Patients were enrolled in PROpel regardless of the presence of HRR gene mutations (HRRm). The radiographic progression-free survival by HRRm status and HRRm determination, however, were predetermined objectives. Although TT testing was the standard method for determining HRRm, ctDNA testing was also carried out due to probable difficulties in getting a significant quantity of high-quality TT.
Over less than 98% of randomly assigned patients provided baseline blood and archival TT samples, which were then analyzed by the FoundationOne®CDx and FoundationOneLiquid®CDx molecular diagnostic assays, respectively. The HRRm status (HRRm, non-HRRm, or HRRm uncertain) was determined by each test result and by combining the findings of the two tests. Pts were labeled as HRRm for aggregate results if an HRRm was found by either test, non-HRRm for results in which no HRRm was found by either test, or HRRm unknown if no valid HRRm test result was available.
Of the 796 patients who were randomly assigned, 782 had TT samples and 794 had blood samples that could be tested. About 535 (68%) of the patients with TT samples had their HRRm status successfully evaluated by a TT test, whereas 734 (92%) of the patients with blood samples had their HRRm status successfully evaluated by a ctDNA test. A total of 778 (98%) patients (226 HRRm, 552 non-HRRm and 18 HRRm unknown) had their HRRm status determined. The positive and negative predictive values were 64% and 94%, respectively, using TT as a reference. The positive-percent agreement was 80%, the negative-percent agreement was 87%, and the overall-percent agreement was 85%. Around 51 participants in the HRRm by ctDNA subgroup were not HRRm by TT. According to the negative predictive value, ∼174/186 (94%) of the patients in the non-HRRm by ctDNA subgroup would also be non-HRRm by TT test. In the entire non-HRRm aggregate population, there were therefore ∼12 possible false negative findings by ctDNA test (2% of 552 points).
Between matched TT & ctDNA test findings in PROpel, there was high agreement. By maximizing the number of patients with known HRRm status while limiting the number of potentially false negative findings, it was feasible to do a rigorous analysis of HRRm status using aggregate test results.