the M184V and L90M mutations in subjects undergoing structured treatment interruptions. J Infect Dis
15. Lawrence J, Mayers DL, Hullsiek KH, et al. Structured treatment interruption in patients with multidrug-resistant human immunodeficiency virus. N Engl J Med
16. King MS, Brun SC, Kempf DJ. Relationship between adherence and the development of resistance in antiretroviral-naive, HIV-1-infected patients receiving lopinavir/ritonavir or nelfinavir. J Infect Dis
17. Harrigan PR, Hogg RS, Dong WW, et al. Predictors of HIV drug-resistance mutations in a large antiretroviral-naive cohort initiating triple antiretroviral therapy. J Infect Dis
18. Roge BT, Barfod TS, Kirk O, et al. Resistance profiles and adherence at primary virological failure in three different highly active antiretroviral therapy regimens: analysis of failure rates in a randomized study. HIV Med
19. Richard N, Juntilla M, Abraha A, et al. High prevalence of antiretroviral resistance in treated Ugandans infected with non-subtype B human immunodeficiency virus type 1. AIDS Res Hum Retroviruses
20. Juenther SN, Williamson C, Ristig MB, et al. Nonnucleoside reverse transcriptase inhibitor resistance among antiretroviral-naive HIV-positive pregnant women. J Acquir Immune Defic Syndr
21. Weinstock HS, Zaidi I, Heneine W, et al. The epidemiology of antiretroviral drug resistance among drug-naive HIV-1-infected persons in 10 US cities. J Infect Dis
22. Bardeguez AD, Shapiro DE, Mofenson LM, et al. Effect of cessation of zidovudine prophylaxis to reduce vertical transmission on maternal HIV disease progression and survival. J Acquir Immune Defic Syndr
23. Frenkel LM, Wagner LE II, Demeter LM, et al. Effects of zidovudine use during pregnancy on resistance and vertical transmission of human immunodeficiency virus type 1. Clin Infect Dis
24. Servais J, Lambert C, Karita E, et al. HIV type 1 pol gene diversity and archived nevirapine resistance mutation in pregnant women in Rwanda. AIDS Res Hum Retroviruses
25. Larbalestier N, Mullen J, O'Shea S, et al. Drug resistance is uncommon in pregnant women with low viral loads taking zidovudine monotherapy to prevent perinatal HIV transmission. AIDS
26. Mandelbrot L, Landreau-Mascaro A, Rekacewicz C, et al. Lamivudine-zidovudine combination for prevention of maternal-infant transmission of HIV-1. JAMA
27. Little SJ, Holte S, Routy JP, et al. Antiretroviral-drug resistance among patients recently infected with HIV. N Engl J Med
28. Shet A, Berry L, Mohri H, et al. Tracking the prevalence of transmitted antiretroviral drug-resistant HIV-1: a decade of experience. J Acquir Immune Defic Syndr
© 2007 Lippincott Williams & Wilkins, Inc.