Phase 3 registrational data is not sufficient for roll-out of new ARVs in low- and middle-income countries

Polly Clayden, HIV i-Base

Phase 3 randomised trials for drug approval in high-income countries do not provide sufficient evidence to support the widespread use of new antiretrovirals in low- and middle-income countries (LMICs), where the majority of people with HIV live.

Once again this issue of missing evidence was highlighted in a paper published ahead of print in AIDS 9 May 2018. The paper summarises discussions from the Third Conference on Antiretroviral Drug Optimisation (CADO 3) that took place at the end of 2017. [1] A summary meeting report is also available on the WHO website. [2]

This meeting focused on optimised second- and third-line ART for adults and the sequencing and recycling of key products: tenofovir pro-drugs (tenofovir disoproxil fumarate (TDF) and tenofovir alafenamide (TAF), dolutegravir (DTG), and darunavir/ritonavir (DRV/r).

Key missing evidence is typically for pregnant women, people with HIV/TB coinfection and people who have not had resistance testing before starting ART. “We need to work in partnership with originator companies to design and perform these studies earlier, so that results can support rapid expansion of new treatments in LMICs” the authors write.

  • For DTG, results from new clinical trials to evaluate the transition from efavirenz (EFV)-based to DTG-based regimens are needed to ensure safety and efficacy for the nearly nine million people on NNRTI based ART in LMICs who have either detectable or unknown viral loads.
  • More evidence is needed on the use of TAF in pregnant women (in this case, considerably more evidence) and people with HIV-TB coinfection taking rifampicin-based treatment to support widespread use of the new TAF/3TC/DTG combination in LMICs.
  • Data from randomised phase 3 GEMINI trials of DTG/3TC will be available in 2018, but the overall disadvantages of this approach appear to outweigh the advantages for people with HIV in the setting of LMICs.
  • For new drugs, there is substantial additional research required to evaluate an antiretroviral for widespread use in LMICs. This should be important for new antiretrovirals in development.

CADO3 participants looked at the growing body of programme data on the safety of DTG in first-line – particularly from Botswana and Brazil – and agreed that these data support further expansion of a DTG regimen as a preferred first-line option. But they also agreed that a robust research programme to provide the remaining missing evidence to support its widespread use, as well as careful collection and analysis of programme data, were essential in parallel to countries’ transition.

The strategy of universally switching people who are currently stable an EFV-based regimen and DTG’s role in second-line was still debatable.

Participants judged EFV 400 mg to be the alternative first-line option for people who cannot tolerate DTG or for countries where it cannot be accessed because of cost and patent protection.

They did not support the use of two-drug regimens for adults based on the trial data currently available. As ever, recent and ongoing studies of two-drug regimens do not consider the usual important populations who will be treated in LMICs: pregnant women, people with TB and HBV co-infections, people diagnosed in advanced HIV infection; and populations with no or limited access to viral load and/or resistance testing.

Many of the two-drug investigations are switching studies. In real-life such strategies would require both widespread use of viral load testing as well as the procurement of two products – there would be many programmatic challenges.

The role of DTG in people who previously failed NNRTI-based regimens and whether or not TDF and TAF could be recycled were defined as key priorities. Following the DAWNING results there was much discussion at CADO3 on whether these results could be duplicated in a public health setting with no genotyping. Trials to answer this question, and whether DTG will perform similarly to a protease inhibitor in the context of NRTI resistance – as in the EARNEST, SELECT and SECOND-LINE studies – or will need to be combined with different NRTIs were judged by the participants to be essential.

Dose optimisation studies on use of low dose DRV/r in second-line for people who either failed first-line or were stable on another second-line regimen was also considered to be a priority.

Long acting drugs (oral, injection or implants) and nanoformulations were judged high-priority in the longer term. But, once again, in order for these or any pipeline products to be usable in LMICs, studies will need to include pregnant and women of child-bearing age, adolescents, people coinfected with TB and on treatment, and other co-morbidities.

And any investigations should consider the circumstances in which these products are likely to be given, such as no or limited viral load monitoring nor access to other tests, highly trained experts or laboratories.

CADO 3 defined a prioritised portfolio of new adult ARV products. See table 1. This product portfolio will be updated on a regular basis, like the PADO list, as new information is made available on existing products or on new products.

Table 1: CADO 3 prioritised optimised products

Short-term 1–2 years Medium-term 2–5 years Long-term 5+ years
TDF/XTC/DTG TAF/XTC Long-acting formulations (entry inhibitors and INSTIs)
TDF/3TC/EFV400 TAF/XTC/DTG Maturation and capsid inhibitors
DRV/r 400/50 mg New DRV/r formulations* bNAbs

*Low dose standard formulation (400/100 mg) or standard dose nanoformulation (800/100 mg).

Other lower priority products might be considered if data suggests superiority to existing products.


  1. Vitoria et al. The transition to dolutegravir and other new antiretrovirals in low- and middle- income countries – what are the issues? AIDS, Publish ahead of print 9 May 2018.
DOI: 10.1097/QAD.0000000000001845.
  2. WHO. Third conference on antiretroviral drug optimisation (‎CADO 3)‎: summary meeting report, 29 November to 1 December 2017, Rosebank Crowne Plaza, Johannesburg, South Africa. (PDF)

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