Nevirapine resistance – the cautionary tales continue

Polly Clayden, HIV i-Base

…following labour

Two posters from Susan Eshleman’s group reported further analyses of nevirapine (NVP, Viramune) resistance from the HIVNET 012 trial in Uganda in which it was famously demonstrated that single dose NVP given as prophylaxis to a mother in labour followed by a single dose to the infant can reduce mother to child transmission. However following analyses of NVP resistance in 111 of the women, 21 (19%) had detectable mutations at six to eight weeks post dose, most commonly K103N.

In an expanded evaluation of the rate of NVP resistance, paired samples collected at seven days and six to eight weeks were compared [1]. NVP mutations were detected in 66 (24%) out of 271 women and a similar number had mutations at both time points. However patterns of mutations were different.

Y181C was detected in 13 (87%) of 15 women with NVP resistance at seven days, but in only four (22%) of 18 women with NVP resistance at six to eight weeks. In contrast, K103N was detected in six (40%) of 15 women with NVP resistance at seven days, but was detected in all 18 women with NVPR at six to eight weeks. Analysis of paired samples suggests that of the most common mutations associated with NVP resistance Y181C is selected early, but fades from detection in most women by six to eight weeks. By contrast, K103N is more likely to be detected at six to eight weeks than at seven days.

The investigators noted that “the pattern of NVPR mutations detected after single-dose NVP depends on the timing of sample collection”. And they explained that “more rapid emergence and fading of Y181C vs K103N may reflect differences in the NVP susceptibility and fitness of HIV-1 with these mutations. Analysis of cloned variants reveals that diverse populations of HIV-1 variants with NVPR mutations are selected as early as seven days following single dose NVP”.

A second poster from the same group evaluated the samples by subtype using a univariate analysis and their findings suggest that the rate of NVP resistance was higher in women infected with subtype D than with subtype A HIV-1 [2]. Unsurprisingly high baseline viral load and low CD4 were associated with likelihood of acquisition of NVP resistance. The authors speculated that “the rate of NVPR following single dose NVP prophylaxis may vary from region to region, depending on which subtypes are prevalent.”

In addition a Thai poster assessed the development of resistance in pregnant HIV-infected women and their infants receiving both short-course zidovudine (ZDV) therapy and single-dose intrapartum/newborn NVP [3]. The authors reported that amongst 133 ARV-naive women receiving both prophylaxis strategies, 20 demonstrated NVP and one a ZDV genotypic resistance mutation at one month postpartum and of the three HIV-infected infants tested, one demonstrated NVP resistance.

…and breastfeeding

Breastfeeding may account for as much as a third of mother to child transmission (MTCT) with estimated transmission rates at 0.5%-2% per month. MTCT breastfeeding interventions have included a second dose of NVP to the mother (conferring a threefold increase in incidence of resistance from 19-67% women receiving prophylaxis) [4], and NVP prophylaxis to breastfeeding infants.

In an oral presentation Dr Lee from Stanford University reported findings from a small study conducted in Zimbabwe, comparing the relative concentrations of NVP resistant virus in 33 women enrolled in HPTN 023 who received single dose NVP at the onset of labour [5]. Plasma and breast milk samples were obtained at two, eight, 16 and 20 weeks post partum. At eight weeks 23/33 (70%) women had detectable HIV RNA in plasma. Sequences were available for 33 of the plasma and 20/33 of the breast milk samples. Detection of NVP mutations was significantly higher in the breast milk 3/20 (65%) than the plasma 8/33 24.2%).

K103N (the mutation most commonly associated with NVP resistance) was the most frequently observed mutation in both breast milk (12/20) and plasma (5/33). Only 4/20 pairs of plasma and breast milk demonstrated the same mutations and all samples were found to be subtype C. The investigators also observed: “The 20 breast milk and plasma samples from each woman were more closely related to one another than to sequences from other women.”

And they concluded: “Significantly higher frequency of resistance mutations in breast milk compared to plasma provides evidence for differential selection and expression of NVP resistance in the BM compartment after single dose nevirapine.”

Questions raised following the presentation concerning the limitations of the assays and the possibility that the investigators may be missing minor variants, implications for transmission of resistant virus and the limits these findings may have on the use of NVP prophylaxis for a breastfeeding infant were largely unanswered.

Rapid test

Finally, a poster from Susan Eshleman’s group reported results from an evaluation of NVP resistance in women in HIVNET 012 using the rapid assay Amp-RT (this assay measures reverse transcriptase (RT) enzymic activity and NVP resistance directly in plasma) compared to genotyping using the Applied Biosystems ViroSeq HIV-1 Genotyping System.

Twenty-nine plasma samples from 17 women, including pre- and post-NVP samples were tested. They found that 17 samples had no detectable NVP mutations (wild type) and 12 samples had minor variants with NVP resistance mutations.

Results were obtained for 26 (90%) of the 29 samples, including 16 wild type (WT) samples and 10 samples with minor NVP variants. The other three samples had undetectable RT activity. The Amp-RT assay detected NVPR in six (60%) of the 10 samples with minor NVPR variants. Thirteen (13; 81%) of the 16 wild type samples were susceptible to NVP in the Amp-RT assay, and three had reduced susceptibility. Two of the samples with reduced susceptibility were pre- and post-NVP samples from the same woman who was antiretroviral drug naïve prior to NVP administration. The pre-NVP sample had a lower level of NVP resistance than the post-NVP sample.

The authors reported that “in this study, results from the Amp-RT assay were concordant with results from genotyping in the majority of cases” (including detection of minority variants). They speculate that the finding of reduced susceptibility to NVP in the three wild type samples may indicate that mutations other than those defined in subtype B analysis may cause NVP resistance in other subtypes. Using this rapid assay does not require time isolation and culture and therefore can provide results in one or two days in contrast to conventional phenotype. They also suggest that further studies be undertaken to characterise the full genetic correlates in non-subtype B HIV and to assess the utility of biochemical testing in resource poor settings.


Amongst the discussion points the from the HIVNET study the investigators explained that ‘Selection of NVP resistance in this setting is not unexpected… Because women in HIVNET012 had relatively advanced HIV-1 disease, they may be more likely to develop NVP resistance than women in other cohorts. Data from HIVNET 012 indicates that NVP resistance fades in women over time. Clinical studies are needed to determine whether the emergence of nevirapine resistance after single dose NVP prophylaxis will affect the efficacy of NVP prophylaxis in subsequent pregnancies. It is not clear whether the selection of NVP resistance will limit the use of NVP or other NNRTIs for subsequent treatment of HIV-1 infection.’

They continue ‘ If antiretroviral drugs become more widely available in those countries women who received NVP could be offered treatment with other drugs…’ (our italics).

We would urge for more caution with this approach. By our understanding these findings clearly do warrant a change in policy. Resistance does not fade. Other treatments will not be readily available.

As for the rapid test, given the ease at which NVP resistance appears to develop in both mothers and infants receiving prophylaxis and the not inconsiderable numbers of programmes using this intervention such an assay may prove to be invaluable.


Unless stated otherwise, all references are to the Programme and Abstracts of the 10th Conference on Retroviruses and Opportunisitc Infections (CROI), 10–14 February 2003, Boston.

  1. Eshleman S. H., Cunningham S. P, Jones D et al.  Analysis of nevirapine resistance seven days after single-dose nevirapine prophylaxis: HIVNET 012. Abstract 856
  2. Eshleman SH, GuayLA,. Mwatha A et al.  Extended analysis of nevirapine resistance in women with subtype A vs D HIV-1 6-8 weeks after single dose NVP prophylaxis: HIVNET 012. Abstract 857
  3. Chaowanachan T, Chotpitayasunondh T. Vanprapar N, Resistance mutations following a single-dose intrapartum administration of nevirapine to HIV-infected Thai women and their infants receiving short-course zidovudine. Abstract 855
  4. Sullivan J. South African intrapartum nevirapine trial: selection of resistance mutations. XIV International AIDS Conference, Barcelona 2002. Abstract LbPPeB9024
  5. Lee E, R. Kantor1, E. Johnston Breast milk shedding of drug-resistant subtype C HIV-1 and among women receiving single-dose nevirapine. Abstract 96
  6. Eshleman S. H., Cheingsong R, Garcia G. Evaluation of a rapid phenotypic assay for nevirapine resistance in Ugandan women who received single-dose NVP prophylaxis in HIVNET 012. Abstract 581

Links to other websites are current at date of posting but not maintained.