236P Predicting Oncotype DX Recurrence Scores Using Clinicopathologic Features: Tennessee University Nomogram: an External Validation

The 21 gene recurrence score (RS) assay, Oncotype Dx (ODX) was developed to predict the benefits of chemotherapy in hormonal receptor (HR) positive early breast cancer (EBC). The Tennessee University Nomogram (UTMC), based on clinic-pathologic features (tumor grade, progesterone receptor, histologic type and age). estimate the probability that a patient's (p) with EBC (stage I-II), has a low-risk (LR 1-25) or a high-risk (HR-26-100) ODX-RS in female (p) >50 years old; in ≤50 years old, is applicable only for estimation of the probability for a HR-ODX RS. (Orucevic A et al The Breast (2019) 46:116-125). The aim of our study was to evaluate the concordance between UTCM and ODX in our population. From September 2012 to December 2019, we retrospectively analyzed a cohort of 199 pts >50 years with (EBC) pT1b-T2, pN0, HR positive, Her2 negative with ODX assay at the Basurto University Hospital (Spain). Clinico-pathologic variables were abstracted from pathology reports: Tumor size, Grade, Progesterone receptor and histologic type. We analyzed the correlation between ODX score and the UTMC in our population by the Pearson correlation. N=199, median age was 62 y (range: 51-78). Stage IA 83.4%. Grade 1, 2, and 3 were 13%, 74%, and 11%, respectively; half of p had Ki 67 <20%. Most p were ER+/PR+ (91.5%), LR-ODX RS: 171 p (85.9%) and HR-ODX RS: 28 p (14.1%). UTMC was calculated in 195 p. The probability for a LR-ODX-RS was in 165 p (with a 95% confidence between 85%-100%). In our population when the nomogram calculated probability is within 85% to 100% range, the calculator correctly assigns 82.05% of cases to the LR-ODX: 149/165 p or HR-ODX RS: 11/27 p. Our study found a moderate correlation between ODX RS and UMTC (Pearson correlation 0.484). In our serie, Luminal EBC pT1-2. pN0, The Tennessee University Nomogram (UTMC) correctly assigns 82.05% of cases to the LR-ODX when the nomogram calculated probability is within 85% to 100% range. The UMTC may be useful in health care systems with limited resources but we could not validate it as a universal alternative for OncotypeDx.