Data sources and applied methods for paclitaxel safety signal discernment

Area of Science:

• Cardiovascular devices
• Medical device safety
• Regulatory science

Background:

• A late mortality signal identified with drug-coated balloons (DCB) and drug-eluting stents (DES) prompted FDA review.
• An FDA advisory panel concluded existing clinical data are insufficient for comprehensive evaluation of this signal.
• Paclitaxel-coated DCBs and DESs are the focus of this safety signal investigation.

Purpose of the Study:

• To identify and summarize clinical studies on paclitaxel-coated DCBs and DESs.
• To assess the quality of available data sources for evaluating these devices.
• To present methodologies for effective safety signal discernment.

Main Methods:

• Searched for English-language studies published January 2007-November 2021 with >36 months follow-up.
• Included randomized trials, cohort studies, registries, claims, and electronic health records.
• Assessed data quality and described advanced statistical methodologies for signal detection.

Main Results:

• Identified 39 studies and data sources, including 19 randomized trials and 8 registries.
• Summarized methodologies such as individual-level data analysis, machine learning, and Bayesian approaches.
• Highlighted the need for advanced statistical techniques to analyze diverse data sources.

Conclusions:

• Despite numerous studies, existing data are insufficient to fully assess the late mortality signal associated with DCBs and DESs.
• High-quality data sources and advanced statistical methods are crucial for timely detection of safety signals.
• Enhanced data analysis is needed for regulatory bodies to act promptly on detected safety concerns.