Thoughts on CROI 2015

The 22nd Conference on Retroviruses and Opportunistic Infections (CROI) was held between Feb. 22-26, 2015, during which “the latest studies, important developments, and best research methods in the ongoing battle against HIV/AIDS and related infectious diseases” were shared.  A good summary of the HIV research presented at the conference can be found here, courtesy of Dr. Paul Sax.  And webcasts of many of the presentations (including slides +/- audio and mp3 format) can be found here on the conference website.

Here are a few thoughts meant to provoke and discuss:

1.     Financial incentives in HIV/ART programs:  El-Sadr et al (Effect of Financial Incentives on Linkage to Care and Viral Suppression: HPTN 065) found that financial incentives did not increase linkage to HIV care in U.S. settings but did improve viral suppression (VS) in those on ART at specific types of sites:  hospital-based clinics, facilities having fewer patients, sites where patients had lower VS at baseline and at peak of intervention.

Should MSF try using incentives in some of the HIV/ART programs that we support, for example in settings where improvements are necessary in HIV testing, linkage to care, and adherence (i.e. viral suppression)? Important questions would include which type of incentive (financial or other), how to determine the amount of incentive given…and to who?  For example, a financial incentive could be offered to adolescents on ART, paid out half before and half after achieving an undetectable viral load.

Any incentive would have to represent both significant value to the individual and a cost-effective improvement to the program, i.e. it would have to be something that would make sense to the Ministry of Health and/or donors in order to be sustainable.  Although one could list many reasons not to use financial incentives, is anyone keen to trial this in a resource-limited setting?  Bearing in mind the cost of second-line ART, not to mention potential savings from prevention of transmitted infections and AIDS-related morbidity, there are real gains to be made.  Perhaps something to discuss with our modelling friends….

2.     Are FTC and 3TC equivalent?  Until now, we had considered them interchangeable, and have been told that both molecules should be equally effective for pre-exposure prophylaxis (PrEP).  But an observational study by Rokx et al (Virological Responses to Lamivudine and Emtricitabine in the Nationwide ATHENA Cohort) raises a big question, since it showed better virological responses in those using FTC + TDF compared to 3TC + TDF in a Dutch cohort, regardless of the third ARV (EFV, NVP, or boosted PI).  For example, virological failure rates at week 48 were 3.6% and 10.8% in those taking FTC and 3TC respectively, when either was taken in a regimen with Efavirenz.

There was no mention of whether a fixed dose combination (FDC) was used, as this could have played a confounding role if those in the FTC group were more likely to have been prescribed a single tablet.  Although further information is needed to explain the difference, it now appears that FTC is a better option than 3TC in first-line ART.

3.     Welcome findings for HIV/ART programs in resource-limited settings, in which delays in switches to second-line ART are often de facto!  Paton et al (Impact of NRTI Cross-Resistance on Second-Line PI + NRTI Therapy Outcomes in Africa) examined the impact of baseline NRTI resistance on responses to subsequent PI + NRTI use in second-line ART and found that genotypic NRTI resistance had no relation to short-term efficacy of the NRTIs used in second-line treatment; even when there was no predicted activity (based on genotypic testing), NRTIs made a major contribution to the efficacy of the second-line ART.  Whether or not NRTI resistance was detected, virological suppression in the second-line regimen was equally good, and better than without an NRTI (i.e. PI only), even when NRTI resistance was proven.

In other words, delays that result in multiple NRTI resistance do not seem to have a negative impact on early efficacy or the added value of NRTIs being used in second-line ART.  A huge question remains though about any effect on long-term efficacy.  But what is clear is that in the short-medium term, the thing that matters most is ADHERENCE — we need to do all we can to ensure optimal adherence in the HIV/ART programs that we support!

These findings also touch on the thorny question of how long it’s acceptable to delay a switch to second-line ART, while working to improve adherence and viral suppression in someone on first-line ART with a high viral load.  If the individual is doing well clinically and immunologically, one could argue that the efficacy of a second-line ART regimen is put at risk more by an ‘early’ switch (i.e. without adequate attention to improving adherence) than by a delay in switch while working on adherence.  Once the switch to second-line ART has occurred, there may be less motivation to work on adherence due to a false sense of security in both patient and health worker.  Having said that, there is no doubt that in the real world of resource-limited settings, many patients who are not switched will ‘slip through gaps’ in immunological screening.  In these people, the delay in switch may lead to higher risks of morbidity and death.

Whatever happens, it is critical to maximize adherence support in anyone whose viral load is not undetectable on first-line ART.  The debate continues…

4.     Individual benefit of a ‘test and treat’ ART strategy?  Danel et al (Early ART and IPT in HIV-Infected African Adults With High CD4 Count, TEMPRANO trial) added to the body of evidence that early ART reduces morbidity, even when the CD4 count was between 500-800 cells/μl.  The findings of this study are tantalizing but a problem lies in that the control group — people following WHO guidelines related to ART — includes many from earlier years, when they had to wait for their CD4 count to drop below 200 or 350 cells/μl before being initiated on ART.  If many in the WHO (control) arm were below those lower CD4 counts, that can be expected to wipe out the significance related to the benefit for those who we are really interested in:  people starting ART when their CD4 count is above 500 cells/μl versus those starting above 350 cells/μl.  There are unlikely to be enough in the 350-500 group to compare <500 with <800; if there aren’t, then the findings don’t really tell us anything about whether ‘test and treat’ makes sense.

It looks like we will have to wait for the full publication before having an answer to the critical question of CD4 level at ART initiation in the  WHO arms.  Until then, ‘late presenter’ early publication bias can not be discounted.  We all want to believe in the individual benefit of a ‘test and treat’ ART strategy!


Please feel free to comment on any of the above musings.


Tom Ellman

Head of the Southern Africa Medical Unit (SAMU)