The FDA, Juno Therapeutics, and the ethical imperative of transparency

On July 7, Juno Therapeutics announced that the US Food and Drug Administration (FDA) had placed a clinical hold on the company’s phase 2 trial of its investigational immunotherapy, JCAR015, as a treatment for patients with relapsed or refractory B cell acute lymphoblastic leukaemia (ALL). The trial planned to enrol 90 ALL patients. However, at the time of the hold, only 20 patients had been enrolled, three of whom had died unexpectedly from cerebral oedema.1

The FDA lifted its hold three days later without further public comment, and Juno representatives explained that the three deaths were thought to be related to a mid-trial modification to the protocol. The original protocol included the administration of cyclophosphamide as a sensitising agent to enhance the efficacy of JCAR015. However, after the trial was underway, a second sensitising agent, fludarabine, was added on the basis of promising evidence from the company’s other immunotherapy studies. In its public statements, Juno attributed the deaths to the combination of fludarabine and JCAR015.2

Before a new drug can be tested on US patients, the sponsor submits an Investigational New Drug application (IND) to the FDA, which automatically goes into effect after 30 days, allowing clinical trials to begin. The IND includes a summary of pre-clinical research and can include informal plans for the conduct of clinical trials, as well as descriptions of how informed consent or local trial oversight will be conducted.

If a major patient safety problem is apparent from the submission or emerges after the start of a trial, the FDA can issue a clinical hold on an IND, but holds are relatively rare; one review of company press releases found reports of only 29 holds in an eight year period, compared with the thousands of drugs being tested at any one time.3

Despite the public health significance of a clinical hold, information related to the event is relatively sparse because, owing to its current interpretation of what qualifies as confidential commercial information, the FDA does not disclose the existence of INDs (the European Medicines Agency does not require INDs for early stage clinical investigations, leaving the oversight task to member countries). Nor does it routinely comment on clinical holds or their resolution, leaving the company’s voice as the only source of information.

Companies have a strong financial incentive, however, to downplay problems related to their investigational drugs; one study of the complete response letters that the FDA issued—but did not publicly release—for drugs denied market approval found that the sponsoring manufacturers’ press releases often did not fully explain the reasons for rejection.4

Instead of being considered the protected trade secret of the sponsoring manufacturer, the existence of an IND and the details of a clinical hold should be promptly disclosed. The hold should serve as an important moment for reflection that allows the research and clinical communities to re-examine the existing rules and safeguards and, when necessary, to refine or establish new guidelines or regulations that will better ensure the safety of clinical trial participants and promote thoughtful design, conduct, and oversight of trials. For example, in the Juno case, what role (if any) did local institutional review boards or data safety monitoring boards have in evaluating the protocol adjustment? Were there preparatory preclinical investigations that could have been taken that might have predicted the outcome of adding the second agent? Given the risks that can be associated with major adjustments made midway through a trial protocol,5 it is imperative that these types of questions are adequately answered.

Unfortunately, because the FDA does not make its decision making process public, no one has insight into whether these issues were addressed and no one knows if the agency agreed with Juno’s explanation. A systematic and transparent decision making process for clinical holds is critical to maintaining public trust in the integrity of the research and regulatory enterprises. Although fast and efficient regulatory decision making leading to resumption of the trial with its original design seems a positive, the brevity of the hold in the Juno case (previous holds have lasted for an average of eight months)³ is particularly puzzling and seems like a missed opportunity to show that the agency has conducted a sufficiently rigorous review.

In addition, the FDA should have a predetermined set of steps to clarify the source of the failure and the protections needed for future patients receiving the investigational drug. Such a regulatory checklist could be developed to mirror those used in surgery, which were themselves developed in response to the persistence of similar (thankfully rare) major medical system failures, such as retained instruments in surgery or operations on the wrong body part.6 But because the FDA lifted its hold without providing a public explanation as to what happened and why, this case becomes a challenge to the integrity of the research enterprise, rather than a learning opportunity.

Increased transparency related to INDs, clinical holds, and the decision making process in these and other cases could promote protection of patients to ensure a supply of volunteers for future trials, improve the experimental process for new therapies, and enhance appreciation for FDA’s vital oversight role.