Interview with Richard Pazdur, MD

New oncology drugs have, for decades, followed a well-trodden path of sequential clinical trials to get from the laboratory to the patient. This model has proven its worth by delivering effective new treatments while also protecting patient safety, but newer “seamless” trial designs have the potential to modernize the drug testing process and allow patients to access promising new treatments more quickly. Richard Pazdur, MD, who has been the FDA’s Director of the Office of Hematology and Oncology Products since 2005, recently co-authored an article in the New England Journal of Medicine1 discussing the main issues surrounding seamless drug development. IASLC Lung Cancer News spoke with Dr. Pazdur about how this new model will affect oncology drug development programs moving forward.

Q: Can you explain what are the most significant differences between conventional sequential trial designs used in the development of anticancer drugs versus the newer seamless approach used for drugs like pembrolizumab?

A: The primary difference is that seamless drug trials are intended to obtain information on drug dosing, activity, and efficacy in one trial and avoid the delays of conducting separate phase 1, phase 2, and phase 3 trials. Seamless design allows greater fluidity and flexibility in the clinical trial, and it eliminates some of the administrative burden on the part of the sponsor and participating institution. It is a much more fluid approach that allows expansion cohorts to be added after the early dose-finding phase of the trial to, for example, investigate activity in different diseases or to start examining various biomarkers to define which populations are most likely to respond to the drug.

Q: What changes in the drug development field have prompted the emergence of the seamless trial design in drug development now?

A: I think the reason why it is happening now is that the drugs are better, and there is also an urgency to get these drugs developed rapidly and get them on the market as expeditiously as possible. We have a better scientific rationale for the drugs that are being developed, and we are seeing activity much earlier than before, in phase 1 studies or the phase 1 components of these seamless trials. This creates an incentive to avoid the “start-stop” process of sequential trials and move toward a more fluid development process in which different disease cohorts can be added as the trial proceeds. The bottom line is that the drugs have gotten better.

Q: What are some of the disadvantages of the seamless trial design?

A: Well, as we emphasized in the NEJM article, one of the things that we really want to make sure of is that we do not lose patient protections that are afforded by the traditional phase 1, phase 2, phase 3 design. Specifically, we want to make sure that patients are adequately informed, that there are revisions to the informed consent, and that there are more interactions between the investigators, the institutional review boards (IRBs), and the FDA. This will require a more open dialogue and more frequent communications on the part of the regulators, the IRBs, and the investigators. It will require more meetings; whether those are conducted in person or in teleconferences will depend on the specific situation, but there is going to be a need for greater communication. There are also unique issues that need to be addressed for seamless trials. Namely, there should be an adequate statistical plan for cohorts being added, including the number of patients to be added, and predefined measures of success and failure. We think these issues can be addressed with careful planning.

Q: Besides progression-free survival and overall survival, what other endpoints should be taken into account with respect to new drug approval?

A: One of the things we are looking at is response rate, obviously. Many times sponsors are coming in with single-arm trials from these expansion cohorts. They are not randomized populations, and response rate is the primary endpoint we look at in these situations. It is an endpoint we commonly see submitted to the agency for approval, especially for some “breakthrough therapy” drugs.

Q: Do regulators, researchers, and trial sponsors have any additional or different responsibilities when participating in seamless, rather than sequential trials?

A: Yes, and again, the key is greater openness and communication on the parts of the investigators, the IRBs, and the regulators. The FDA will also need to modify the format of its oversight teams, which have traditionally been disease-specific. Seamless trials may involve cohorts for several diseases, such as melanoma, sarcoma, lung cancer, breast cancer, all in one trial, so we may have to do some reorganizing of our staff for a specific protocol.

Q: Is there anything else you would like to share with our readers about seamless drug development?

A: I think this model provides an excellent springboard for studying populations that have been excluded from traditional trials, particularly pediatric patients. I’m very interested in highlighting pediatric populations as a possibility for expansion cohorts to investigators, especially after the adult dose has been established. One could also expand to cohorts in renally or hepatically impaired patients. I think that we are just at the beginning of how to look at these trials and use them more effectively in the broader picture of drug development. So far, seamless trials have been confined to the development of anticancer drugs, but I cannot see any reason why this could not be expanded to other diseases outside of oncology as well.


1. Prowell, TM, Theoret, MR, Pazdur, R. (2016). Seamless Oncology-Drug Development. N Engl J Med. 2016. doi:10.1056/NEJMp1603747