The results of clinical research studies are often published without details of the scientific and ethical discussions that precede their final approval. The Children with HIV Early Antiretroviral Therapy (CHER) study initiated in 2005 and conducted in South Africa is a case in point.1 The study sought to evaluate immediate versus delayed treatment of infants with HIV infection using antiretrovirals (ARVs). Questions concerning both the ethics and science were almost immediately raised by scientists and advocates outside the study but the NIH/US government supported study nevertheless proceeded. More than 10 years after the study was initiated, a recent article in the New England Journal of Medicine by DeMets and Ellenberg provides a delayed behind the scenes look at the scientific and ethical debate surrounding the study design.2 DeMets and Ellenberg had access to information available to them but not the general public.

The CHER study enrolled HIV-infected infants 6 to 12 weeks of age with a CD4 lymphocyte percentage of 25% or more and randomly assigned them to receive combination antiretroviral therapy when the CD4 percentage decreased to less than 20% or clinical AIDS occurred or to immediate treatment regardless of the CD4 percentage or clinical findings. Subsequently the study was halted by the Data and Safety Monitoring Board (DSMB) because of excess infant mortality in the infants whose treatment was delayed. It should be noted that this result was not surprising to anyone, including in all likelihood the DSMB and the researchers, since a delay in treatment of an infectious disease such as HIV or for that matter any infectious disease, would result in worsening of the disease and if severe enough, end in death. This of course raises questions as to why the study was initiated to begin with.

DSMBs (also called data monitoring committees or DMCs) were established following the recommendation of the Greenberg report in 1967. The responsibilities of DMSBs have increased with time as the complexity and resources for evaluating clinical research have increased dramatically. DeMets and Ellenberg summarize what are considered responsibilities of DSMBs. DSMBs serve a different function from the Institutional Review Boards (IRBs) which provide intimal review and approval/disapproval of clinical research protocols.

“… independent groups of experts have monitored the progress of many clinical trials for early definitive evidence of benefit, convincing evidence of harm, or sufficient evidence of no potential benefit to render continuation of the trial to be futile. Such monitoring is motivated primarily by an ethical imperative; for trials of treatments intended to prevent or delay serious outcomes, one would want to stop the trial and make the superior treatment available as soon as the evidence was definitive. The assessment of risk versus benefit throughout the course of the trial requires frequent access to accumulating data on efficacy and safety…the committees require multidisciplinary expertise and experience, including knowledge of statistical methods for interim data monitoring; they also pay attention to reach recruitment progress and the general quality of the trial with respect to adherence to the protocol and completeness of data collection and follow-up.”

The “behind the scenes” look at the debate of the DSMB concerning the design of the CHER study provides a disturbing insight into what perhaps is undo influence of clinical researchers on the role of DSMBs to protect research subjects. A crucial debate was what statistical method should be used to determine how many infants would be allowed to die in the delayed treatment group before the study was halted.3,4 The clinical researchers chose a method that would result in more infant deaths from HIV infection in order to strengthen their research results. In a callous indifference to the increased number of deaths that would result, the DSMB capitulated to the wishes of the clinical researchers.

From the outset, the ethics of the study were debated. The study was initiated in 2005 almost a decade after international HIV experts called for treating all HIV-infected individuals including infants with immediate antiretroviral treatment. The CHER study researchers ignored the imperatives to treat HIV infection early to prevent disease progression and death, stating that separate studies in infants were required in poor countries. In doing so, the researches ignored the vast evidence that early initiation of treatment was of greatest benefit (and standard of care) for all infectious diseases including HIV. Their claim that separate studies of infants were required countered decades-old standards of care for treating infectious diseases and FDA guidelines that declared that separate studies of therapeutic agents, such as antiretrovirals, were not required for children with life-threatening diseases caused by the same agent as in adults. Their arguments would never have held up in a US court of law.

Rather than using a study endpoint of infant deaths, the clinical researchers could have used a simple blood test to measure the HIV viral load as a surrogate endpoint obviating the need to wait until infants died before ending the study.

The recent revelation by DeMets and Ellenberg of the issues surrounding the design of the CHER study raise troubling ethical and scientific concerns. The fact that the clinical researchers could dissuade the DSMB from using a study design that would have avoided the deaths of infants by instead using viral load as a surrogate marker for efficacy, suggests that the protection of vulnerable infants in poor countries from ethical abuses by the DSMB was inadequate. An issue of conflict of interest was also present if researchers were able to manipulate the DSMB into approving a clinical research study design that are more costly, required larger number of patients, and was of longer duration — factors that were more likely to benefit the clinical researchers then the research subjects. Finally, the behind the scenes information, unavailable to the public, patient advocates, and those providing research consent should be viewed as a failure to provide impartial oversight of clinical research that placed vulnerable infants at an unacceptable risk of participating in clinical research.

References

  1. Violari A, Cotton MF, Gibb DM, et al. Early antiretroviral therapy and mortality among HIV-infected infants. N Engl J Med 2008; 359(21): 2233-44.
  2. DeMets DL, Ellenberg SS. Data Monitoring Committees – Expect the Unexpected. N Engl J Med 2016; 375(14): 1365-71.
  3. Haybittle JL. Repeated assessment of results in clinical trials of cancer treatment. The British journal of radiology 1971; 44(526): 793-7.
  4. Peto R, Pike MC, Armitage P, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. I. Introduction and design. British journal of cancer 1976; 34(6): 585-612.