Emergence of Antiretroviral Resistance in HIV-Positive Women Receiving Combination Antiretroviral Therapy in Pregnancy


Fiona E. Lyons; Suzie Coughlan; Christina M. Byrne; Susan M. Hopkins; William W. Hall; Fiona M. Mulcahy Authors and Disclosures

Abstract and Introduction Abstract

Background: Antenatal antiretroviral therapy is integral to preventing vertical transmission of HIV-1. The impact of temporary triple antiretroviral therapy in pregnancy on the emergence of antiretroviral resistance has not been studied. Objective: To determine the impact of temporary triple antiretroviral therapy in pregnancy on emergence of antiretroviral resistance. Methods: Pregnant HIV-1 infected women with a pre-treatment CD4 cell count > 300 × 106 /l initiated triple antiretroviral therapy in the third trimester and discontinued postpartum. Genotypic resistance testing was performed after antiretroviral cessation and on pretreatment samples when postpartum samples showed primary mutations. Results: In a cohort of 50 women who initiated antiretroviral therapy in pregnancy, 39 (78%) had postpartum HIV-1 nucleotide sequences available for analysis: 35 of these (90%) were previously antiretroviral naive. Seven primary mutations, V106A (one), Y181C (two), G190A (one), K101E (one), M184V (one), T215S (one) were detected in five (13%) women. All five were on regimens that included nevirapine and all were antiretroviral therapy naive prior to the index pregnancy. Four had no mutations detected pretreatment (one did not have a pretreatment analysis available; viral load 83 copies/ml). The median duration of antiretroviral exposure was 70 days. Conclusion: Emergence of genotypic resistance is significant in this cohort of pregnant women. All mutations detected were in those that took nevirapine-containing regimens. The clinical implications of these mutations are unknown. Introduction International guidelines recommend combination maternal antiretroviral therapy to reduce mother-to-child transmission of HIV-1[1,2] (F. Lyons, K. Butler, S. Coulter-Smith and F. Mulcahy, unpublished data). Combination therapy is suggested for its greater efficacy in preventing mother-to-child transmission[3] and for concerns for antiretroviral resistance with therapy with one or two drugs.[4,5] In Ireland, women commencing antiretroviral therapy in pregnancy who do not require such therapy for their own health discontinue therapy postpartum. The impact of this strategy on the emergence of genotypic resistance and future antiretroviral therapy response is unknown. Emergence of genotypic resistance may also pose a risk for mother-to-child transmission of resistant virus.


The Irish HIV in Pregnancy Guidelines recommend that HIV-infected women be offered triple antiretroviral therapy in pregnancy regardless of immune status (F. Lyons, K. Butler, S. Coulter-Smith and F. Mulcahy, unpublished data). Women with a pretreatment CD4 cell count of > 300 × 106 /l initiate antiretroviral therapy as close as possible to 28 weeks of gestation and discontinue after birth. Genotypic resistance testing is performed approximately 6 weeks following treatment cessation. This study included all women that discontinued triple antiretroviral therapy and returned for genotypic resistance testing. Antiretroviral therapy was dispensed from the HIV clinic and attendance records were maintained. Non-attenders were contacted and another appointment scheduled. Before antiretroviral therapy was initiated, women were counselled about therapy adherence, with reinforcement at each visit. Women were asked to contact the clinic if they experienced medication difficulties and not to stop therapy beforehand. Women were seen every 2 weeks for the first month of therapy and every 4 weeks thereafter. Because of the different pharmacokinetics, women taking regimens that included nevirapine were asked to stop nevirapine immediately postpartum and continue the remainder of the regimen for 5 days. The importance of staggered therapy cessation was highlighted in communication with obstetric colleagues for women taking nevirapine-containing regimens. Postpartum women were asked to return to the clinic within 6 weeks of treatment discontinuation for genotypic resistance testing. Postpartum plasma samples were obtained for HIV-1 viral load measurement and samples > 500 copies/ml were included.

Virological Parametres

Plasma HIV-1 RNA was quantified using VERSANT HIV-1 bDNA 3.0 (Bayer, Berkeley, California, USA) or Amplicor HIV-1 Monitor v1.5 (Roche Molecular Systems, Branchburg, New Jersey, USA). Genotypic resistance testing (Bayer TruGene HIV-1 Genotyping Kit) was performed after cessation of antiretroviral therapy and on pretreatment samples when postpartum samples showed primary mutations. The CS-1 primer set (Bayer) was used to amplify HIV-1 non-B subtypes when necessary. The protease gene and codons 1-246 of the reverse transcriptase (RT) gene were sequenced and each resistance-associated codon examined. HIV-1 subtype was determined using the web-based BLAST subtyping algorithm. Mutations were classified according to the IAS/USA drug resistance mutations group[6] and Guidelines v6.0 (Bayer). Demographic information, previous antiretroviral therapy exposure, immunological and virological parameters were recorded. Statistical analysis was performed using SPSS, version 11 (SPSS, Chicago, Illinois, USA).


Fifty women were eligible for inclusion, 46 (92%) from sub-Saharan Africa. The median age was 27 years (range, 16-32). The median pretreatment CD4 cell count was 480 × 106 /l (range, 300-1083) and HIV viral load was 2689 copies/ml (range, 50-34753). Four women (8%) took antiretroviral therapy in a previous pregnancy (one zidovudine, three a combined lamivudine/zidovudine tablet). From this cohort, 39 (78%) RT and 36 (72%) protease sequences were generated. The protease sequences generated were in women in whom RT sequences were obtained. These samples were included in the analysis. The remaining 11 women (22%) were excluded, nine because the postpartum viral load was < 500 copies/ml and two because pol was not amplifiable. Of the 39 women for whom postpartum sequences were obtained, 28 had initiated zidovudine, lamivudine and nevirapine; 10 had initiated zidovudine, lamivudine and nelfinavir and one initiated didanosine, zidovudine and nevirapine (zidovudine and lamivudine in a previous pregnancy with the M184V mutation postpartum). No changes in nucleoside reverse transcriptase inhibitor drug (NRTI) were made. Eight women switched from nevirapine to nelfinavir: one because of rash, two because of rash and abnormal liver enzymes and five electively because of concerns for nevirapine toxicity in pregnancy.[7] This interchange had no virological or immunological consequences. The median time on antiretroviral therapy was 70 days (range, 3-114). The median viral load reduction was 1.76 log10 copies/ml (range, 0-2.74). The viral load was < 1000 copies/ml in 31 of 32 (97%) with available results at 36 weeks and the other was 12 788 copies/ml. Seven women (18%) had no viral load from 36 weeks of gestation: three had a very late presentation; three delivered before the final appointment and one was a non-attender. The median time from antiretroviral therapy cessation to genotypic resistance testing was 42 days (interquartile range, 57-33). Table 1 shows clade type. Seven primary mutations were detected in the RT segment in five (13%) of the postpartum sequences from the 39 women ( Table 2 ). All five were antiretroviral therapy naive before this pregnancy. In four women, pretreatment samples did not demonstrate any relevant mutations. In one woman, pretreatment sequencing was unavailable (viral load < 500 copies/ml). Of the 35 protease sequences generated, six (17%) demonstrated M36I [subtypes C (one), B (two), F2 (one), G (one) and B/G (one)], two demonstrated L10I (5.7%) [subtypes G (one), B (one)] and one demonstrated K20R (2.8%) (subtype C). All women were protease inhibitor naive. Those with no mutations postpartum did not differ from those that did with respect to pretreatment CD4 cell count, pretreatment viral load, duration of antiretroviral therapy, predelivery viral load or time from treatment cessation to genotypic resistance testing.


Emergence of significant mutations in women taking temporary antiretroviral therapy in pregnancy may jeopardize future maternal HIV care. Furthermore, in the event that mother-to-child transmission occurs, resistant virus may be transmitted. This study identified five women (13% of the cohort) with primary RT mutations following temporary triple antiretroviral therapy. None of these mutations was detectable pretreatment. We had postulated that these women would be low-risk for emergence of genotypic resistance as they are a group with good immunological parameters and low median pretreatment viral load. Statistically, those who developed resistance did not differ from those who did not when comparing pretreatment immunological and virological parameters, duration of therapy and virological response to therapy. Those with postpartum RT mutations were on the same regimen (zidovudine 300 mg twice daily, lamivudine 150 mg twice daily and nevirapine 200 mg once daily for the first 14 days of therapy and 200 mg twice daily thereafter). In four of the five women (90%), the primary RT mutations conferred resistance to non-nucleoside reverse transcriptase inhibitor (NNRTI) drugs; one woman had two such associated mutations. Nevirapine has a low genetic barrier for emergence of resistance and, therefore, nevirapine monotherapy is not advised.[8] Nevirapine is metabolized primarily by the CYP3A family of the cytochrome P450 system. Additionally it induces this group of enzymes, effectively inducing its own metabolism. This autoinduction decreases the terminal-phase plasma half-life from approximately 45 h after a single dose to approximately 25-30 h after multiple dosing.[8] Importantly, the steady-state pharmacokinetics of nevirapine do not appear to be altered in pregnancy.[9] NRTI drugs have shorter half-lives. These pharmacokinetic differences may predispose to a period of nevirapine monotherapy when combination antiretroviral therapy is stopped simultaneously. To avoid this in our study, women continued taking NRTI’s for 5 days following nevirapine discontinuation postpartum. Recent work demonstrated that efavirenz levels may persist for up to 2 weeks following antiretroviral therapy cessation, suggesting that 2 weeks of NRTI use may be necessary to avoid efavirenz monotherapy and selection of NNRTI resistance.[10] However, a more recent study of nevirapine decay after cessation did not demonstrate any resistance in a cohort of nine patients that continued NRTI use for 5 days.[11] All women with protease segment mutations were protease inhibitor naive. The M36I mutation was identified in 17% of the sequences generated. This has previously been described as a polymorphism with high prevalence rates in non-B HIV clades[12] and may predispose to failure of protease inhibitor regimens through emergence of the L90M mutation.[13] The L10I and K20R accessory mutations may have similar implications for protease inhibitor response. In recognition of this, the Irish guidelines for HIV in pregnancy now recommend that all pregnant women have pretreatment resistance testing (F. Lyons, K. Butler, S. Coulter-Smith and F. Mulcahy, unpublished data). This study may underestimate the number of mutations present following temporary antiretroviral therapy in pregnancy. The range of time to performance of postpartum resistance testing was wide, with a median of 42 days and one woman not returning until 198 days postpartum. The longer the time to performance of postpartum resistance testing, the greater the likelihood that the virus will have reverted to wild type. Furthermore, the profile of mutations detected may change in the days and weeks after antiretroviral therapy cessation with selection and fading of mutations at different time points.[14] Consequently, genotypic resistance testing at more than one time point may be more fruitful. There are a number of limitations to the study. The small sample size may contribute to the inability to detect differences between those that developed mutations and those that did not. Adherence to therapy was not formally assessed. Previous authors have identified poor adherence rates with antiretroviral therapy in pregnancy, citing that teenagers and minority groups had poorer adherence rates than white women and those on antiretroviral therapy before pregnancy.[15] In our cohort, 46 women (92%) came from sub-Saharan Africa and all those with postpartum mutations came from sub-Saharan Africa. Furthermore, all those with mutations were diagnosed with HIV infection through antenatal screening in the index pregnancy. Nonetheless, at each clinic visit no adherence issues were identified. The absence of genotypic resistance data on all women pretreatment is a further limitation as significant mutations may have been present before treatment. The majority of the women came from sub-Saharan Africa and were, therefore, less likely to have acquired resistant virus than those acquiring infection in countries with more widespread antiretroviral use. However, resistance may evolve spontaneously and there are conflicting data on the prevalence of naturally occurring mutations within the non-B subtypes of HIV-1. One study found no mutations in a cohort of pregnant women with clade C infection[16] while other workers found more frequent genetic variation at the NNRTI resistance-associated positions in antiretroviral drug-naive individuals with C subtypes versus B subtypes.[17] The impact of these sequence changes on subsequent treatment response is unknown but they may alter frequency and pattern of drug resistance mutations selected under NNRTI pressure. Genotypic resistance testing has limitations; it has poor sensitivity and detection of mutations is dependent on the mutation being present in 25% of the prevailing viral population. More sensitive methods may prove valuable in the future.[18] In conclusion, this study identified a significant number of primary mutations in the RT gene in a cohort of women taking nevirapine-containing regimens in pregnancy. Mutations were not detected in women taking nelfinavir-containing regimens, but this may be because of the small sample size. This study suggests that the strategy of using triple antiretroviral therapy in pregnancy may not protect drugs with known low-genetic barriers such as nevirapine and the clinical implications of this remain to be seen. Larger studies of the temporary use of triple antiretroviral therapy in pregnancy with drugs with higher genetic barriers than nevirapine are warranted. http://www.medscape.com/

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s