The U.S. Food and Drug Administration (FDA) allows for some flexibility related to changes in clinical trial protocols—for example, with adaptive trials. However, as one sponsor has found out, certain changes can lead to some pointed questions from agency officials.
TransMedics Inc., an Andover, Mass.-based device company, is waiting for a decision from the FDA on the potential approval for its Organ Care System (OCS)-Lung. This portable monitoring system, which the company has said would mark a step forward in terms of organ preservation, aims to maintain donor lungs in a near physiologic state ahead of transplantation. It also is intended to enable surgeons to perfuse and ventilate the organ between the donor and recipient sites. TransMedics has said OCS-Lung should offer a beneficial alternative to cold static storage, the current standard of care (SOC), by allowing for the following:
- reducing ischemic injury;
- optimizing the lung condition; and
- allowing for ex vivo functional assessment of the donor lung during preservation and prior to transplant.
The Gastroenterology-Urology Devices Panel of the Medical Devices Advisory Committee heard from both the FDA and the sponsor about the system during a hearing in May. Ultimately, members voted 11 to 2 in favor on the device’s safety. In addition, members determined that the sponsor had demonstrated the system’s effectiveness (8 to 5) and that the benefits outweighed its risks (9 to 4).
The FDA is not obligated to follow the voting recommendation of the advisory committee, but it may do so once all information is considered.
Admonition from the panel
Despite these votes, panel members had reservations about the sponsor’s decision to change the primary effectiveness endpoint during the pivotal INSPIRE study. One member even cautioned the sponsor never to conduct such a change during future clinical trials—an admonition with which TransMedics agreed.
Broadly, studies of devices posing significant risk to the health, safety or welfare of a subject require both FDA and institutional review board (IRB) approval. To request an FDA sign-off, sponsors must submit an investigational device exemption (IDE) application, and the agency should issue a positive or negative decision within 30 calendar days. The agency’s decision process is outlined in the guidance document titled “FDA Decisions for Investigational Device Exemption Clinical Investigations.” After a review, it may issue one of the following decisions:
- approval—study may begin upon IRB approval;
- approval with conditions—study may begin upon IRB approval, but sponsor must respond to issues identified by the FDA within 45 days; or
- disapproval—study may not begin until issues identified by the FDA have been addressed.
During its briefing, the FDA noted that it cannot disapprove an IDE application if it believes that the study design is not adequate enough to support a future premarket approval application, 510(k) submission, humanitarian device exemption application or de novo classification. That said, it may discuss aspects of the trial with the sponsor. As the FDA noted in its briefing documents, the agency approves IDEs for clinical studies based on the safety of study subjects. It added that if changes to study endpoints do not have an impact on the safety of the subjects, a new protocol may be approved, even if the agency has concerns. The FDA would address these concerns in approval letters.
The multinational INSPIRE study included 407 subjects, randomized to receive either OCS-Lung (n = 208) or SOC (n = 199). That figure represents the intent-to-treat (ITT) population. Screening failure, resulting in exclusion of patients, occurred in the OCS arm (21%) and the SOC arm (7.5%). The changes left a modified ITT population of 349 randomized subjects. No results were collected from the excluded patients. In addition, the FDA initially approved a composite endpoint of “patient survival at day 30 post-transplantation, and absence of 2005 International Society for Heart and Lung Transplantation (ISHLT) Primary Graft Dysfunction (PGD) grade 3 at 72 hours post-transplantation.” However, TransMedics changed that endpoint to “survival at day 30 post-transplantation and absence of ISHLT PGD3 within the first 72 hours post-transplantation.” The agency advised against the change, as it could “jeopardize the integrity of the study and may invalidate the data collected thus far,” according to briefing documents posted ahead of the meeting. The sponsor did not take this advice.
Selection bias?
As a result of the change, the sponsor changed the primary analysis population from modified ITT to per protocol (PP). The FDA regarded the modified ITT population as primary and the PP as the secondary analysis population. The agency said the PP population is susceptible to selection bias.
During its presentation at the meeting, the FDA noted that it had statistical concerns for the modified endpoint, particularly relating to good clinical trial design. The agency said that primary endpoints and hypotheses should be pre-specified before any medical device clinical trial begins. In addition, the FDA noted that prior to the primary endpoint change, the sponsor had access to unblinded outcome data and 227 of 320 subjects had been transplanted. Access to data related to these subjects, the FDA hypothesized, might have led to the endpoint change.
In addition, excluding patients after randomization potentially could weaken the comparability of the study arms produced by randomization. Further, inferences based on random or ad hoc subgroups present challenges regarding interpretation and generalization.
It remains to be seen how the FDA will decide on OCS-Lung. However, the FDA does not discourage sponsors from making changes. A sponsor may modify the device or clinical protocol without approval of a new IDE application or supplement if it meets certain criteria. Examples include emergency use to protect the life of a subject and certain developmental changes in the device that do not constitute a significant change in the design or basic principles of operation. In these cases, the sponsor must provide notice to the FDA within five working days either of learning of the emergency or making the changes.
The FDA also has provided its thinking on adaptive trials in joint guidance, titled “Adaptive Designs for Medical Device Clinical Studies,” from the Center for Devices and Radiological Health and Center for Biologics Evaluation and Research. The guidance notes that an adaptive design allows for prospectively planned modifications as data are collected without weakening the study’s integrity and validity. The guidance offers other tips, including what factors should be taken into consideration when choosing an adaptive design.
For more information on regulatory developments in Europe and North America impacting the medical device industry, we invite you to join a complimentary webinar on October 10, Medical Devices: Reviewing Latest Regulatory Changes in the EU and the US. Expert presenters will discuss options to help you steer away from risk and uncertainty and toward a more certain path to market.
Editor’s Note: This article was originally published in the Journal for Clinical Studies.
