12 month response to HAART in women following exposure to prevention of mother to child transmission regimens

4 April 2007. Related: Conference reports, Pregnancy, CROI 14 (Retrovirus) 2007.

Polly Clayden, HIV i-Base

In an oral late breaker Francois Dabis presented findings from an evaluation of treatment response in women exposed to single dose nevirapine (sdNVP) and short course AZT alone or plus 3TC, for prevention of mother-to-child transmission (PMTCT) in Cote d’Ivoire between 2003 and 2006. [1]

It has previously been found that sdNVP exposure for PMTCT can compromise subsequent NNRTI containing treatment but that acquisition of NVP resistance can be reduced with “tail” coverage with AZT/3TC for 4-7 days post partum. [2]

Additionally, in DITRAME Plus 1.1, this group had found that adding 3TC to AZT at 32 weeks gestation and continuing for 3 days post partum produces low rates of transmission and NVP resistance but reported 8% overall frequency of 3TC associated resistance. But the subsequent consequences on future HAART were not known. [3]

In this study, from a prospective cohort of the Abidjan MTCT-Plus programme, eligible women were HIV positive, had had at least one pregnancy, had initiated HAART with d4T or AZT plus 3TC and NVP or EFV and had at least 12 months follow up on treatment.

“Exposed” women had received PMTCT prophylaxis (of either short course AZT/sdNVP or short course AZT/3TC/sdNVP) during previous pregnancy and “unexposed” women had no previous history of PMTCT prophylaxis. Genotypic resistance testing was performed at week-4 post partum. The investigators defined immunological failure as a decline of more than 30% from peak CD4 cell count at 6 months and virological failure by a plasma HIV RNA>500 copies/mL at 12 months after HAART initiation.

The investigators reported that of 247 HAART treated women, 109 (44%) were unexposed, 86 were exposed to short course AZT/3TC/sdNVP and 52 were exposed to short course AZT/sdNVP. The women in the three groups were comparable for age and CD4 count but women exposed to short course AZT/sdNVP were more advanced by WHO clinical staging.

None of the women tested had detectable AZT mutations. 3TC resistance mutations were detected in 11 out of 73 (15.1% [7.8-25.4%]) women, for whom data were available, exposed to short course AZT/3TC/sdNVP.

Three out of 78 (4.3% [0.5-12.0%]) women, exposed to short course AZT/ 3TC/sdNVP and 16 out of 42 (38.1% [23.6-54.4%]) of women receiving short course AZT/sdNVP had detectable NVP resistance mutations.

Professor Dabis noted that in the 52 women receiving short course AZT, exposure was for a median of 30 (IQR: 28-34) days and treatment was initiated after a median of 28 (24-35) months. And for the 86 women receiving short course AZT/3TC, exposure was a median of 56 (IQR: 37-61) days (with 3 days post partum) and treatment was started after 15 (9-21) months. Overall HAART was initiated 21 (13-26) months after exposure in the 138 exposed women.

Out of 219 women, virological failure was identified in 42 (19.2%) at 12 months. Of this group 50% of women with detectable 3TC resistance; 18.9% 3TC exposed but no detectable resistance and 16.3% of unexposed women experienced failure (p=0.03 in univariate analysis).

Additionally, 27.8%,18.6% and 15.5% of women with detectable NVP resistance, NVP exposed but no detectable resistance and unexposed respectively, experienced treatment failure (p=0.27 in univariate analysis).

In multivariate analysis the investigators found that factors associated with virological failure were: self-reported poor adherence (AOR: 12.68, 95% CI: 2.98 to 53.86, p=0.001); 3TC-resistance mutations post partum (AOR: 6.86, CI 1.1 to 42.93, p=0.05) and baseline CD4 count <200/mm3 (AOR 0.34, CI 0.15 to 0.78, p=0.01).

Detectable NVP resistance mutations/exposure to NVP, maternal age, WHO clinical stage, and hemoglobinemia were not associated with virological failure. 3TC-exposed women who did not develop resistance mutations post partum were not at increased risk of virological failure (p= 0.11 in adjusted analysis).

Exposure to sdNVP was not statistically associated with virological failure (AOR 1.77, CI 0.48 to 6.48, p=0.39) for NVP resistance; AOR 0.35, CI 0.09 to 1.42, p=0.14 for NVP exposure without resistance).

Immunological failure was found in 26 women out of 235 (11.1%); the only associated factor was poor adherence (AOR 12.3; CI 3.2 to 47.8).

Exposure to 3TC and NVP with or without detectable post-partum resistance mutations to these drugs did not predict immunological failure (p= 0.57 and 0.12, respectively in multivariate analysis), nor did baseline CD4+ count, WHO clinical stage, age and hemoglobinemia.

Professor Dabis noted that this study had limitations including small sample size: only 10 women with detectable 3TC resistance (5 had virological failure out of a total of 42 failures) and 18 women with NVP resistance (4 virological failures). Additionally, the investigators were unable to perform an analysis stratified by time interval between PMTCT exposure and initiation of HAART (interval >1 year in 95% of women).

In conclusion he stressed that, “A public health priority is to provide fully suppressive ART for all eligible pregnant women.” And “New PMTCT drug regimens should be investigated for non-ART eligible women in resource limited settings.”

Comments from the floor included Mark Weinburg who described the results as “a bit depressing” and suggested that performing an ultra sensitive analysis for TAMS might be more instructive as TAMS are probably responsible for treatment failure in this cohort along with the 184V. John Mellors agreed and suggested that the non-association with NVP might change with a more sensitive assay.

Comment

The association with 3TC is perhaps not really surprising, since the median length of AZT/3TC (in a non-virologically suppressive regimen) was 56 days. Non-association with NVP was perhaps more of a surprise but 86/138 women received three day “tail” coverage and, as John Mellors noted, this might change with a more sensitive assay.

There was a much shorter time from PMTCT prophylaxis regimen to initiating HAART in the AZT/3TC/sdNVP group and conversely the exposure period was shorter for the AZT group, so straightforward comparisons of the two regimens are difficult.

The results highlight three things. Once again, the need to invest more in identifying women in need of ART for their own health and getting them started on ongoing fully suppressive therapy during pregnancy. Secondly that any potential small efficacy gain in adding 3TC into short course AZT/sd NVP is probably outweighed by the potential risk of 3TC resistance. And that the use of a postpartum tail added to the AZT/sdNVP regimen, which only involves one week of 3TC and is much less likely to lead to 3TC resistance, appears to make this an effective and safer option for women in settings where triple therapy for PMTCT alone is not yet available, as recommended in the 2006 WHO guidelines.

However, as Francois Dabis concluded, none of the current short course PMTCT regimens is perfect and the need to investigate other potential regimens, such as with tenofovir, remains a priority.

References:

1. Coffie P, D Ekouevi D, Chaix ML et al. Short-course Zidovudine and Lamivudine or single-dose Nevirapine-containing PMTCT Compromises 12-Month Response to HAART in African Women, Abidjan, Cote d’’Ivoire (2003-2006). Oral abstract 93LB.
2. McIntyre JA, Martinson N, Gray GE et al. Addition of short course Combivir to single dose Viramune for the prevention of mother to child transmission of HIV-1 can significantly decrease the subsequent development of maternal and paediatric NNRTI-resistant virus. 3rd IAS Conference on HIV Pathogenesis and Treatment, Rio de Janeiro, 2005. Abstract TuFoO2O4.
3. Chaix ML, Ekouevi DK, Rouet F et al. Low risk of nevirapine resistance mutations in the prevention of mother-to-child transmission of HIV-1: Agence Nationale de Recherches sur le SIDA Ditrame Plus, Abidjan, Cote d’Ivoire. J Infect Dis 2006; 193(4):482-7.].