Women should be offered the choice to avoid treatment with tenofovir and emtricitabine during pregnancy owing to a higher risk of stillbirth and early infant death associated with these drugs, according to new recommendations published this week in the British Medical Journal (BMJ).
The authors – a panel of experts in maternal and child health, and of women living with HIV – say that women's strong preferences to avoid early infant death and have a healthy infant are not given sufficient weight in other recommendations, and that in order to reduce this risk, antiretroviral treatment during pregnancy should be based on zidovudine and lamivudine, not tenofovir and emtricitabine, especially when combined with lopinavir/ritonavir.
Women should be informed of all the potential benefits and harms of treatment options and should share in the decision-making about which drugs they should take during pregnancy, say the authors.
The recommendations recognise that guidelines that take a public health approach, such as those of the World Health Organization, will place greater emphasis on efficient use of resources, and will seek to minimise the number of different drug options used in a health system so as to simplify the delivery of treatment.
But, in an accompanying BMJ Opinion blog post, Alice Welbourn of the Salamander Trust, a long-standing activist for the rights of women living with HIV, strongly criticises the ways in which the World Health Organization and other international agencies have formulated guidance on treatment for pregnant women with HIV. In particular, she says, “Women’s intrinsic rights over our bodies should be prioritised. Women should be offered informed choices just as men are offered voluntary medical male circumcision for HIV prevention.”
In 2014 the World Health Organization adopted a policy, Option B+, which encouraged national health systems to offer antiretroviral treatment for life to all pregnant women living with HIV. In 2016, the World Health Organization recommended treatment for everyone living with HIV, regardless of CD4 cell count, preferably with a combination of tenofovir, either lamivudine or emtricitabine, and efavirenz.
The combination of zidovudine and lamivudine was recommended as an alternative option where the preferred option is contraindicated or unavailable.
Tenofovir and emtricitabine are co-formulated as part of the three-drug combinations Atripla (with efavirenz), Eviplera (with rilpivirine) and in the four-drug pill Stribild (with elvitegravir and cobicistat). Numerous generic products that contain tenofovir are also available in lower-income countries, and increasingly in Europe and the United States.
Due to lack of evidence, the guidelines panel is unable to state whether or not the findings of the systematic review apply to the new formulation of tenofovir, tenofovir alafenamide, contained in the two-drug combination pill Descovy (with emtricitabine) and in the combination pills Odefsey (with emtricitabine and rilpivirine) and Genvoya (with emtricitabine, elvitegravir and cobicistat).
The new guidelines were developed as part of the BMJ Rapid Recommendations series, which is designed to produce up-to-date evidence-based recommendations that involve patient representatives and independent experts. The process was supported by an international consortium, MAGIC, which aims to improve shared decision-making by doctors and patients using the best available evidence.
The guidelines panel highlights evidence from the PROMISE study, which randomised women to zidovudine alone or to receive zidovudine/lamivudine or tenofovir/emtricitabine. All participants randomised to receive zidovudine/lamivudine or tenofovir/emtricitabine also received lopinavir/ritonavir, with a 50% increase in the lopinavir/ritonavir dose in the third trimester.
The PROMISE study showed that women exposed to tenofovir/emtricitabine were significantly more likely to give birth very early (< 34 weeks) than women exposed to zidovudine/lamivudine. The infants of mothers who took tenofovir/emtricitabine were significantly more likely to die in the first week after birth.
The guidelines panel commissioned a systematic review of studies in which the maternal outcomes of tenofovir/emtricitabine treatment or an alternative nucleoside reverse trancriptase inhibitor (NRTI) combination were compared. Studies which evaluated the outcomes of infant exposure to tenofovir/emtricitabine treatment or an alternative NRTI combination were also analysed.
The researchers identified ten studies which met the inclusion criteria. Meta-analysis of these studies showed no difference in the vertical transmission of HIV between NRTI backbones, but the reviewers assessed the evidence on this question to be of low certainty, due to the non-randomised nature of the studies from which it was drawn.
A pooled estimate, derived from three randomised trials, concluded that infants of mothers who had taken tenofovir/emtricitabine were approximately four times more likely to be stillborn or die in the first week after birth (RR 4.40, CI 1.75-11.01).
The researchers calculated that for every thousand mothers exposed to tenofovir/emtricitabine, approximately 66 would have a stillbirth or lose a child in the first week after birth in high-income countries, compared to a background rate of 15 per 1000 in women with HIV not exposed to tenofovir/emtricitabine.
In low-income settings, the absolute number of losses would be much higher, at 304 per 1000, compared to a background rate of 69 losses per 1000. The researchers say the certainty of this estimate is moderate to low and admit that the adverse effect on stillbirths and neonatal death is likely to be an overestimate.
They also note that although another recent systematic review failed to show an increased risk of stillbirth or neonatal death, that review pooled findings from randomised trials and from low-quality observational studies, a methodology described as 'inappropriate' by the authors of the BMJ Rapid Recommendations.
The PROMISE study was the only study identified that provided high-quality evidence regarding the risk of premature birth. It showed that infants of mothers who had taken tenofovir/emtricitabine were approximately twice as likely to be born very prematurely (< 34 weeks) (RR 2.4, CI 1.06-4.97).
Why tenofovir/emtricitabine treatment should lead to worse outcomes is unclear, say the authors, although the increased rate of very premature births might also explain the higher rate of early infant mortality. Two-thirds of all deaths in the PROMISE study were a consequence of premature birth, they note.
Preferences of women living with HIV
A second systematic review also commissioned to inform the recommendations, looked at studies of women’s values and preferences regarding antiretroviral therapy (ART) and pregnancy, including studies which looked at barriers to adherence. The reviewers identified 15 qualitative studies in which six themes were identified:
- A desire to reduce the risk of HIV transmission
- A desire for the child to be healthy
- Concern about side-effects to the child
- Concern about side-effects to the mother
- Desire for oneself to be healthy
- Pill burden.
The reviewers concluded that “many women living with HIV probably consider the health of their child as the most important factor for taking ART during pregnancy, but have to balance this with their own health.”
One reason for preferring a tenofovir-based antiretroviral regimen is that it is also active against hepatitis B, and so it will prevent hepatitis B transmission. Although lamivudine is also active against hepatitis B, there is a risk of developing resistance to the drug if it is taken for a long time. In women with resistance to lamivudine, or in previously untreated women with a high hepatitis B virus viral load (> 200,000 IU/ml), tenofovir might prevent more infant infections than lamivudine, but this benefit is speculative, the panel concluded. The systematic review found no significant difference in hepatitis B transmission risk between women treated with lamivudine and women treated with tenofovir.
The guidelines panel makes a strong recommendation that women should be offered an alternative to treatment with the combination of tenofovir, emtricitabine and lopinavir/ritonavir during pregnancy. Due to the strength of the evidence, it makes a weak recommendation that zidovudine/lamivudine should be offered in preference to tenofovir/emtricitabine during pregnancy.
They say that based on current knowledge regarding drug interactions, it is not plausible that the combination of tenofovir with lopinavir/ritonavir explains the higher risk of stillbirth and neonatal death in the PROMISE study, and for this reason, the recommendation to avoid tenofovir/emtricitabine should apply to all women living with HIV.
Interactive infographic of the recommendations
Siemieniuk R et al. Antiretroviral therapy in pregnant women living with HIV: a clinical practice guideline. BMJ Open 358: j3961, 2017.
Siemieniuk R et al. Antiretroviral therapy for pregnant women living with HIV or hepatitis B: a systematic review and meta-analysis. BMJ Open 7: e019022, 2017.
Lytvyn L et al. Values and preferences of women living with HIV who are pregnant, postpartum or considering pregnancy on choice of antiretroviral therapy during pregnancy. BMJ Open 7: e019023, 2017.
Fowler M et al for the IMPAACT 1077BF / 1077FF PROMISE Study Team. Benefits and risks of antiretroviral therapy for perinatal HIV prevention. New Engl J Med 375: 1726-37, 2016.