Interpreting Real-World Data and Cross-Trial Comparisons in CLL Studies

Dr. Lipsky addresses the framework for interpreting real-world data and cross-trial comparisons in the absence of head-to-head trials. He acknowledges the inherent tension in oncology between wanting prospective data and relying on indirect comparisons, noting several important caveats for interpreting real-world evidence.

Real-world data is most valuable when answering questions ahead of prospective studies, for example, retreatment after time-limited therapy, where community experience may exceed available trial data. Key considerations when evaluating real-world datasets include data source quality, granularity of clinical information available (claims databases can capture broad outcomes but not specific toxicity details), and whether industry-sponsored analyses have access to proprietary granular data for one drug but only published summary data for the comparator.

He emphasizes looking for consistent trends across multiple real-world analyses rather than relying on single studies, noting that methodological inconsistencies across studies producing conflicting directional findings should prompt skepticism. When consistent trends emerge across independent datasets, this strengthens confidence in real-world observations.

Dr. Shadman notes that the best treatments only work if patients can remain on them, reinforcing that treatment selection must account for tolerability and adherence. He highlights the importance of the covalent BTKi class as the most important drug class for CLL, with second-generation agents acalabrutinib and zanubrutinib demonstrating significantly better tolerability than ibrutinib. Prospective data show patients can switch between covalent BTKis for intolerance, and rare patients intolerant of all covalent agents may tolerate pirtobrutinib. Dose modifications for both BTKis and BCL-2 inhibitors are important tools for keeping patients on effective therapy, particularly in elderly and frail populations.