Integrase inhibitors give greater chance of viral suppression at delivery in pregnant women

Integrase inhibitor-based treatment with either raltegravir (Isentress) or dolutegravir (Tivicay, also in Triumeq) reduces viral load more rapidly than efavirenz if started in pregnancy, findings from two randomised studies presented at the Conference on Retroviruses and Opportunistic Infections (CROI 2019) in Seattle show.

Investigators on the raltegravir study say that their findings support the use of raltegravir-based antiretroviral treatment during pregnancy, especially for women who start treatment late in pregnancy.

Professor Saye Khoo of Liverpool University said that the findings of the DOLPHIN-2 study of dolutegravir initiated in late pregnancy suggest that dolutegravir is a drug that could be used in this high-risk situation for mother-to-infant transmission.

Many women with HIV learn of their HIV status when they are tested during pregnancy, often after the first trimester. Reducing viral load rapidly during pregnancy is essential to achieve an undetectable viral load at the time of delivery. Undetectable viral load at delivery maximises the chance that HIV will not be transmitted during delivery.

But achieving undetectable viral load at delivery requires diagnosis of HIV and treatment initiation in time to suppress viral load. In South Africa, one-fifth of pregnant women start treatment in the third trimester of pregnancy (week 24 onwards).

Integrase inhibitors reduce viral load more quickly than other antiretroviral drugs, so a combination that includes an integrase inhibitor might be expected to increase the likelihood of viral suppression by the time of delivery in women who start antiretroviral treatment during pregnancy.

However, the safety and efficacy of integrase inhibitor treatment in pregnancy has not previously been tested in a randomised trial.

As integrase inhibitor-based treatment becomes the standard of care in many countries, it is important to know whether an integrase inhibitor is more effective than efavirenz (the currently recommended standard of care for pregnant women) in suppressing viral load during pregnancy.

Safety is also a major concern. Although integrase inhibitors have been used to treat tens of thousands of people, there are few data on their use in pregnant women, especially from randomised studies.

Preliminary data from an observational study in Botswana showed an increased risk of neural tube defects in infants exposed to dolutegravir in the first 12 weeks of pregnancy. The foetus is vulnerable to birth defects due to drug exposure during the first 12 weeks of pregnancy as organs and tissues form. In later pregnancy (after week 12) drug exposure may affect infant growth or lead to premature delivery or may cause side-effects for the mother.

Raltegravir-based ART started in pregnancy

The NICHD P1081 study randomised previously untreated women initiating antretroviral therapy (ART) in later pregnancy (after week 20) to receive either raltegravir- or efavirenz-based ART.

The trial was carried out in South America, Africa, Thailand and the United States between 2013 and 2018. It recruited 408 women, of whom 307 had no genotypic resistance to any study drug at study entry. The trial recruited an even number of women who commenced treatment between 20 weeks and 28 weeks of gestation, and women who commenced treatment between 28 weeks and 37 weeks of gestation.

The study population was predominantly Hispanic (53%) and black (36%) with a median viral load at study entry of 3.9 log₁₀ copies/ml (just under 8000 copies/ml). The median gestational age at study entry was 29.6 weeks.

The primary analysis of the study was confined to 307 women without genotypic resistance. In this population, significantly more women randomised to raltegravir had an undetectable viral load below 200 copies/ml at delivery (94% vs 84%, p < 0.001) and this association was strongest in women who had started treatment after week 28 (93% vs 71%). The rate of viral load suppression did not differ at the time of delivery in women who initiated treatment prior to week 28.

When the investigators examined responses according to speed of viral load reduction and persistence on therapy, it became clear that achieving very rapid viral load reduction and staying on treatment up to delivery most strongly differentiated between raltegravir and efavirenz.

Ninety-two per cent of women randomised to raltegravir achieved either a 2log decline in viral load, or a reduction viral load below 200 copies/ml at week 2 and maintained a viral load below 1000 copies/ml at all time points after week 4 and remained on study drug to the time of delivery. In comparison, only 64% of women randomised to efavirenz achieved all these outcomes. The difference was most pronounced in rapid viral load reduction; 93% of women on raltegravir achieved either a 2log decline in viral load or a reduction viral load below 200 copies/ml at week 2, compared to 69% of women on efavirenz.

The median time to viral suppression below 200 copies/ml was 8 days in women receiving raltegravir and 15 days in women receiving efavirenz.

There was no difference in any adverse outcome for mother or infant, including stillbirth, pre-term birth or grade 3 or 4 maternal adverse events, between the two study arms. Six infants in efavirenz arm and one infant in the raltegravir arm were found to be HIV PCR positive after delivery; all had been born to mothers who enrolled at 28 weeks of gestation or later.

Dolutegravir-based ART started in late pregnancy

The DOLPHIN-2 study randomised previously untreated women initiating ART from week 28 of pregnancy to receive either dolutegravir- or efavirenz-based ART.

The study recruited 268 women, of whom 237 were evaluated for efficacy. There was no significant difference in baseline viral load (4.5 log₁₀ copies/ml) or median gestation at baseline (31 weeks), nor in the median duration of ART prior to delivery (52 days dolutegravir, 59 days efavirenz).

At delivery, a significantly higher proportion of women in the dolutegravir arm had a viral load below 50 copies/ml (the primary outcome) (73.8% vs 42.6%, p < 0.0001). Women in the dolutegravir arm were 66% more likely to have an undetectable viral load by the time of delivery (adjusted RR 1.66, 9%% CI 1.32-2.09). Baseline viral load and week of pregnancy at initiation did not affect the result and dolutegravir was also superior to efavirenz when assessed according to the proportion of women with a viral load below 1000 copies/ml at delivery (93% vs 83%).

There was no difference in maternal adverse events between the two study arms, nor in median gestational age at delivery (39.9 weeks), nor in premature births. No neural tube defects were reported and there was no difference in congenital abnormalities. Approximately 10% of infants were diagnosed with postnatal infections and eight infants died, consistent with the poor outcomes seen in other studies in which mothers had initiated antiretroviral treatment late in pregnancy.

Three cases of mother-to-child HIV transmission occurred in this study, all in the dolutegravir arm. The investigators say it is likely that these transmissions occurred in utero, not at the time of delivery.