PreciseBreast is prognostic for breast cancer recurrence in a 21-gene assay managed cohort

PreciseBreast (PDxBR) is an AI-based digital test that provides individualized risk prediction for invasive breast carcinoma (IBC) independent of a gene expression test and yields important prognostic information that may well guide decision making, especially in patients who do not have access to genomic assays. We aimed to assess the accuracy of risk prediction between 21-gene expression assay (21-GEA) and PDxBR for disease recurrence within six-years. This retrospective longitudinal study on patient data and samples utilizing COTA, Inc’s real-world database. Hematoxylin and eosin-stained tumor samples were sent from Miami Cancer Institute to PreciseDx for scanning and image processing using the PDxBR platform to generate an algorithm-validated disease recurrence score. PreciseDx remained blinded to 21-GEA results during the PDxBR risk assessment. Patients diagnosed with IBC between 2014 and 2018 with 21-GEA results and six-year distant- and local-recurrence outcome data were identified. Four hundred twenty-five patients with stage I-II, hormone receptor (HR)-positive, HER2-negative IBC were included. Risk stratification, adjuvant treatment choice, and event classification between PDxBR and 21-GEA were evaluated. Key outcomes included breast cancer-free interval and overall survival. The median age was 59 years (73% post-menopausal). Ninety percent of tumors were <2.5cm, 50% clinical grade 2, and 81% lymph node negative. Twenty percent of patients received adjuvant chemotherapy, 99% endocrine therapy, and 56% radiation therapy. 21-GEA classified 87% (n=371) as low-risk vs. PDxBR 79% (n=337) low-risk. There were 11 deaths, 19 distant recurrences, and 10 local-regional recurrences. The negative predictive value (low-risk for an event) for 21-GEA vs. PDxBR was 0.93, 0.93; positive predictive value (high-risk) was 0.26, 0.17 and hazard ratio was 4.26, 2.41 (p=0.22). The avoidance of adjuvant chemotherapy due to 21-GEA (82%) was similar to the PDxBR (77%) low-risk patients. 21-GEA classified 35% as high-risk vs. 38% for PDxBR. For distant and local-recurrence, 21-GEA identified 45% as high-risk vs. PDxBR 38%. PDxBR is comparable to 21-GEA in providing prognostic risk of recurrence for patients with IBC. Once validated in ongoing studies, applying PDxBR would enable individualized risk assessment and effective guidance of disease management, including adjuvant therapy choice.