The case against a when-to-start antiretroviral therapy trial.

Dr. Schouten is a member of the Community Constituency Group (CCG) and one of two CCG representatives to the Adult Executive Committee of the Adult AIDS Clinical Trials Group (AACTG). The views expressed here are his own and do not necessarily reflect those of the AACTG. Although SMART is not designed to determine the optimal time to start antiretroviral therapy, the design is nevertheless relevant to that question for treatment-naive patients. Accordingly, the issues Dr. Schouten raises below are appropriate to SMART volunteers who have never before taken therapy.

In early November 2000, a group of about 20 HIV treatment activists sent a letter to Anthony S. Fauci, MD, Director of the National Institute of Allergy and Infectious Disease (NIAID), expressing serious reservations about a possible federally-funded when-to-start trial (WTST) for antiretroviral therapy. Jules Levin, from the National AIDS Treatment Advocacy Project (NATAP), and I wrote this letter. We felt it important to express our view, shared by many in the HIV-infected community, regarding the feasibility and opportunity costs of a large, simple WTST.

While there is no doubt that the question of when to start treatment for HIV is a very important one, not all questions can be answered in a clinically useful manner. All HIV-infected people struggle over when to start therapy. In the end, people evaluate available data and make a decision with which they are comfortable. Very few people would allow such a critical decision to be made by the "flip of a coin." Whether or not to begin therapy is one of the most significant decisions one has to make while living with HIV. Even if enough people joined a WTST, the answers derived from such a trial may not provide information meaningful to the individual. There are many factors involved in making that decision, and many of them are psychosocial, not medical. Additionally, a number of new developments will emerge during the study, rendering the study's results less meaningful. These developments include new drugs, new classes of drugs, new understandings of and interventions for metabolic and morphologic changes, and possibly other unforeseen issues.

One of the most important medical factors not adequately addressed in any proposed WTST is the individual rate of disease progression. An option for correcting this deficiency would be to include a 3- to 6-month observation period, prior to randomization, to identify rapid and slow progressors. Thus, study results could be interpreted by stratification or subset analysis based on differential disease progression rates. This could be problematic, as slow progressors would more likely comply with a randomization to deferred therapy, while rapid progressors would drop out and begin therapy. However, without that observation period, the results will not be clinically applicable to the individual.

When counseling people on the difficult issue of when to start therapy, many providers currently tell their patients that the best time to start therapy is when they are ready. Thus, there may be a high dropout rate after randomization to the immediate therapy arm. Having a person buy into the need for therapy, as well as the appropriate timing for initiating that therapy, is critical to ensure maximal adherence. The standard of care in most communities has already moved away from aggressive early therapy towards initiating therapy when there is evidence of moderate immunologic deterioration. In fact recent changes in the US Department of Health and Human Services treatment guidelines now recommend deferred therapy.

Many health care providers are concerned about enrolling patients with early HIV disease in a WTST in which they could be randomized to premature, immediate treatment. If many knowledgeable providers would not enroll their patients in a study, where treatment decisions will be made by the flip of a coin, then it is ethically questionable to let less knowledgeable providers enroll their patients. The same ethical considerations apply to going overseas to conduct the trial if it cannot be conducted in the US.

Another major concern is the potential cost of the trial: well over $100 million of federal funds, as well as opportunity costs. What other critical trials will not be done if we embark on this trial? There must be a meaningful assessment of the opportunity costs of a large, simple, but expensive WTST. What will the effect of a WTST be on enrollment of treatment-naive subjects in other important studies that try to answer key questions concerning treatment choices? While some WTST designs have proposed allowing co-enrollment in ongoing cooperative group and industry trials, recent experience shows that many people are willing to commit to only one study at a time.

Perhaps most important is the question of how informed consent will be conducted in a meaningful way in a WTST. At the January 2000 meeting between NIAID and the community, many of the most vocal supporters of a large, simple WTST admitted that they would never allow the timing of their own treatment to be determined by the flip of a coin. This raises the question: if the most informed HIV-positive people insist on determining their own starting time, how will this trial accrue if fully informed consent is required? There is a flaw even in conducting surveys of patients and doctors about their willingness to enter such a study. What people say they will do is often different than what they will do when faced with a critical life-decision.

In summary, a WTST will not be feasible, it will not accrue, the results will not actually assist in clinical practice, and the real costs plus opportunity costs are too high. Rather than investing in a study that is unlikely to give us information we need, money ought to be spent on other areas with more cost-effective results, such as education for health care providers or effective adherence programs for patients.

Jeffrey T Schouten, MD, JD Seattle Treatment Education Project (STEP)