From Before to After: Preparing and Concluding your FDA Data Submission
“A good start is half the battle” (the Before) when submitting data to the FDA and there are a couple of cherries to put on top (the After) when your regulatory group has finally submitted the eCTD to the FDA . A good start is to have early discussions with the agency by regularly meeting them and sharing the status of your clinical data standards. While, the cherry on the top is the continuous support you need to guarantee to your submission project to promptly react when the reviewers come back with questions and additional requests during the review process.
The Good Start: The Study Data Standardization Plan (SDSP)
In the Study Data Technical Conformance Guide, section 2.1 ‘Study Data Standardization Plan’, the FDA recommends including a plan describing the submission of standardized study data at the time of the IND, and updating it in any subsequent communication. This applies for both CDER and CBER divisions. The CBER also requires an appendix in the SDSP, providing more details about the proposed SDTM domain/variable usage, supplemental domain usage and proposed analysis datasets. The FDA also recommends using the template provided by the PhUSE initiative.
The SDSP is a living document and requires continuous updates, including:
New studies – These are to be added as soon they are planned or updated i.e. final version of medical dictionary used.
Data integration plan - Over the drug development process, the submission and its “target” will become clearer and will help in defining the pooling/integration strategy. This will require an update in the SDSP where the sponsor will anticipate how it intends to integrate data from different studies, for either the Integrated Summary of Safety (ISS) or Integrated Summary of Efficacy (ISE). For example, the sponsor could specify if they plan to integrate data first at the SDTM level, prior to creating integrated ADaM datasets, or if they plan to integrate data into an integrated ADaM datasets directly from individual study SDTM(s) or from individual study ADaM(s).
Discussion history with FDA - The SDSP contains a section where the sponsor can track discussions that have happened with the FDA during the various meetings, so that there is a record of the past decisions that were made.
Noncompliance - Any violations to the standards requiring the feedback of the agency can be also tracked in the SDSP.
Some sponsors have recently started to share their experience of submitting the SDSP to the agency,, which implies that creating SDSP and using it when meeting the agency has become a common practice. However, if you have not used it yet, it is never too late to start doing it. The SDSP can also be created retrospectively; this is what we recommend and where we have helped some sponsors when they requested us to support with their submission i.e. at the time of ISS and ISE.
Why test your data package?
With the FDA, you have the possibility of making a test submission of your data submission package (or mock submission). This usually includes only one study and possibly with real data (making a mock submission with only test data does not make sense at all!). Which study you use to make the test submission does not matter to the FDA as this is purely a technical test. However, we do recommend using a study with some sort of complexities, for example when complex transformation of legacy data is made in SDTM, or a study where a standard might be deviated and you want to get a confirmation from the FDA. This is really your opportunity to seek technical advice from the technical team at the FDA and very often, in our experience, we received some good suggestions from them.
The Cherry on the Top: Supporting the Bioresearch Monitoring (BIMO) Inspections
The FDA has the right to audit clinical research sites. Through onsite inspections, the FDA ensures that clinical investigators, sponsors, and Institutional Review Boards (IRB) comply with the FDA regulations, while developing investigational drugs or biologics. Medical reviewers, who are responsible for approving or disapproving a product, consult with BIMO reviewers to choose which clinical trial sites to inspect.
The site selection is based on data submitted and on the information requested by the “Standardized Format for Electronic Submission of NDA and BLA Content for the Planning of Bioresearch Monitoring (BIMO) Inspections for CDER Submissions”, and their recent technical conformance guide where they provided more details and clarified the requirements.
In these guidances they essentially ask for three types of information:
Clinical Study Level Information
Subject-Level Data Listings by Clinical Site (By-Site Listings)
Summary-Level Clinical Site Dataset (the so called CLINSITE)
This is usually requested for major studies only i.e. pivotal studies. This information, particularly the By-Site Listings, is used during the inspection visits at the selected site. Of note, the agency has a plan to generate these listings automatically in the future, from the submitted data, either SDTM or ADaM or both.
Quite often, this additional information is requested by the agency during the review period, after the final eCTD has been submitted. However, the recommendation again is to plan ahead and eventually anticipate details of how you intend to submit this additional information in the SDSP; this includes the trial(s) for which you intend to provide such additional information.
For more technical details, in addition to the above two guidances, I recommend reading the sponsor experience recently presented at PharmaSUG.
Planning ahead is a good step forward for a successful FDA data submission. Planning an early discussion with the agency is something your regulatory group will take care of, but we as biostatistician, statistical programmer and data-manager, can influence them by pushing for earlier technical discussions that include the format of the data we intend to submit. While it is never too late to start data standards discussion with the agency, if you haven’t done it yet, you can consider it since early discussions can have a positive impact on agency trust. Most importantly, it will reduce the need for corrections if for example, CDISC data conversion approaches or the way of handling deviations from the standards are not shared with the agency well in advance.
Angelo Tinazzi is Senior Director, Statistical Programming, Clinical Data Standards and Clinical Data Submission at Cytel. He is a well- published and recognized expert in statistical programming with over 20 years' experience in clinical research. The application of CDISC standards in different therapeutic areas is part of his core expertise since 2003 in particular in the context of data submission to health authorities such as the FDA and PMDA.
Angelo is an authorized CDISC instructor and member of the CDISC ADaM Team as well as the CDISC European Committee where he also manages the Italian-speaking CDISC User Network.