Abstract P5-02-05: Prediction of response to neoadjuvant chemotherapy in ER+ breast cancer: Magee equation recurrence versus oncotype DX® recurrence score

Abstract

Abstract Background and Aim: The Magee equation recurrence score (MS) based on tumor pathological characteristics can be used to estimate the actual Oncotype DX® recurrence score (RS). This study's aim was to test the correlation of MS with RS and their usefulness in predicting tumor response to neoadjuvant chemotherapy (NCT) in patients with estrogen receptor (ER) positive, HER2 negative invasive breast carcinoma (BC). Methods: Pathological data required for MS calculation such as Nottingham Score, H-scores for ER and PR, HER2, Ki 67 and tumor size were obtained via pre-therapy slide review or pathology/electronic medical record. Actual RS was measured (at Genomic Health) on core biopsies from 60 patients who received NCT. Pre-therapy tumor size was measured using imaging. Percentage tumor volume reduction (%TR) was measured. %TR was based on pre-therapy tumor size by MRI and US and detailed pathology evaluation of the resection specimen. Substantial %TR was defined as at least 50% reduction in tumor size (>50%TR). MS was calculated using Magee equation (http://path.upmc.edu/onlineTools/ptvr.html). MS and RS were categorized by score <18 (referred as MS18 and RS18, respectively) and ≥31 (referred as MS31 and RS31, respectively). Correlation between MS, RS and >50%TR were determined by using the Spearman's correlation coefficient. The mean levels of pathological parameters and score levels which had >50%TR or <50%TR were compared by ANOVA at a 2-sided α level of 0.05. The impact of the scoring systems for >50%TR by the scoring systems was assessed by calculating the area under (AUC) the receiver operating characteristic curve (ROC); AUC values between 0.7–0.8 represent considerable discrimination. Results: The mean patient age was 52 ±13 years. The mean pre-NCT tumor size was 48 ±36 mm.>50%TR was observed in 27 (45%) patients, but there was no pathological complete response. Low level of ER H-score (<100) was significantly associated with >50%TR (p=0.01). The Spearman's correlation coefficient (rs) for the MS and RS was 0.522 (p<0.001) with an average coefficient of determination (r2) of 0.298. Neither, MS nor RS correlated with >50%TR (p=0.08, p=0.2, respectively). The AUC values were calculated from the MS, RS, MS18, RS18, MS31 and RS31 (AUC=0.625, 0.602, 0.589, 0.572, 0.525 and 0.522, respectively).Conclusions: MS and Oncotype DX® RS have a good correlation in this small ER+ neoadjuvant cohort. However, neither MS nor RS correlated with %TR. Lower ER expression as a single variable is an excellent predictor of response to NCT and may be used alone in making therapy decisions in small number of cases. For moderate to strongly ER+ cases, the decision between primary surgery versus neoadjuvant systemic therapy may be difficult, but MS may be used in lieu of actual Oncotype DX® RS when it is not available. Citation Format: Soran A, Johnson R, Menekse E, Bhargava R, McGuire K, McAuliffe P, Ahrendt G, Dabbs D, Bonaventura M. Prediction of response to neoadjuvant chemotherapy in ER+ breast cancer: Magee equation recurrence versus oncotype DX® recurrence score. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-02-05.