Miami Breast Cancer Conference® Abstracts Supplement
While the RxPONDER trial was a landmark study demonstrating that Oncotype DX (ODX) can guide adjuvant chemotherapy use in postmenopausal women with 1 to 3 positive nodes and an ODX score of 25 or less, ODX was not predictive in premenopausal women. Therefore, current NCCN guidelines recommend ODX as a consideration but leaves decision-making up to the discretion of the physician. We sought to determine the prognostic accuracy of Ataraxis Breast RISK (ATX-RISK), an artificial intelligence-based test that integrates clinical data with pathology features extracted from whole-slide H&E images.
ATX-RISK scores were generated for 150 patients with hormone receptor–positive (HR+)/ HER2-negative (HER2–) breast cancer who received standard-of-care treatment. Patients were classified into ATX high and low risk groups using a cutoff of 0.1. To compare the performance of ATX to ODX, analyses were narrowed to 43 patients who had ODX recurrence scores available. The primary end point was disease-free interval, evaluated using Kaplan-Meier curves with log-rank testing.
Among 150 patients who were premenopausal node-positive, ATX high patients had a lower 5-year disease-free interval (84%; 95% CI, 73%-90%) than ATX low patients (91%; 95% CI, 68%-98%). This difference was significant (Plog-rank = 0.04) (Figure 1A).
Among the 43 patients with both ATX and ODX scores, both scores were significantly associated with disease-free interval. However, ATX showed superior discriminatory performance (C-index= 0.69; 95% CI, 0.24-0.94), compared with ODX (C-index = 0.53; 95% CI, 0.14-0.88) (Figure 1B). At 5 years, 7% of patients concordantly classified as ATX low plus ODX low recurred, compared with 17% of ATX high plus ODX high patients. Notably, 4 of 23 (17%) patients in the discordant ATX high plus ODX low group experienced an event (Figure 1C). No patients were classified as ATX low plus ODX high.
Overall, of the 29 patients classified as ATX high, 5 (17%) recurred, compared with 1 of 14 (7%) of patients who were ATX low who recurred. In contrast, the recurrence rate between ODX groups was nearly identical where 5 of 37 (14%) of ODX low patients experienced a recurrence event within 5 years compared with 1 of 6 (17%) of ODX high patients. Of interest, 4 out of 6 (67%) of recurrence events occurred in the discordant ODX low plus ATX high group (Figure 1D).
ATX demonstrates promising prognostic performance in premenopausal node-positive patients with breast cancer and can identify aggressive disease not recognized by genomic assays.
