Data Integration Strategy: The Cornerstone Of A Successful Regulatory Submission

, , ,

A regulatory submission marks the end of a long, exciting, and often challenging journey. It’s the final and most pivotal step in clinical development which, if not done correctly, can be a barrier to bringing your valuable product to market. For smaller companies in particular, submission can be a make-or-break situation.

It’s especially true, at this time, when clinical development has never been more complex. Industry guidance is changing for many therapeutic areas and COVID-19 has accelerated the evolution in trial design, giving rise to innovative approaches such as digital, decentralized, and virtual trials. Additionally, there are more data sources than ever before, which demand expert integration prior to submission.

The more things change, however, the more they stay the same. The key to submission success continues to be delivering a clear and effective integrated summary of safety (ISS) and efficacy (ISE) as quickly as possible. This requires a multidisciplinary team of experts—clinical, statistical, regulatory, medical writing and more—who understand the compound and regulatory requirements and can do the hard work of generating and integrating the data to address all safety and efficacy questions.

Plans are never set in stone

It’s important to start with the end in mind, though the inherent nature of clinical development is fluid. In many cases the data, not the plan, provide the direction. Teams won’t really know what they’re facing until they see: first, the data from individual studies; and second, those data in an integrated fashion.

As is standard practice in the clinical trials industry, multiple rigorous studies are required to prove the safety and efficacy of a treatment:

  • Phase I: First-in-human studies in healthy volunteers to determine if a treatment is safe, determine how it works and understand any side effects
  • Phase II: Studies in patients with the disease/condition to see if the treatment is effective and to understand the optimal dose/administration
  • Phase III: Large randomized studies of patients that take place across regions or countries to understand how the treatment compares to any existing treatments

An expert data integration strategy is critical because many forks in the road can and will arise. Product acquisitions are commonplace after a successful Phase II, impacting both plans and data as the scope inevitably changes. Additionally, there may be early-stage trials that are not beneficial to the overall program, so it’s important to be able to consider that data in context and weed it out during data pooling.

A question of timing

Timelines shown are typical but may be compressed in cases such as emergency use authorization or fast track, breakthrough, or other accelerated designations.

Phase III is a critical milestone where positive results mean the product will likely move forward with a submission. Waiting until the end of Phase III to create your data integration strategy is not a feasible option because doing so risks adding months to the process and delaying regulatory approval.

Having the ISS and ISE ready soon after the last Phase III pivotal study database lock can expedite the submission and help speed the new medication to the market. This means that ISS and ISE activities must happen in parallel to the individual pivotal study analysis.

Starting ISS/ISE activities once Phase III pivotal study results are complete can streamline the integrated reporting process, which can reduce effort and costs. But this approach does extend the submission timeline.

Timing is a perennial challenge. Plan too early and you may waste time, money, and resources if initial study results are not optimal. Plan too late, and you lengthen the time it takes to get your product to market.

The data charts your course

The accepted industry guidance for study data submission is Clinical Data Interchange Standards Consortium, or CDISC. These standards determine the format data must be submitted in, for both individual studies and integrated reporting. Although CDISC is used industrywide, there are different interpretations by company and region. The process will mean submitting to one region and then using that information to amend and submit, region by region.

Adherence to the data standards/guidance is critical to helping the U.S. Food and Drug Administration and other agency reviewers review and understand the data. If the data are not in the correct format, the submission will be rejected and will need to be configured and submitted again—costing valuable time, money, and loss of sales revenue.

Good data integration means converting and consolidating data from different studies on multiple systems in different formats. This is not as easy as it sounds because it requires resources who know every study, every format, every endpoint, and the full development process. Regulators around the world will want the data in the standard format for quick interpretation and answers to their questions. This takes specialized knowledge.

Resource across the process

The compound development plan drives the overarching resourcing approach. It’s a delicate dance, being able to estimate what specific resourcing will be required for a successful submission: how many people, what specific experience level, when?

Key roles

There are certain roles to cover through the clinical development journey. We will not discuss what is required for individual studies but for integration it is important that a submission-experienced programming leader is in place. This person can set the programming strategy that will be deployed alongside the integration lead statistician. Involving the medical writer preparing the integrated reports and an experienced regulatory lead early also is essential. They along with the statistician and programmer will help drive the appropriate strategy.

Other key players to involve are medical experts who will help define pooling approaches that are applicable given the available data and medical need/practice plus other experts depending on the need (e.g., pharmacokinetics, data managers for recoding activities).

For data integration and statistical programming, you’ll need people who know what data can go together and are familiar with how to post and interact with regulators. This includes knowledge of platforms and folder structures to keep everything ticked off and in order.

Key skills across the journey

Once the integration strategy is set and ISS/ISE timelines are determined, a team with the right experience level and the right size becomes key to a successful submission.

Those who take on the challenges of an ISS/ISE must possess compound knowledge, therapeutic experience, and statistical and programming technical expertise, but also have a deep understanding of regulatory requirements and know how these vary from country to country.

Throughout the ISS/ISE production period—especially when it is happening in parallel with individual study reporting effort—there will be times when rapid resourcing ramp up becomes necessary to accommodate new or changing analysis requirements across the program. How can the team effectively deal with the peaks and troughs of the resourcing needs to avoid unnecessary delay of the overall submission timelines?

Once the submission goes out the door, you must plan for the period between the filing and regulatory approval. Some regulatory agencies may require rapid responses to review questions throughout the review period. There also will be 120-day updates. These additional requirements must be resourced by members of the core team to ensure efficiency and speed in responses.

There is no one answer to these questions; rather, various scenarios should be worked through with the submission team to determine the best course of action.

Expertise is everything

Statistical/data integration expertise is your primary concern at this stage (as opposed to data management, which occurs during the clinical trial process itself—the exception being to update coding terms to the same versions of dictionaries before integration). Even so, there are many other factors that drive the need for having other skilled experts on the team.

  • The submission strategy for every therapeutic area, indication and region is fully unique—a cardiology study is vastly different from oncology.
  • If you’re working in rare diseases, there will be a small population and pool of data. Do you have a partner who understands this and has proven success in this area?
  • There are always follow-ups and updates from regulators! Submission isn’t the end of the road, so resource accordingly.

Experience and relationships with regulators around the world often go hand in hand. Proactive engagement with agencies is a success factor because there are many open channels to engage with regulators in advance. For instance, you can share your preliminary plan, give them early looks at the data and let them point out any potential red flags or concerns. Take advantage of that so you can see questions coming, make any course corrections and plan for a positive response.

Above all, getting to yes with regulatory agencies involves a deep knowledge of the mechanics of the submission as well as how to interact with regulators. This involves the entire process starting from program development and individual study design, all the way to submission and beyond.

Craig McIlloney is Vice President, Biostatistics and Programming; Robert King is Executive Director; Hunter Everton is Senior Director, Biostatistics and Programming; and Lindsey Lian is Senior Director, Biostatistics and Programming, all of PPD