Streamlining the Integrated Summary of Safety (ISS) Process for Oncology Medications with Multiple Ongoing Submissions

Photo by National Cancer Institute on Unsplash

Introduction

With the increasing complexity of clinical trials for oncology medications, the need for efficient and streamlined processes to create Integrated Summary of Safety (ISS) packages has become crucial. This article focuses on the processes and techniques for generating ISS packages for compounds with multiple ongoing submissions, such as an oncology medication being submitted for multiple tumor types or indications.

Purpose and Scope

The purpose of this article is to:

• Explain ways to address the unique challenges for ISS packages with multiple ongoing submissions.
• Suggest processes and techniques to efficiently build and maintain a reusable library of reference safety datasets (safety stack) that reflect the compound's safety profile.
• Discuss methods to harmonize datasets to support integrated analysis with multiple data sources, ensuring harmonized variable attributes, compliance of integrated ADaM datasets, alignment with pivotal trial Clinical Study Reports, and usage of common dictionaries.

Reference Safety Dataset Library Overview

The Reference Safety Dataset (RSD) Library is a centralized repository containing submission-ready ADaM datasets for all clinical trials identified in the compound's RSD population. The library is organized by MedDRA version and is regularly updated to accommodate additional trials, ADaM standard updates, and new dictionary versions. This library can be reused across individual submissions, ensuring efficiency and consistency.

High-Level Process for Reference Safety Dataset Library Creation and Maintenance

The process of creating and maintaining the RSD Library involves the following steps:

1. Define: Gather and document all requirements for the ISS package, including the list of trials, cut-off dates, analysis dataset specifications, TLF mockups, and validation plans. Program validation for each component is planned and documented.
2. Develop: Use the requirements from the Define phase to create SAS programs for ADaM, Pooling of datasets, and TLFs.
3. Validate: Programs are validated in the test area according to the validation plan using the requirement specifications and analysis plans. A dry run is conducted to ensure draft output meets expectations.
4. Operations: Validated programs are executed with production data to generate/create analysis datasets and tables, listings, and figures. Change Management/Maintenance is done as needed to address new requirements or issues.
5. Reference Safety Dataset Library Uses

The RSD Library serves as a valuable resource for the following purposes:

• Ensuring consistency across submissions, as it uses a central ISS team to create and validate submission-ready ADaM datasets for each individual trial.
• Reducing the time and effort required to create ISS packages, as the library can be reused across each individual submission.
• Facilitating updates and additions of new trials, ADaM standard updates, and new dictionary versions.

Analysis Populations

The analysis populations for the ISS package are defined by the regulatory authorities based on the submission and approval status of each clinical trial. These populations are used to generate the RSD Library and ensure that the ISS package accurately reflects the compound's safety profile across all relevant clinical trials.

Best Practices for ISS Datasets

To ensure the quality and consistency of the ISS datasets, several best practices should be followed:

• Use the CSR analysis datasets as the starting point for creating the RSD Library.
• Establish clear standards and templates for the RSD Library to maintain consistency and compliance with regulatory requirements.
• Utilize a centralized ISS team to create and validate submission-ready ADaM datasets for each individual trial, ensuring consistent application of standards and validation processes.
• Regularly update the RSD Library to incorporate new trials, ADaM standard updates, and new dictionary versions.

Reference Safety Dataset Library Templates and Standards

Developing templates and standards for the RSD Library is essential for maintaining consistency and ensuring compliance with regulatory requirements. These templates and standards should include:

• Analysis dataset specifications, outlining the structure and content of the datasets.
• A standardized format for naming and organizing datasets and subfolders within the library.
• A clear process for updating and maintaining the library, including procedures for incorporating new trials, ADaM standard updates, and new dictionary versions.

Integrated Analysis Standards and Templates

In addition to the RSD Library standards and templates, integrated analysis standards and templates must be developed to support the harmonization of datasets across multiple trials. These should include:

• Guidelines for harmonizing variable attributes across multiple trials, ensuring consistent data representation.
• Templates for pooling datasets from different trials, facilitating the generation of integrated ADaM datasets.
• Standardized formats for tables, listings, and figures, ensuring uniform presentation of results across different trials and submissions.

Programming Important Points and Checks

To ensure the accuracy and quality of the ISS package, it is crucial to perform several checks and validations during the programming phase, including:

• Verification of dataset specifications, ensuring that they align with the submission and analysis plans.
• Validation of SAS programs for ADaM, pooling of datasets, and TLFs, confirming that they meet the requirements specified in the validation plan.
• Review of draft outputs, identifying any discrepancies or issues that need to be addressed before finalizing the ISS package.

MedDRA Leveling and Dictionary Conversion

As the RSD Library is organized by MedDRA version, it is essential to perform MedDRA leveling and dictionary conversion during the dataset creation and harmonization process. This involves:

• Mapping the data from different trials to a common dictionary version, ensuring consistent coding and reporting of safety data.
• Leveling the MedDRA terms, facilitating the aggregation and comparison of safety data across multiple trials.

Unique Challenges and Solutions

Creating ISS packages for compounds with multiple ongoing submissions presents several unique challenges, including:

• Harmonizing variable attributes and datasets from different trials, which can be addressed by following established integrated analysis standards and templates.
• Ensuring compliance of integrated ADaM datasets, which can be achieved by adhering to the RSD Library templates and standards and utilizing a centralized ISS team for dataset creation and validation.
• Aligning the ISS package with the pivotal trial Clinical Study Reports, which can be facilitated by using the CSR analysis datasets as the starting point for creating the RSD Library.

Quality Control of Outputs

To ensure the accuracy and quality of the ISS package, a robust quality control process should be implemented, including:

• Independent validation of SAS programs and output, verifying that they meet the requirements specified in the validation plan.
• Review of draft outputs by a separate team, identifying any discrepancies or issues that need to be addressed before finalizing the ISS package.

Conclusion

Creating ISS packages for compounds with multiple ongoing submissions can be a complex and time-consuming process. By establishing a reusable RSD Library and following best practices for dataset creation, harmonization, and validation, the process can be streamlined and made more efficient. Utilizing a centralized ISS team, adhering to established standards and templates, and implementing robust quality control processes can ensure the generation of high-quality and consistent ISS packages. Ultimately, these strategies can help to effectively address the unique challenges associated with creating ISS packages for oncology medications with multiple ongoing submissions, supporting the timely and successful approval of these life-saving treatments.