Abstract P3-08-47: Prediction of the high-risk oncotype Dx recurrence score from clinicopathologic factors: A report from a cancer reference center in Brazil

Abstract

Abstract Background: Oncotype DX Recurrence Score (ODXRS) is a 21-gene test that has been validated as an assay to effectively predict recurrence and chemotherapy benefit for hormone receptor positive (HR+), human epidermal growth factor receptor 2 negative (HER2-) and lymph node negative (pN0) breast cancer. It has also been used in patients with 1-3 positive lymph nodes (pN1) in clinical practice. Methods: A single-institution retrospective study was conducted to include all patients who had an ODXRS test in our cancer center from 2005 to 2019. Clinical and pathology data were extracted from electronic charts. The predicted ODXRS was recorded and correlated with selected clinicopathological characteristics potentially related to a high-risk profile. A multivariate logistic regression model was performed to identify independent predictors of high-risk ODXRS (≥26) and ROC analysis was done to provide performance of the final model, found through exhaustive screening using Information Criteria (AIC). Associations with p<0.05 were considered to be statistically significant. Results: Two hundred and seventy-five patients with estrogen receptor positive, HER2 negative, pN0 or pN1 breast cancer were included. Median age was 55 years (range, 29 - 79). The ODXRS was low (<11), intermediate (11-25) and high (>25) in 58 (21.1%), 178 (64.7%) and 39 (14.2%) patients, respectively. On univariate and multivariate analyses, progesterone receptor negative status (p<0.001 for both), high histological grade (p<0.001 for both) and KI-67 proliferation rate (p<0.001 and p=0.002, respectively) were statistically significantly associated with high-risk ODXRS. In our prediction model, high-risk ODXRS could be predicted with a sensitivity of 67%, specificity of 83%, and an area under the curve of 0.82. Age, multifocal disease, tumor histological type, T-stage, N-stage, lymphovascular invasion and body mass index were not associated with ODXRS (Table 1). Conclusions: In our study, a prediction model based on progesterone receptor status, histological grade and KI-67 proliferation rate can be used to predict high-risk ODXRS. In resource-constrained healthcare systems, such model might help identify high-risk patients who would benefit from adjuvant chemotherapy without incurring the costs of the ODXRS test. Multivariable logistic regression Models for predicting High-Risk (>25) scores.Clinical and pathological variablesUnivariate OR (95%CI) pMultivariate OR (95%CI) pPatient age ≤ 50> 501.16 (0.58-2.36)0.682Multifocal tumorAbsent Present1.78 (0.81-3.70)0.1332.38 (0.86-6.34)0.085HistologyDuctalLobular1.14 (0.43 – 2.69)0.778Other0.65 (0.18 – 1.79)0.447T staging T1T2/31.25 (0.53-2.72)0.595N stagingN0N1 0.62 (0.20-1.55)0.347High-grade histologyG1/2G310.44 (3.97-28.25)<0.00110.51 (3.11-36.68)<0.001Lymphovascular invasionAbsentPresent1.18 (0.51-2.55)0.6881.15 (0.43 – 2.88)0.765PR positiveNegativePositive0.18 (0.07-0.47)<0.0010.10 (0.03-0.32)<0.001Ki-67, mean (±SD)1.06 (1.03-1.09)<0.0011.06 (1.02-1.09)0.002BMI (Kg/m2), mean (±SD)0.97 (0.88-1.05)0.479 Citation Format: Leandro JonataCarvalho Oliveira, Thais BacciliCury Megid, Marina Sahade, Andrea Kazumi Shimada, Daniele Xavier Assad, Joao Vicente Horvat, Antonildes Nascimento Assunção, Artur Katz, Max Senna Mano. Prediction of the high-risk oncotype Dx recurrence score from clinicopathologic factors: A report from a cancer reference center in Brazil [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-08-47.