PROMISE results support WHO recommendations for pregnant and breastfeeding women: more needs to be done to improve ART acceptability and adherence

AIDS 2016 combined logoPolly Clayden, HIV i-Base

Continued antiretroviral treatment was safe in women with higher CD4 counts after delivery and associated with improved maternal health – but virologic failure rates were high – according to data from the PROMISE study, presented at AIDS2016.

The study also revealed that acceptance of ART was low in this population and women needed time to consider starting treatment. And it found that ART during breastfeeding essentially eliminates vertical transmission by breast milk.

PROMISE (Promoting Maternal-Infant Survival Everywhere) is a multi-country, multicomponent study, that began in 2010 and included 5398 asymptomatic HIV positive pregnant women who were not eligible for antiretroviral treatment (ART) at the time of enrollment. The study included breast feeding (BF) and formula feeding (FF) women, depending on local guidelines. Participants were randomly assigned different antiretroviral strategies to look at vertical transmission during pregnancy and post-partum, infant safety, and maternal health.

Research questions include:

  • Antepartum (1077BF/FF) Among HIV positive women who do not meet criteria for starting ART for their own health, what is the best intervention to prevent in utero and intrapartum HIV transmission to infants? Maternal ART vs a single drug prophylaxis regimen.
  • Postpartum (1077BF) Among HIV positive women who do not meet criteria for starting ART for their own health, what is the best intervention to prevent transmission of HIV to infants during breastfeeding? Maternal ART vs single drug infant prophylaxis.
  • Maternal health What is the best intervention to preserve maternal health after delivery? Stopping vs continuing ART.

The first part of the study (finally) provided randomised controlled trial (RCT) data to show that taking three drug ART in pregnancy was more effective in preventing vertical transmission than taking one drug during pregnancy, another in labour and two after delivery. [1, 2, 3]

These findings were reported on 4 November 2014 during a scheduled interim review of PROMISE by an independent data and safety monitoring board (DSMB). The US National Institutes of Health issued a press release explaining the results on 17 November 2014. [4]

Standard of care also changed during the course of the study: as PROMISE was ongoing, the START study showed ART in people with CD4 500 cells/mm3 or more reduces the risk of HIV disease progression. [5] PROMISE participants were informed of these results by the study group and women not receiving ART were strongly recommended to start immediately for their own health.

World Health Organization (WHO) guideline subsequently changed to recommend “Treat All” reflecting the START results. [6]

Several analyses from PROMISE were presented at AIDS 2016 – this commentary summarises results from studies looking at stopping or continuing ART postpartum, ART acceptability among women with higher CD4 counts and transmission during breastfeeding. The results broadly support the recommendations in the WHO guidelines – including lifelong ART for all pregnant and breastfeeding women – but also show that work needs to be done to improve adherence support and acceptability of ART among asymptomatic women.

Continuing ART postpartum benefits maternal health

In 2008/2009 when PROMISE was designed the health benefits of postpartum ART for women with high CD4 counts had not been evaluated in RCT. Judith Currier presented results from the component of PROMISE (1077HS) designed to assess the risks and benefits of continuing vs stopping ART among non-breastfeeding women after delivery. [7] She showed these findings in an oral late breaker presentation on behalf of the study team.

HIV positive, non-breastfeeding, postpartum women with no indication for ART based on local guidelines, and who had received ART during pregnancy for at least four weeks in the main study, were randomised to continue or stop ART within 42 days of delivery. Women were followed for 84 weeks after the last enrollment. Those who stopped were restarted when their CD4 count dropped below 350 cells/mm3 or otherwise clinically indicated.

ART was provided by the study. The majority of participants received a regimen of lopinavir/ritonavir (LPV/r) plus tenofovir DF and emtricitabine (TDF/FTC). Atazanavir/ritonavir and, in some settings, rilpivirine and raltegravir were also available. Overall 90% of women were on PI-based ART.

The primary composite endpoint included: death, time to AIDS event (WHO stage 4), and serious non-AIDS events (cardiovascular, renal or hepatic). The primary safety endpoint was time to first targeted grade 2, 3 or 4 event. Key secondary endpoints were: a composite of HIV/AIDS related or WHO 2/3 events; or time to WHO 2/3 events.

The planned study sample size of 2000 participants provided 90% power to detect a 50% reduction from an annual primary event rate of 2.07% (calculated from other clinical trials). But in November 2014 the DSMB approved stopping enrollment at 1630 participants due to longer than expected time to enrol (the longer follow up was expected to make up for smaller sample size). All analyses were intent to treat.

Participants were informed of the START results and all offered ART in June 2015.

There were1652 participants enrolled from 52 sites across eight countries: Argentina, Botswana, Brazil, China, Haiti, Peru, Thailand and the US between January 2010 and November 2014. The majority were from Brazil (31%), Botswana (28%) and Thailand (18%).

Of the total, 827 and 825 women continued and stopped ART for a median of 2.31 vs 2.35 years; 79 (9.6%) vs 70 (8.5%) respectively discontinued the study. Adherence to randomly assigned treatment was generally good: 15% of women stopped ART in the continue arm and 12% restarted ART before the study threshold in the stop arm.

Median age was 28 years and 28% were black African, 16% Thai (16%) and 15% white. Median CD4 count at study entry was 696 cells/mm3, median ART exposure before delivery was 19 weeks and 91% had entry viral load <1000 copies/mL. During follow up 31% of the stop arm started ART at a median CD4 of 372 cells/mm3.

For the primary efficacy outcome events were very rare and not significantly different between arms. The events included: two cases of cervical cancer, and two deaths (one homicide and one unknown); and two cases of extrapulmonary TB, one toxoplasmosis and four deaths (one hepatic encephalopathy and one unknown), in the continue and stop arms respectively. The rate of safety endpoints was higher in the continue arm compared to the stop arm but this was not statistically significant.

WHO Stage 2 and 3 events were almost halved (44% reduction) with continued ART. The key events were: 16 herpes zoster and four bacterial infections; and six pulmonary TB, 43 herpes zoster, four thrombocytopenia, 10 oral candidiasis and 11 bacterial infections, in the continue and stop arms respectively. See Table 1.

Table 1: PROMISE 1077HS endpoints
Endpoint (time to first event) Continue ART Stop ART Hazard ratio (95% CI) p-value
n rate/100 py n rate/100 py
Primary efficacy outcome 4 0.21 6 0.31 0.68 (0.19 to 2.40) 0.54
Primary safety endpoint 260 18.4 (15.7 to 21.4) 232 15.4 (13.1 to 18.2) 0.08
Secondary endpoints
Composite endpoint 57 3.09 99 5.49 0.56 (0.41 to 0.78) <0.001
WHO 2/3 events 39 2.02 80 4.36 0.47 (0.32 to 0.68) <0.001

Toxicity rates were higher in the continue arm but the difference was not statistically significant.

Among participants randomised to continue ART, 189/827 (23%) had virologic failure at or after 24 weeks of treatment. Of the 155 (82%) with resistance testing, 103 (66%) failed with no evidence of resistance to their current regimen (suggesting non-adherence). Of the 52 with evidence of resistance: 22 had resistance to one of the drugs in the failing regimen; 14/25 (11%) failing a PI regimen and 8/27 (30%) failing an NNRTI regimen.

Low acceptance of ART among women with higher CD4 counts: they need more time to consider

Following the START results, in June 2015, all women not receiving ART at that time were recommended to start for their own health.

In a second oral late breaker, Lynda Stranix-Chibanda showed PROMISE participants’ response to these recommendations and their reasons to accept or decline ART. [8]

The study used a mixed methods approach to collect responses from participants receiving the START information. Study staff contacted participants to return to the clinic and gave START results. This was done using a structured script, the language was chosen by the participant and the staff assessed comprehension. Information included that about the trial aims, study location and results.

Participants also attended a counselling session to discuss the implication of START for them as individuals. Those not receiving ART discussed the offer of starting with the study staff and decided whether or not to accept during that session.

Women selected their primary reason for accepting or rejecting the offer from a set of options. The results were recorded and categorised by the study staff.

All 1483 women not on ART were advised to start: 984 women (66%) accepted the offer but 499 (34%) declined. Acceptance rates varied by country, quite broadly, with a mean of 66% (Brazil) and range of 100% (Peru) to 37% (Tanzania). Reasons for accepting or declining ART after initial counselling session are shown in Table 2.

Table 2: Reasons for accepting or declining ART in PROMISE
Reason 1077BF/FF 1077HS
For accepting ART
Concerned about health 46% 43%
Understands treatment is now recommended 35% 36%
Concerned about CD4 count 16% 13%
Other reason 2% 7%
For declining ART
Wants more time to consider 44% 33%
Feels well/knows CD4 count is high 13% 28%
Concerned about HIV disclosure 9% 3%
Concerned about commitment to life-long ART 9% 7%
Concerned about potential side effects 8% 8%
Other reason 7% 14%
Knows not indicated in current guidelines 6% 0%
Too busy with childcare or other responsibilities 2% 5%
Concerned about adherence 1% 5%

Dr Stranix-Chibanda explained that a number of women were not willing to start ART after a single counselling session. This was despite exposure to considerable ART education and HIV monitoring within a well-resourced trial setting.

Women particularly needed time to consider starting ART and the researchers noted that the women continued to be offered ART through to study exit and the proportion remaining off ART decreases with each visit.


Finally, data were shown from the PROMISE 1077BF study, designed to compare transmission rates with maternal ART vs infant nevirapine (NVP) during extended breast feeding until 18 months post-delivery (the first randomised trial to do so). Taha Taha and colleagues showed these findings in a late breaker poster.

This postpartum component of PROMISE was conducted in 14 sites in: Indian, Malawi, South Africa, Tanzania, Uganda, Zambia and Zimbabwe.

The study enrolled 2431 mothers and their HIV uninfected infants between June 2011 and October 2014. They were randomised at 6-14 days postpartum to maternal ART plus six weeks of daily infant NVP (n=1220) or daily infant NVP (n=1211). Women were asymptomatic with a median CD4 count 686 cells/mm3 and 97% WHO stage I). They were a median age of 26 years. Infants had a median gestational age of 39 weeks and birthweight of 2.9 kg.

Baseline characteristics were similar across study arms. The median duration of breastfeeding was 15 months was also similar across study arms (p=0.85).

Rates of HIV transmission during breastfeeding were very low and did not differ significantly between arms at 12 months postpartum these rates were 0.5% with maternal ART and 0.6% with infant nevirapine. Rates of infant survival were high (98.9%) and did not differ significantly between arms (p=0.72).


All references are to the programme and abstracts of the 21st International AIDS Conference, Durban, South Africa, 18-22 July, 2016 (AIDS2016), unless otherwise stated.

  1. Clayden P. Three drug regimen superior for preventing vertical transmission in the PROMISE study. HTB. 29 January 2015.
  2. Fowler MG et al. PROMISE: Efficacy and safety of 2 strategies to prevent perinatal HIV transmission. CROI 2015. 2015 Conference on Retroviruses and Opportunistic Infections (CROI 2015), 23-26 February 2015, Seattle. Oral abstract 31LB.
  3. Clayden P. Three drug ART best for preventing vertical transmission to infants: results from the PROMISE study. HTB. 24 March 2015.
  4. National Institutes of Health press release. NIH-sponsored study identifies superior drug regimen for preventing mother-to-child HIV transmission. 17 November 2014.
  5. The INSIGHT START study group. Initiation of antiretroviral therapy in early asymptomatic HIV infection. N Engl J Med 2015; 373:795-80. 7 August 27, 2015.
  6. WHO. Consolidated guidelines on HIV prevention, diagnosis, treatment and care for key populations 2016 update. July 2016.
  7. Currier J et al. Randomised trial of stopping or continuing ART among post-partum women with pre-ART CD4 >400 cells/mm (PROMISE 1077HS). AIDS2016. Oral abstract THAB0103LB. (abstract)
  8. Stranix-Chibanda L et al. Low acceptance of early antiretroviral therapy (ART) among post-partum women enrolled in IMPAACT PROMISE studies across the globe. AIDS2016. Oral abstract THAB0106LB. (abstract)
  9. Taha TE et al. Comparing maternal triple antiretrovirals (mART) and infant nevirapine (iNVP) prophylaxis for the prevention of mother to child transmission (MTCT) of HIV during breastfeeding (BF). AIDS2016. Poster abstract LBPE013.

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