RSNA: RECIST, Choi, and Modified Choi Criteria Fail to Predict Progression-Free Survival in Everolimus-Treated RCC
“Neither RECIST, CHOI nor mCHOI criteria could discriminate responders from non-responder patients in the group treated with everolimus,” reported a team of French coauthors led by Samar A. Raslan, MD.
The authors reviewed computed tomography (CT) images from 105 patients with mRCC who participated in the RECORD-1 trial. During RECORD-1, RECIST criteria were assessed using CT images taken at baseline and every 2 months to assess progression-free survival (PFS). The authors of the new analysis used computer software to reassess partial response (PR) from RECORD-1's CT images using Choi criteria (which defines PR as a ≥10% decrease in tumor size or ≥15% decrease in CT attenuation) and mChoi criteria, which defines PR as a ≥10% decrease in tumor size and ≥15% decrease in attenuation. They then compared Choi and mChoi criteria results with patients' PFS times.
When both placebo- and everolimus-arm patients were included in the analysis, “2D CHOI criteria identified 25% of PR without differences for median PFS compared to non-responders (2.3 vs 1.8 months, respectively), while mCHOI criteria found 0% of PR,” Dr. Raslan and coauthors reported.
Among patients in the everolimus treatment arm, “2% and 98% of patients were considered PR and SD (stable disease), respectively with RECIST criteria (median PFS – 5.5 months for both),” the coauthors reported. “48% of patients were considered PR with 2D CHOI criteria without differences for median PFS compared to non-responders (5.4 and 5.8 months) while mCHOI found 8% PR without significant differences for PFS (5.4 and 4.6 months).”
Similar results were obtained using 3D assessments of the attenuation Choi criterion, Dr. Raslan noted.