Share this article on:

Antiretroviral resistance in viral isolates from HIV-1-transmitting mothers and their infants

Bauer, Greta Ra,*; Colgrove, Robert Cb; LaRussa, Philip Sd; Pitt, Janed,‡; Welles, Seth Lcfor the Women and Infants Transmission Study Team, USA

doi: 10.1097/01.aids.0000242816.80462.81
Basic Science: Concise Communication

Objective: To characterize concordance of resistance mutations to antiretroviral drugs (ART) in mother–infant pairs.

Design: Case series of HIV-transmitting mothers and infants in the Women and Infants Transmission Study, where delivery occurred between April 1994 and December 1999.

Methods: Reverse transcriptase and protease genes were sequenced in stored viral isolates from 32 mother–infant pairs. Mutations were coded as ‘pure mutants’ where only mutant virus was detected or as ‘mixtures’ where a mixed mutant/wild-type population was identified. ART resistance mutations were compared for concordance between mothers and their infants.

Results: Maternal mutations associated with resistance to nucleoside reverse transcriptase inhibitor (NRTI) and minor protease inhibitor (PI) drugs were typically concordant with that of infant, while those associated with non-nucleoside reverse transcriptase inhibitors (NNRTI) and major PI drugs were not. Of five NRTI-associated maternal mutations observed, three pure mutants corresponded with mutant in the infant, while two wild-type-predominant mixtures corresponded with infant wild type. The only NNRTI-associated mutation observed, K103N, was not transmitted, nor were the two major PI-associated mutations, L90M and V82I/V. Transmission of minor PI-associated mutations was consistent with the sole observed or dominant variant for 20 of 21 mutations.

Conclusions: For NRTI- and minor PI-associated mutations, transmission was consistent with relative quantity of variants in maternal virus. However, where NNRTI- and major PI-associated mutations were present in three cases, they were not transmitted, even where only mutant virus was detectable in maternal isolates. This is consistent with evidence of loss of transmission with resistance to NNRTI and PI drugs.

From the aUniversity of New Hampshire, Durham, New Hampshire, USA

bHarvard Medical School, USA

cBoston University, Boston, Massachusetts, USA

dColumbia University, New York, USA.

*Present address: The University of Western Ontario, London, Ontario, Canada.

See the Appendix for study members.


Received 9 March, 2005

Accepted 6 June, 2006

Correspondence to Dr G. R. Bauer, Epidemiology & Biostatistics, University of Western Ontario, K201 Kresge Bldg, London, ON N6A 5C1, Canada. E-mail:

Back to Top | Article Outline


Resistance to antiretroviral drugs (ART) during pregnancy is now common. From 1995 to 2001, the prevalence of resistance in HIV-infected New York City women at first prenatal visit was 53% for ART-experienced and 10% for ART-naive women, despite exclusion of minor mutations to protease inhibitors (PI) [1]. Resistance at conception may represent a direct response to treatment or primary infection with resistant strains that had been increasing over time [2–4]. Additional mutations can emerge during pregnancy, even as a result of short-term triple therapy for perinatal prophylaxis [5]. At delivery, virus from an estimated 17–25% of women will harbor zidovudine resistance mutations [6,7] and an unknown proportion resistance to other ART.

While in-vitro studies have demonstrated comparable or reduced replicative fitness of virus containing various ART resistance mutations [8–12], it is not clear whether this translates into a reduction of in-vivo replicative fitness. Moreover, it is unclear whether this in any way corresponds with vertical transmission fitness, the in-vivo transmission efficacy of a virus relative to other strains. As the relative contributions of viral properties to vertical transmissibility are unknown, transmission fitness cannot currently be measured except through actual transmissions [13].

Mother–infant pairs are useful for in-vivo studies of transmission fitness for several reasons. First, viral donor and recipient are known. Second, perinatal transmission in non-breastfeeding populations is most likely to occur in the intrapartum period, providing a probable window of transmission, and maternal blood specimens are frequently collected at this time. Finally, where a maternal viral population is a mixture of mutant and wild type at a particular codon, the situation provides a naturally occurring in-vivo competition experiment [13].

Data on concordance of resistance mutations in mother–infant pairs are limited. Our previous research documented selective transmission of zidovudine resistance mutations in monotherapy-treated mothers and their untreated infants. Where maternal virus consisted of wild-type/mutant mixed populations, only a single variant was evident in the infant, and mutations other than T215Y were not transmitted [14]. Palumbo et al. [7] examined 22 neonatal infant isolates and found resistance mutations in two; both infant genotypes differed from maternal. While four of six maternal virus mutations were observed in infant virus, there were also three new mutations between the two infants. Variations in treatment regimens between mothers and neonates were not reported, nor was the coexistence of wild-type subpopulations. Based on the limited information available, continued characterization of ART resistance in pregnant women is warranted.

Back to Top | Article Outline


Subjects were mothers in the Women and Infants Transmission Study (WITS) who delivered liveborn singleton HIV-infected infants from April 1994, after the announcement of the successful trial of prophylactic zidovudine [15], to December 1999. The WITS is an ongoing cohort that has enrolled and followed HIV-1-infected pregnant women and their children at six US sites. Informed consent was obtained from all research subjects at 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. A total of 57 mothers transmitted HIV to their infants during this time period. Prior to June 1998, transmission was determined to have occurred if at least two separate cultures of infant peripheral blood mononuclear cells (PBMC) were positive. Two negative cultures and no positive cultures were sufficient for classification as not infected, where both cultures occurred after 1 month of age, and one of the two after 6 months of age [16]. After June 1998, transmission was indicated by a positive polymerase chain reaction (PCR) for HIV-1 DNA, confirmed with either a PCR for HIV-1 RNA or PBMC culture. Three negative PCR tests occurring after 1 month of age and no positive PCR were sufficient for classification as not infected.

Data for these mothers and their infants were obtained from the WITS database, maintained by Clinical Trials and Surveys, Inc. (Baltimore, Maryland, USA). Data included maternal viral loads and ART treatment histories of mothers and infants.

Stored maternal virus isolates from tissue culture supernatants were obtained for the time of delivery. Where these were unavailable, supernatants from a third trimester or immediate postpartum clinic visit were obtained. Stored infant virus isolates were obtained from the first positive post-delivery culture.

HIV-1 RNA was isolated, reverse transcribed, and amplified from supernatants using the HIV Genotyping System kit v.2 from Applied Biosystems (Foster City, California, USA) under recommended conditions. When amplification was insufficient for sequencing, a second nested PCR was employed. Automated software basecalling and direct visual inspection were employed to determine the presence of multiple variants at each codon. Where these were identified, predominant and minority variants were identified by relative peak height.

Resistance mutations were classified based on recommendations from the International AIDS Society-USA Drug Resistance Mutations Group [17]. L63P mutations were excluded from analysis for two reasons. First, it is associated only with use of lopinavir and ritonavir, and none of our subjects used this combination. Second, L63P occurs at high frequency in the absence of ART treatment and so would contribute disproportionately to combined measures.

All mutations were coded as ‘pure mutant’ where a pure population of mutant virus was detected and as mutant/wild-type ‘mixtures’ where a chromatographic mixture of mutant and wild-type subpopulations was clearly identifiable in sequencing reads from both template strands. Mixtures are presented with two amino acids following the codon number, the first indicating the predominant variant. For example, M36M/I would indicate that methionine (M) was the predominant variant at codon 36, and isoleucine (I) the minority variant.

Back to Top | Article Outline


Stored culture supernatants from at or near the time of delivery were available for 36 of 57 HIV-transmitting mothers. Virus was successfully isolated and sequenced from 32 of these supernatants. Of these 32 mothers, 28 had used ART during pregnancy and four had declined treatment.

From these 32 pairs, virus from six mothers (19%) had any resistance mutations associated with nucleoside reverse transcriptase inhibitor (NRTI), non-nucleoside reverse transcriptase inhibitor (NNRTI) or major protease inhibitor (PI) drugs, with multiple such mutations observed in two mothers (Table 1). Of the six mothers, four also had minor PI-associated mutations, as did 11 mothers with otherwise wild-type virus. When minor PI-associated mutations were included, 17 of 32 mothers (53%) had ART resistance mutations at delivery.

Table 1

Table 1

Table 1

Table 1

HIV-infected infants had resistance profiles largely resembling that of their mothers. Since some of the 17 mothers with ART resistance had multiple mutations, 29 total mutations were observed. Maternal–infant concordance by drug class was as follows. Of five maternal NRTI-associated mutations, three pure mutants corresponded with mutant in the infant (in pairs 8 and 24), while two mixtures corresponded with wild type in the infant (pairs 11 and 17); in both mixtures, wild type was the predominant variant. One NNRTI-associated mutation (K103N) and two major PI-associated mutations (L90M and V82I/V) were observed among mothers (pairs 27 and 15). None of these was transmitted. However, 17 of 21 maternal minor PI-associated resistance mutations (81%) were detected in the corresponding infant isolates; of 16 pure mutations in maternal virus, 15 were identical in infant virus, while for five mutant/wild-type mixtures only the quantitatively predominant variant was detected in the infant. In two of the latter, the predominant variant was mutant; in three it was wild type.

In this analysis, a mutant/wild-type mixed population was detected for a total of eight maternal virus mutations; in each case only one variant was observed in the infant. In five of the eight mixed populations, the transmitted sequence was both wild type and predominant; in two, the transmitted sequence was mutant and predominant; and in only one was it wild type and minority. There were no instances of mutant/minority being transmitted.

There was evidence of emergence of additional resistance mutations in infants. Of 15 mothers with wild-type virus at all resistance-associated codons, 11 had infants with fully concordant virus. The four remaining infants had a single resistance mutation each: M36M/I at 127 days (pair 6), T215T/Y at 42 days (pair 10), L74V/L at 170 days (pair 21), and K70R at 31 days (pair 14).

Back to Top | Article Outline


We have characterized resistance mutations found in virus from 32 HIV-transmitting mothers and their infants. While viral genotypes were not always identical, those of the infants closely matched the mother's virus.

Our previous work examined only zidovudine-associated resistance mutations. The present work extends these analyses to resistance mutations currently seen in the multidrug era. Previously, we had observed that the mutant variant solely was transmitted when maternal virus was mixed at reverse transcriptase codons 210 and 215 but wild type was transmitted in three cases where mutant/wild-type mixtures were present at codon 70 [14]. In the present series, we observed two additional HIV-transmitting mothers with K70K/R mixtures; again both infants evidenced only the wild type despite continued drug pressure with infant zidovudine treatment. The predominance of wild-type sequence in both maternal mixtures suggests transmission may be a result of quantitative advantage, rather than a reduction of in-vivo transmission fitness with K70R mutation.

One patient had a K103N mutation indicative of NNRTI resistance. Since this mutation was observed in a specimen collected 15 days postpartum and use of nevirapine in labor may result in rapid emergence of the K103N mutation [18], we conducted a chart review to exclude an unrecorded dose. A review of treatment dates confirmed that the mother received no NNRTI drug during pregnancy or delivery, but she had received efavirenz as part of a research protocol prior to pregnancy. The observance of the K103N mutant is then of note, as it was likely present at delivery but not transmitted.

Neither major PI-associated mutation observed (L90M or V82I/V) was transmitted. This is consistent with in-vitro evidence that mutations at these codons result in loss of replicative fitness [19,20], and this may extend to transmission fitness. However, as only two such mutations were present it may also be purely chance.

Most minor PI-associated mutations were transmitted. Since only one mother used a PI during pregnancy, these likely represent naturally occurring polymorphisms, or primary maternal infection from a source treated with PI drugs.

It appears, consistent with our previous report, that where maternal virus is a mixture of mutant and wild type at a specific codon, only one variant is transmitted. For seven of eight maternal mixed viral populations, the variant transmitted was quantitatively predominant. Therefore, quantity may ultimately be more important than replicative quality, and this hypothesis should be subjected to further research. The exception was the PI-associated major mutation discussed above.

It is important to note that while ART-associated mutations may not be preferentially transmitted, they are transmitted. Mothers in this study did transmit multiple ART-associated mutations to their infants, and such transmission could result in reduced treatment options and altered progression of disease in these infants.

This study has limitations. While the sample represents the largest such series of paired mother–infant virus sequences presented to date, it is not large enough in itself to draw firm conclusions from the data. Nevertheless, the descriptive results suggest a need for additional research, and the findings should be subject to further analysis in other study populations with larger sample sizes to permit statistical testing and evaluation.

In addition, it is possible in some cases that mutations detected do not accurately represent those present in patient viral populations, or that concordant mutations result from reasons other than transmission. Since most vertical transmission in non-breastfeeding populations occurs during labor and delivery, it is frequently not possible to isolate HIV-1 from infants at delivery. In this analysis, infant specimens were obtained a median of 39 days after delivery (range, 0–180). In this post-076 era, all infants were treated with ART from birth, allowing the possibility that virus populations detected may have evolved since transmission. Indeed, several new mutations were present in infants, including the only three infant mutant/wild-type mixtures observed. While these new mutations were consistent with infant treatment, they could also represent transmissions of minor maternal variants. A second limitation is that the use of viral isolates obtained from culture allows the possibility of brief continued viral evolution in the absence of drug pressure. Therefore, it is possible that some mutations present in patients' PBMC could have been ‘lost’ in culture and not detected in this analysis. While it is also possible that new mutations could emerge in culture, the sequence results obtained for 18 patients with the methods described above were compared with paired specimens obtained from direct PCR amplification of virus from patient plasma obtained at the same clinic visit (results not shown) and indicate that this is unlikely. Paired sequences demonstrated strong but imperfect concordance, with all differences in a single direction: mutations in virus sequenced from plasma not being detectable in virus obtained from drug-free culture. Therefore, it is likely that where our results err, they err in failure to detect existing mutations rather than in detection of mutations not existing in patient's virus.

The increasing use of ART therapy worldwide will undoubtedly impose selection pressures for HIV-1 viral populations toward continued evolution of resistance. As the prevalence of resistance mutations increases, so more attention should be paid to the transmission characteristics of resistance-associated variants, in vivo as well as in vitro.

Sponsorship: This study is supported by the US National Institutes of Health (NIAID/NIH) grant RO1 AI39144-01. Women and Infants Transmission Study is supported by 1 U01 A150274-01, University of Puerto Rico, San Juan, Puerto Rico (U01 AI 34858), Boston/Worcester Site, Boston, Massachusetts (U01 AI 34856), Columbia Presbyterian Hospital, New York (U01 AI 34842), State University of New York, Brooklyn, New York (HD-3–6117 and RO-1–IID-25714), University of Illinois at Chicago, Chicago, Illinois (U01 AI 34841), Baylor College of Medicine, Houston, Texas (U01 AI 34840), Clinical Trials & Surveys Corp., Baltimore, Maryland (N01 AI 85339).

Note: Dr. Pitt is deceased.

Back to Top | Article Outline


1. Pitt J, Garretson A, LaRussa P, Colgrove R, Welles S, WITS Study Team. Genotypic resistance of HIV-1 isolates to protease inhibitors, nonnucleoside reverse transcriptase inhibitors, and nucleoside reverse transcriptase inhibitors from anitretroviral experienced and naive pregnant women, 1995–2001. 10th Conference on Retroviruses and Opportunistic Infections. Boston, February, 2003 [poster 633].
2. Little SJ, Holte S, Routy J-P, Daar ES, Markowitz M, Collier AC, et al. Antiretroviral-drug resistance among patients recently infected with HIV. N Engl J Med 2002; 347:385–394.
3. Grant RM, Heckt FM, Warmerdam M, Liu L, Liegler T, Petropoulos CJ, et al. Time trends in primary HIV-1 drug resistance among recently infected persons. JAMA 2002; 288:181–188.
4. Simon V, Vanderhoeven J, Hurley A, Ramratnam B, Louie M, Dawson K, et al. Evolving patterns of HIV-1 resistance to antiretroviral agents in newly infected individuals. AIDS 2002; 16:1511–1519.
5. Lyons FE, Coughlan S, Byrne CM, Hopkins SM, Hall WW, Mulcahy FM. Emergence of antiretroviral resistance in HIV-positive women receiving combination antiretroviral therapy in pregnancy. AIDS 2005; 19:63–67.
6. Welles SL, Pitt J, Colgrove R, McIntosh K, Chung PH, Colson A, for the Women and Infants Transmission Study Group. HIV-1 genotypic zidovudine drug resistance and the risk of maternal–infant transmission in the Women and Infants Transmission Study. AIDS 2000; 14:263–271.
7. Palumbo P, Holland B, Dobbs T, Pau CP, Luo CC, Abrams EJ, for the Perinatal AIDS Collaborative Transmission Study. 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 2001; 184:1120–1126.
8. Larder BA, Kemp SD, Harrigan PR. Potential mechanism for sustained antiretroviral efficacy of AZT–3TC combination therapy. Science 1995; 269:696–699.
9. Back NK, Nijhuis M, Keulen W, Boucher CA, Oude Essink BO, van Kuilenburg AB, 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; 15:4040–4049.
10. Harrigan PR, Bloor S, Larder BA. Relative replicative fitness of zidovudine-resistant human immunodeficiency virus type 1 isolates in vitro. J Virol 1998; 72:3773–3778.
11. Maeda Y, Venzon DJ, Mitsuya H. Altered drug sensitivity, fitness, and evolution of human immunodeficiency virus type 1 with pol gene mutations conferring multi-dideoxynucleoside resistance. J Infect Dis 1998; 70:2146–2153.
12. Nicastri E, Sarmati L, d'Ettorre G, Palmisano L, Parisi SG, Uccella I, et al. Replication capacity, biological phenotype, and drug resistance of HIV strains isolated from patients failing antiretroviral therapy. J Med Virol 2003; 69:1–6.
13. Leigh Brown AJ, Frost SD, Mathews WC, Dawson K, Hellmann NS, Daar ES, et al. Transmission fitness of drug-resistant human immunodeficiency virus and the prevalence of resistance in the antiretroviral-treated population. J Infect Dis 2003; 187:683–686.
14. Colgrove RC, Pitt J, Chung PH, Welles SL, Japour AJ. Selective vertical transmission of HIV-1 antiretroviral resistance mutations. AIDS 1998; 12:2281–2288.
15. Connor EM, Sperling RS, Gelber R, Kiselev P, Scott G, O'Sullivan MJ, for the Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. Reduction of maternal–infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994; 331:1173–1180.
16. McIntosh K, Pitt J, Brambilla D, Carroll S, Diaz C, Handelsman E, et al. Blood culture in the first 6 months of life for the diagnosis of vertically transmitted human immunodeficiency virus infection. J Infect Dis 1994; 170:996–1000.
17. D'Aquila RT, Schapiro JM, Brun-Vézinet F, Clotet B, Conway B, Demeter LM, et al. Drug resistance mutations in HIV-1. Topics HIV Med 2003; 11.
18. Jackson JB, Becker-Pergola G, Guay L, Musoke P, Mracna M, Fowler MG, et al. Identification of the K103N mutation in Ugandan women receiving nevirapine to prevent HIV-1 vertical transmission. AIDS 2000; 14:FT111–FT115.
19. Martinez-Picado J, Savara AV, Sutton L, D'Aquila RT. Replicative fitness of protease inhibitor-resistant mutants of human immunodeficiency virus type 1. J Virol 1999; 73:3744–3752.
Back to Top | Article Outline


Women and Infants Transmission Study scientific leadership core: Kenneth Rich (principal investigator) and Delmyra Turpin (study coordinator). Other members include Clemente Diaz and Edna Pacheco-Acosta (University of Puerto Rico, San Juan, Puerto Rico); Ruth Tuomala, Ellen Cooper, and Donna Mesthene (Boston/Worcester Site, Boston); Philip LaRussa and Alice Higgins (Columbia Presbyterian Hospital, New York); Sheldon Landesman, Edward Handelsman, and Gail Moroso (State University of New York, Brooklyn, New York); Kenneth Rich and Delmyra Turpin (University of Illinois at Chicago, Chicago); William Shearer, Susan Pacheco, and Norma Cooper (Baylor College of Medicine, Houston); Joana Rosario (National Institutes of Allergy and Infectious Disease, Bethesda); Robert Nugent, (National Institute of Child Health and Human Development, Bethesda); Vincent Smeriglio and Katherine Davenny (National Institute on Drug Abuse, Rockville); and Bruce Thompson (Clinical Trials & Surveys Corp., Baltimore).

20. Martinez-Picado J, Savara AV, Shi L, Sutton L, D'Aquila RT. Fitness of human immunodeficiency virus type 1 protease inhibitor-selected single mutants. Virology 2000; 275:318–322.

HIV; anti-HIV drugs; viral RNA; vertical transmission; drug resistance

© 2006 Lippincott Williams & Wilkins, Inc.