Contributed Commentary by Josh Baker

August 13, 2021 | What do pharmaceutical manufacturers, medical device makers, and drug developers need to know when navigating the new drug application (NDA) and biologics license application (BLA) processes? This article will take an insider's look at what is happening inside the FDA and the approach taken by the Center for Drug Evaluation and Research (CDER) statisticians in the review of the drug/device application, including what data are analyzed, the integrated review process, and labeling and post-approval evaluation. Based on experience in multiple FDA advisory committee meetings for new therapeutics, the author will also explore how to be prepared for the interactions between the review team and the applicant during the review cycle.

Seeking FDA approval can be an intimidating process. Typical applicant concerns about ‘what’s happening on the other side’ can be allayed by knowing in advance what information FDA reviewers are seeking and how to be prepared for FDA queries. This article will highlight the demanding role of the statisticians who lead the applicant teams and how they can be energized to successfully react to FDA information requests and findings. 

While the drug/device review process with the FDA can be complicated and sometimes unnerving, the role of statisticians on the FDA teams and those of the sponsor/applicant teams have a key role in moving the process forward smoothly and efficiently. Not only do they lay the groundwork for the review processes, but they also bring insights into how alternative methods can be deployed and are a critical voice when questions arise on various trial aspects such as safety and efficacy concerns.

The important role of these statisticians remains in place as the application moves through the FDA regulatory procedures.

Timing and Who is Involved in the Review

The 21st Century Review Process is generally about a 10-month initiative, but it can be shorter if preliminary work has been carefully incorporated into the initial application. In a typical scenario, when the application has been submitted, it is reviewed by the FDA during the first 30-45 days. Filing and planning meetings occur throughout the process with mid-cycle meetings happening around the fifth month. By month eight, internal FDA action meetings are held. By month 10, input from the action meetings is rolled into a wrap-up. And finally post-action feedback is delivered to the applicant. There are instances where the review process may happen faster—sometimes in just two to three months if urgency is required and submitted applications are complete.

Within the FDA review process there are several NDA/BLA individuals responsible for areas such as statistics, clinical reviews, pharma, toxicology, chemistry, and more. In addition, there are other members of the FDA team that serve in secondary roles but who also are involved in eventual sign-offs. 

Role of the Statistical Reviewer

The role of the FDA statistical reviewer is critical in analyzing submitted data and evaluating it from a statistical viewpoint to determine if the application is complete and that it will meet the regulatory requirements to be filed for the NDA/BLA review. This process requires that the reviewer, upon immediate receipt of the application, is reviewing and accessing the contents, reviewing the label, datasets, reviewer’s guides, annotated case report forms (CRF) and clinical study reports (CSRs) that describe the methodology and results of a clinical trial in NDA/BLA development.

The FDA reviewer may uncover refusal to file (RTF) issues. These issues oftentimes include omission of required data, inadequate content, clear failure of effectiveness compatible with statutes and regulations or lack of missing data assessments. While not common, RTF issues can seriously impact the application as well generate possible fines or loss of research and development investments.

Responding to an Information Request

In addition to meetings and other communications under the 21^st Century Review process, the FDA reviewers and the applicant are frequently interacting through information requests (IRs). These impromptu IRs occur for a variety of reasons and can come from any FDA reviewer or member of the NDA/BLA team during the review cycle. 

Statistical related questions triggering IRs may include unexpected or unclear findings in the statistical review of supported data, questions on the need to consider alternative analyses due to lack of agreement on the statistical approach, risk/benefit dose response or other analyses not originally considered by the applicant. IRs may also question such issues as inconsistencies across study populations, i.e. between trials or among subgroups studied.

Proactively Mapping a Structure Around IR Responses

During the development of the application and before submission, a structure for IR responses should be in place, particularly to assure speed and accuracy. A fully ‘mapped’ process will include all members of the applicant team (primary and secondary responders). This can eliminate challenges typically faced by applicant responses regarding misinterpretation of the IRs or incomplete answers to the IRs. 

Success in responding to a regulatory review requires the applicant to build a core team designated for overall leadership. Statistics and programming leads should be clearly identified by area. A tiered production team needs to be assembled and planning elements of the response should be agreed upon. Finally, all team members should be familiar with the materials presented and how/where to easily access core data. 

A review of regulatory queries includes these steps:

• Review of the FDA query by the full applicant project team
• Review of queries by the biometrics team with reference to submission and data/programming
• Building a framework for response
• Review of the response with the full project team
• Challenge responses/follow-up interactions with regulators

As with the initial applicant submission process, the key to success during the regulatory review process is to be prepared for as many contingencies as possible. Preparation, including mapping of the response mechanisms, includes attention to the critical role of the statistician/statistical review process as well as clear collaboration and communication among applicant team members. 

Joshua Baker, PhD is VP, Biometrics for the CRO PHASTAR. He has many years of industry experience, developing and leading a number of biostatistics departments. He has extensive regulatory experience (IND, CTA, BLA, MAA, and post-licensing) in all geographies, including submission strategy and preparation; and has participated in multiple advisory committee meetings at the FDA for new therapeutics (small molecule and biologics). Josh leads PHASTAR's operations from North Carolina and is actively involved in providing statistical support. He can be reached at joshua.baker@phastar.com.