For biotech and pharmaceutical companies with a clinical development program, understanding the safety and efficacy of a therapy is a key success factor. One of the primary sources of safety information is the scientific literature, and thus, it its imperative to ensure monitoring through all phases of the clinical pipeline.
Monitoring the scientific literature is in fact a regulatory requirement, and missing important adverse events and possible new risks documented in the literature can lead to unsuccessful submissions, delays and inspection findings incurring significant additional effort.
The FDA sets out industry guidance for monitoring and reporting of safety events for drugs and biological products under clinical development in the document Sponsor Responsibilities - Safety Reporting Requirements and Safety Assessment for IND and Bioavailability/Bioequivalence Studies, stating:
[…] sponsor-investigators are responsible for evaluating all safety information available to them, including data from reports in the scientific literature […].
Monitoring the scientific literature in a way that ensures compliance and avoids missing important safety data is challenging: there is a growing volume of articles being published every year and process auditability imposes significant “paperwork” burden on safety teams.
This process is the responsibility of the company, and biotechs are increasing looking to technology and automation to address these challenges. In this case study, we introduce the biologit MLM-AI Platform and use it to effectively screen the scientific literature for adverse events and possible new risks in the literature for CAR T-Cell therapies.
CASE STUDY: CAR T-Cell Therapies
CAR T-cell (or chimeric antigen receptor T-cell) is a type of immunotherapy used to treat certain cancers in particular malignancies like leukemia and lymphoma. CAR T-Cell therapy is an area of growing interest in clinical development, and a particularly active research area with over 600,000 patents having been filed in the last three years.
Literature Screening of Adverse Events and Possible New Risks
This case study applied the biologit Database - a unique, global scientific literature resource covering over 110,000 journals and 135 countries. Its comprehensive reach delivers highly relevant, de-duplicated results for analysis directly into the biologit platform.
The query was designed to include all variants of CAR T-cell products, including generic names and brand names ensuring results with high recall. A total of 1020 articles were retrieved for a period of 3 months.
Using the Platform’s unique AI capabilities, articles can be prioritised based on the presence of safety-relevant information and irrelevant ones removed. From the original result set 660 articles were automatically excluded as not relevant, containing no safety information.
Further refinement is possible by using models that detect the presence of case reports - a relevant criteria for ICSR (Individual Case Safety Report) assessments - resulting in 149 articles of interest. The assessment of these articles by pharmacovigilance specialists resulted in 24 articles containing valid ICSRs.
The assessment workflow in the platform is facilitated by tailor made and customisable highlighters further improving speed and quality of decisions made by specialists:
Conclusions
AI technologies can be successfully deployed in a safe and compliant fashion for pharmacovigilance use cases, improving the literature review process for safety surveillance.
This case study illustrates how AI capabilities, coupled with an easy to use workflow and comprehensive database available in the biologit platform delivers high quality results without additional overhead to specialists, and significant savings to busy pharmacovigilance teams.
Comprehensive Literature Monitoring with biologit
Ensuring literature monitoring processes are compliant and important safety data is not missed is challenging: the volume of articles being published grows every year, and auditable processes imposes significant “paperwork” burden on safety teams. Manual methods of literature monitoring are resource-intensive, time consuming and carry a risk of an important safety event being missed.
biologit MLM-AI is a complete platform for completing regulated literature monitoring activities effectively and in full compliance. It is designed from the ground up with the needs of safety teams. To learn more get in touch and book a demo today.
biologit Database: Scientific Literature Database for your Regulatory Searches
The biologit Database is a comprehensive scientific literature resource with global and regional reach: updated daily, hosting over 50 million citations and 110,000 journals and fully integrated into the platform. Broad, high quality literature searches are accomplished seamlessly, and results delivered for assessment free of duplicates.
Easy to use SaaS + Fully Configurable Workflows = Productive Teams
From a simple to use web enabled interface, product configuration and assessments are easily completed in a fully auditable environment. A flexible workflow designed for safety monitoring teams supports any literature screening process, and unique AI-powered automation features ensure productive workflows for busy teams, delivering up to 70% efficiency gains compared to manual processes.
Compliance as Standard
biologit MLM-AI Platform is a fully validated and compliant platform built for strict CFR-11 traceability and security requirements. Extensive audit ready features including activity logging, reporting and fine-grained access controls.