What are RECIST 1.1 progressions made of? Variability in double-read oncology trials

Area of Science:

• Oncology Clinical Trials
• Radiology and Imaging
• Biostatistics

Background:

• Blinded Independent Central Review (BICR) with double reads and adjudication is vital for data quality in imaging-based clinical trials.
• Discrepancies in assessing progressive disease (PD) using RECIST 1.1 criteria can impact trial outcomes, especially in Phase 3 oncology studies.
• Understanding discordance in the date of progressive disease (DoPD) across RECIST components is crucial for accurate survival analysis.

Purpose of the Study:

• To examine discordance in the date of progressive disease (DoPD) across RECIST components (target lesion, non-target lesion, new lesion).
• To explore the impact of these discrepancies on survival curves in oncology clinical trials.
• To determine if discrepancies stem from timing differences or true/false PD detection.

Main Methods:

• Retrospective analysis of data from five clinical trials involving 1932 lung cancer patients.
• Utilized Blinded Independent Central Review (BICR) with double reads and adjudication.
• Examined RECIST components to assess DoPD concordance, discrepancies, adjudicator acceptance, detection timing, and impact on survival curves.

Main Results:

• A 39.3% discordance rate in DoPD assessments was observed between readers.
• New lesions (NL) were the primary cause of discordance (41.4%), followed by target lesion (TL) diameter increase (33.3%).
• In 49.2% of discrepant cases, PD was reported late, often within one treatment cycle, and 62.5% of disputed PD cases were accepted by adjudication.

Conclusions:

• Different RECIST components contribute variably to discordance and adjudicator acceptance.
• New lesion detection is a key indicator of progression but a significant source of inter-reader disagreement.
• Integrating multiple RECIST components enhances the reliability of progressive disease assessment in clinical trials.