SDC Insights: How to Guide Clinical Research Sponsors Through an Effective ISS/ISE Strategy

🗨️  What does ISS/ISE mean?

The Integrated Summary of Safety and Integrated Summary of Efficacy (ISS/ISE) is an integral part of a successful New Drug Application (NDA) submission by a pharmaceutical or biotech organization to the FDA.  It is a comprehensive integrated analysis of the effectiveness and safety of a study drug.

These analyses allow reviewers to easily compare cumulative outcomes, tracking subjects’ results across the entire clinical development lifespan of the investigational product.

ISS/ISE summaries can be useful specifically for:

• Identifying broader correlations and trends across overall populations
• Detecting potential safety signals that may not be highly visible from a single study
• Performing subgroup analysis for pediatrics and demographic subcategories and other subgroups of interest

Thus, ISS/ISE offers insight into the complete clinical research program beyond what is observable in any individual clinical trial within the research program.

The first step in any good strategy is to understand the clinical data, the designs of the studies you're trying to integrate, and then to devise the best process for pooling the data appropriately.

The goal of an ISS/ISE is to tell the complete story of your program to the regulatory agency.

The first letter in the acronym ISS/ISE is “I” which stands for Integrated, and this can be the biggest challenge in the process. Statisticians are reconciling different Phases of trials, different study designs, perhaps even crossover designs versus parallel designs. They also need to understand the nuances of the data in each individual study to appropriately integrate analysis data.

As the 2015 FDA Integrated Summary of Effectiveness guidance specifies, integration summarizes in a single document all the information known about the efficacy of the investigational product, primarily focusing on results from pivotal studies but can also include published literature on that investigational product. The purpose of the integrated analysis is to help the reviewer understand the overall evidence for efficacy.

🗨️  Who is involved in developing a proper ISS/ISE strategy?

When it comes to developing the strategy, SDC partners with the Sponsor and all essential voices from the clinical and biometrics team.  To determine the studies to integrate, we focus on the study pools, and group studies with similar designs together.  If there are multiple Phase 1 or healthy volunteer studies, we integrate those together. Phase 3 studies with similar designs are integrated together. Some studies can stand on their own. For example, I had a client with an unblinded long-term extension study.   Its design was unique as compared to the others in the collection, therefore it stood alone.  

Client engagement is critical during this process. For example, a couple of years ago, I was working on an ISS/ISE integration for a client who had 36 studies. I presented an integration strategy based on what I thought was a natural conclusion for the types of studies in the collection. The client reviewed the plan and after providing clinical insight into the indications of their product, we retooled our strategy for a more effective outcome. It was with their collaboration, that we arrived at the best strategy.

When addressing studies that were completed in disparate data systems, the bottom line is that whatever platform you use, the outcome should be consistent SDTM data based on regulatory standardization guidance.

SDC engages with the providers who manage those disparate data systems and ensures open communication with the team, the Sponsor, and the providers to drive consistency in specs and derivations. Consistency means the SDTM's, ADaM datasets, and key variables from one study are consistent with the others and that is achieved through the ADaM data conversions. This is also true for CDISC conversions.

When I am asked about medical coding up-versioning, I always recommend an upgrade to the latest version possible and working with a skilled medical coder.

A common mistake I’ve seen is when a less experienced biometrics team engages with a Sponsor organization and is too eager to jump in and start working without conducting their own thorough analysis to develop a plan.

I also see challenges occur when a team is trying to put a statistical plan together without having a good pooling plan.

Additionally, I see challenges when the team has not established an open line of communication with the Sponsor to understand all the stakeholders, their viewpoints, and any nuances to the investigational product that only the Sponsor may know about.  At SDC, in our ISS/ISE kick-off meetings we build out the communication strategy and make sure it remains collaborative throughout the duration of our relationship.  This also helps us with understanding the displays needed.

🗨️  What does the process flow of a good ISS/ISE strategy look like?

The process flow of a good ISS/ISE strategy begins with understanding the study program based on the following:

• What were the objectives of each study?
• How was the data collected?
• How was it analyzed? And why?

Everything flows from the answers to those questions. If this is our first exposure to the program, then our biostatistics and programming team will coordinate closely with the Sponsor to develop the integration plan.

For programs where SDC has worked on the studies, we already understand the nuances to the data, the Study Data Tabulation Model (SDTM) and Analysis Data Model (ADaM) data. This consists of descriptions of the potential ADaM datasets to be pooled, preparation for the integration of the analysis datasets, and specifications for those analysis datasets.

The next step is developing the data pooling plan. Pooling or the combined quantitative analyses refers to combining data from multiple studies into a single analysis dataset so that analyses can be run on that new integrated dataset. 

Pooling is intended to provide a clearer understanding of responses across multiple studies, different populations (e.g., demographic, specific concomitant illness), and dosing regimens where the individual studies would have too few subjects with these characteristics to support meaningful conclusions.

Ideally, you're going to have clinical data ranging from healthy volunteer studies through to Phase 3 studies where you have proved your hypotheses. The subjects in the healthy volunteer studies didn't have the indication for the investigational product, so in this pool, statisticians may want to look at some of the pharmacokinetic/pharmacodynamic (PK/PD) data or maybe some of the safety endpoints to make sure there were no safety trends in healthy volunteers.

That data may be treated differently than data in later Phases where the drug was studied in patients with the indication for the investigational product. When building your pooling, consider:

• How to pool the data?
• How many pools are needed?
• Does the ADaM data need to be pulled?

The last step of your ISS/ISE strategy is developing the Statistical Analysis Plan (SAP). If you have built your process well from the beginning with strong dataset integration specification, a thoughtful pooling plan, and a visual plan to present in your Tables, Listings, and Figures (TLFs), then your SAP development should go smoothly. 

SDC teams have a very collaborative process with our clients where we build the documentation together considering any nuances to a particular trial. After all, our client’s complete clinical program is at stake, so our clients need to be able to trust what SDC has developed for them.  Understandably, stakeholders want to get to regulatory submission as quickly as possible, while at the same time recognizing the time it takes to produce a submission package with the highest quality to meet or exceed the expectations of any regulatory agency.  

🗨️  What would you say are the most important takeaways from this conversation?

While preparation is key, if I had to boil it down to 3 points, they would be the following: 

• Know the studies and the data within those studies so that you can prepare a pooling plan that makes sense to your team, your client, their clinicians, and the regulatory agencies
• Anticipate the regulatory agency's questions
• Prepare the story in a submissible format

Additional helpful resources: