START Making Sense
By David A. Wohl, MD – June 12, 2015
Last week investigators conducting the Strategic timing of Anti-Retroviral Treatment, or START, trial comparing the initiation of antiretrovirals earlier versus later in the course of HIV infection, announced that an independent data and safety monitoring board had found significantly higher rates of progression to AIDS or death, as well as serious non-AIDS events, among those assigned to delay treatment. This board recommended that those not already on therapy in the deferral arm be immediately offered HIV treatment.
The study team is understandably excited that their large trial (more than 4,600 participants) finally arrived at a conclusion. However, reading the reports of their results one would think they had stumbled on to some heretofore unknown and well-hidden scientific truth. On the contrary, the greatest novelty in their demonstration that earlier is better when it comes to HIV therapy – regardless of CD4 cell count – is that they were so late to accept this was possible, if not probable.
At the time the START trial was initiated in 2009, the US Department of Health and Human Services (DHHS) CD4 cell count threshold for initiating antiretroviral therapy was 350/mm3 and lopinavir/ritonavir was among the ‘preferred’ agents. Since then, well-designed observational studies, randomized trials of immediate versus delayed treatment at lower CD4 cell counts, and a host of investigations pointing to the pro-inflammatory effects of HIV have consistently concluded that HIV-induced CD4 pool depletion does not do a body good. So, while the START trial chugged along, the mounting data made most experts in HIV therapeutics increasingly nervous about waiting to start HIV therapy. Further, while earlier thresholds for treatment initiation attempted to approximate the tipping point between the risk of complications from HIV infection and those caused by combination HIV therapy, the development of cleaner and more potent antiretroviral regimens also tilted the balance away from postponing treatment.
The START trial team, however, has stuck to a purist position that a randomized controlled trial and only a randomized clinical trial, could answer the ‘when to start’ question. With the HIV Prevention Trials Network (HPTN) study 052 finding that antiretroviral therapy administered to an infected individual protected their sexual partners from infection, START investigators were motivated to fortify their defense of their trial and its design – publishing in BMC Medicine a pointed critique of the DHHS and other guideline panels for moving the CD4 cell count antiretroviral-starting line higher and higher (the DHHS panel removed CD4 thresholds for antiretroviral initiation altogether in 2013) in the absence of the results from START.
However, at UNC, we became more and more uncomfortable enrolling patients into the trial, which called for a delay of treatment until a CD4 cell count of 350/mm3 for those randomized to defer treatment. To us, any equipoise between starting HIV treatment at such a low count compared to those above 500/mm3 had long ago evaporated. We eventually stopped recruiting new participants and over time all of our patients randomized to the deferred arm who agreed to start HIV therapy did so.
Our decision was fortunate as the most astonishing finding of the START trial is that it detected any difference in outcomes between the study arms, at all. Given the hesitancy of many clinicians to defer treatment, those enrolled in the trial were, by nature of low HIV RNA levels (median entry viral load was ~13,000) and stable CD4 cell counts, generally unlikely to develop any of the adverse outcomes which were the study endpoints. That is, by enrolling healthy people doing very well with their HIV, the study was biased toward not finding a difference between starting sooner rather than later. Therefore, that such a difference was seen – not only in both in the developing regions where tuberculosis ruthlessly preys on those deficient in CD4 cells, but also in developed nations – makes clear that the beneficial effect of antiretroviral therapy not only is present but is quite strong.
The ramifications of START in the US will be minimal, other than being fodder for general ‘told you so’ smugness. More data will almost certainly emerge further supporting the benefits of early therapy and we will get reams of papers looking at the effects of therapy on everything from thinking to breathing. In the developing world, though, these results are huge and they made it even more difficult for the WHO and others to avoid advising for antiretroviral therapy for all with HIV.
The story of the START trial will continue to be told for a long time to come. For some it will be a tale of rigorous perseverance in the face of strong counter-prevailing headwinds and ultimate arrival at a result that is solid and conclusive. Others will see a single-minded and aggressive defense of a trial by investigators who refused to accept not only the obvious but also the evidence that rendered their design obsolete and even unethical. Either way, now we can all agree that HIV is a progressive, dangerous, and contagious pathogen and treating it early makes perfect sense.