The Top ISS and ISE Mistakes That Delay Regulatory Submissions
The Integrated Summary of Safety (ISS) and Integrated Summary of Efficacy (ISE) are key components required under US Federal Regulation for a U.S. New Drug Application (NDA) and are highly recommended for Biologics License Applications (BLAs)1,2.
These documents provide integrated analyses of clinical data across studies supporting a drug or biologic (hereafter referred to as “drug”). This usually involves the analysis and development of pooled datasets that harmonize data across trials to enable a comprehensive assessment of the drug’s safety and efficacy profile, allowing the FDA to holistically evaluate the drug.
Over the years, our biostatistics and regulatory teams have supported hundreds of ISS and ISEs, giving us firsthand insight into the challenges sponsors commonly encounter. While many of these challenges are avoidable with the right planning and strategy, if left unaddressed, they can lead to significant rework and regulatory delays.
Here are seven of the most common ISS/ISE mistakes that we see and how sponsors can avoid them.
1. Failing to Plan ISS/ISE Activities Early Enough
A common and costly mistake is delaying ISS and ISE planning until late in the development program. Because these integrated summaries rely on data from multiple studies, they often become critical-path deliverables for regulatory submissions. Initiating planning at the End-of-Phase 2 (EOP2) meeting and incorporating ISS/ISE considerations into pivotal study design allows teams to align data standards, define pooling strategies, and anticipate analytical requirements early. This proactive approach can significantly reduce rework, prevent submission delays, and support a smoother path to regulatory approval.
2. Waiting Too Long to Align with FDA on Pooling Strategy
Early alignment with FDA on the pooling strategy for the integrated analyses is critical. Because pooling data across studies is both time-consuming and technically complex, late-stage changes can have significant downstream effects on submission timelines.
Sponsors that wait until the pre-NDA/BLA meeting to discuss pooling approaches may find themselves needing to implement substantial modifications to study inclusion criteria, pooling methodologies, or analysis plans. These changes can delay not only the ISS and ISE, but also the overall submission. An even greater risk occurs when no FDA feedback is sought in advance and the Agency later requests changes through an Information Request (IR) during the review cycle, resulting in costly post-submission rework.
FDA explicitly encourages applicants to seek feedback on the pooling strategy through a dedicated Type C meeting prior to the pre-NDA/BLA meeting. In recent training materials, the FDA outlines common questions for discussion in this meeting3. The FDA recommends this Type C meeting to occur approximately 1 year before submission, allowing sufficient time to adjust analyses without impacting timelines.
Proactively aligning with the FDA on key elements of the integrated development strategy, such as study inclusion, pooling approach, and analytical methods, can significantly de-risk a delay in the review of the submission and reduce the likelihood of major post-submission rework.
3. Pooling Data Without Adequately Evaluating Study Heterogeneity
Another common mistake is assuming that studies should be pooled simply because they evaluate the same product. However, studies across a single program often differ in meaningful ways, including dose or treatment regimen, duration of exposure, administration route, manufacturing processes, and patient populations. If these differences are not appropriately evaluated and addressed, they can confound and limit the interpretability of integrated analyses, especially when data is pooled.
When significant differences exist across studies, additional analyses, such as stratification, sensitivity analyses, or modeling approaches may be required to understand the impact of key variables. Failure to do so can lead to questions about the validity of conclusions, resulting in rework, delayed review, or limitations to labeling.
A well-defined pooling strategy should clearly articulate when and why data are combined, based on study design, population, and scientific relevance. Establishing a logical, defensible approach, and aligning it with FDA expectations is essential to avoid unnecessary work and regulatory delays.
4. Underestimating Data Standardization and Legacy Data Remediation Efforts
One of the most common and underestimated challenges in ISS/ISE preparation is the effort required to standardize data across studies. With multi-year development programs, studies frequently use different coding systems or data standards (or different versions of the same standard [e.g., MedDRA or CDISC]), as regulatory requirements evolve and standards are updated.
Because integrated analyses require consistent data structures and coding, these differences must be reconciled before datasets can be pooled. While this process can be relatively straightforward when a strategy for data standards is established early in development, legacy studies often require substantial remediation to align with current standards. This can require re-mapping variables, updating medical coding, and resolving ambiguous or incomplete data.
Given the complexity, reconciling legacy data can take weeks to months and should be proactively assessed and incorporated into submission timelines. To minimize delays, applicants should establish a program-level strategy for data standards and medical coding early in development, align on target standards and dictionary versions, and proactively assess legacy studies for remediation needs well before submission preparation begins.
5. Failing to Prospectively Define ISS/ISE Analysis Populations
A key advantage of integrated analyses is the ability to evaluate subpopulations with greater statistical power by pooling data across studies, particularly in small subpopulations or in rare disease settings. However, these populations must be prospectively defined to ensure analyses are unbiased and interpretable. Post hoc definition of subgroups can raise concerns around data integrity and limit the ability to support labeling claims.
Clearly specifying analysis populations within the ISS/ISE SAP helps ensure consistency, transparency, and regulatory confidence in the results. Failure to do so can lead to challenges during review and reduce the impact of integrated analyses.
6. Creating Standalone ISS/ISE Reports When There is No Regulatory Value
While standalone ISS and ISE reports submitted in Module 5 are generally expected for U.S. marketing applications, there are scenarios where creating separate reports provides limited regulatory value.
For example, in programs with only a single pivotal efficacy study or programs with small phase 1/2 studies, which is common in rare or life-threatening diseases, there may be little benefit in producing a standalone ISE report. Similar considerations may apply to ISS in programs with a single multinational study or heterogeneous datasets where pooling is not meaningful. In such cases, integrated analyses can often be sufficiently addressed within the Clinical Summary documents (e.g., Modules 2.7.3 and 2.7.4) and individual CSRs without creating separate ISS and ISE reports.
Rather than defaulting to standalone reports, applicants should evaluate whether standalone reports add value and seek FDA alignment early (see point 3).
7. Failing to Define Clear Key Messages for the Integrated Analysis
Consistent messaging across the marketing application is critical to a smooth review. Misalignment between the ISS/ISE and other components of a marketing dossier, such as CSRs and clinical summaries, can lead to increased information requests and rework.
ISS/ISE outputs are particularly vulnerable, as they are often finalized later in the dossier. Without clearly defined and proactively managed key messages, discrepancies can emerge among the individual clinical study reports, the ISS/ISE documents, and the Modules 1 and 2 documents, including labeling and clinical summaries.
Establishing and maintaining key messages for the integrated analysis and updating them as data evolves helps ensure consistency, strengthens the overall narrative, and avoids unnecessary delays during review.
Conclusion
ISS and ISE development is not simply a final deliverable, it is a program-long strategy that requires early planning, cross-functional alignment, and thoughtful execution. Many of the delays seen in regulatory submissions stem not from a single major issue, but from a series of avoidable missteps in how integrated analyses are approached.
By planning early, aligning with FDA, ensuring data consistency, and applying sound clinical and statistical judgment, applicants can proactively address these common pitfalls and reduce risk to submission timelines as well as strengthen the overall regulatory package.
Ultimately, a well-executed ISS/ISE does more than meet a requirement, it is a clear, credible, and cohesive evaluation to support a drug’s benefit-risk profile, enabling a more efficient and confident regulatory review.
References:
1. U.S. Food and Drug Administration. Guidance for Industry Integrated Summaries of Effectiveness and Safety Location Within the Common Technical Document. April 2009.
2. U.S. Food and Drug Administration.; Integrated Summary of Effectiveness Guidance for Industry. October 2015.
3. Pei, Y. Veronica. FDA Type C Meetings on ISS Safety Analysis Strategy and Related Data Requirements. U.S. Food and Drug Administration, 7 Mar. 2024
