Abstract P1-08-37: Magee equations predict pathologic response to neoadjuvant chemotherapy

Abstract

Abstract Introduction: Neoadjuvant chemotherapy is used in locally advanced breast cancer to downstage the tumor, facilitating surgical management. Oncotype DX (ODX) is used to estimate the risk of distant recurrence for ER-positive breast cancers, allowing selected patients to avoid the toxicity of chemotherapy. ODX is often not possible on the small core biopsy samples. Klein et al. have shown that standard histological variables, combined with semiquantitative ER, PR, HER-2, and Ki-67 results, can provide information similar to that with ODX, using equations derived by linear regression analysis (Magee equations). We applied a modification of these equations to pretreatment core biopsies in women who received neoadjuvant chemotherapy to determine if the risk scores were predictive of pathologic response. Methods: 25 patients who received chemotherapy for receptor positive locally advanced(21), inflammatory(3), or metastatic(1) breast cancer followed by surgical treatment of the primary site were identified from a prospective breast cancer database. Pretreatment core biopsies were reviewed by a breast pathologist and Nottingham grade, ER and PR status (% of cells staining and intensity of staining), and Her-2 status by IHC and/or FISH were recorded. Clinical tumor size was defined as the average of sizes derived from mammogram, ultrasound, MRI, PET-CT and clinical breast examination. Using these data in a modified Magee equation, the patient's recurrence score was calculated. 0-18 was considered low risk (LR), >18-<30 was considered intermediate risk (IR), and ≥30 was considered high risk (HR). Resection specimens were reviewed to define pathologic response. A good pathologic response to chemotherapy was defined as a complete pathologic response (3 cases), near complete response (2), or a response with one or more of the following; reduction in the post-treatment size of the tumor by greater than 50% compared with pretreatment imaging, a significant reduction in tumor cellularity in the tumor bed, and an inflammatory lymphohistiocytic infiltrate with tumor necrosis (6 cases). For the remaining 14 cases, the response was defined as poor (no histopathologic evidence of response to treatment). Risk scores were compared between good and poor responders using T-Test. Comparison between risk groups (HR vs IR vs LR) were made using Chi Square analysis. Results: Magee scores ranged from 13.8-41.6 (mean 27.4) and were significantly lower in the poor responders (mean = 23, range 13.8-41.6) compared to the good responders (mean = 33, range 22-41.3, p = 0.003). Table 1 shows the distribution of response by Risk Group (p = 0.018). Table 1: Response by Risk CategoryMagee Risk GroupLRIRHRPoor Response563Good Response038 73% of patients with high risk Magee scores had a good response to chemotherapy, compared to 21% of patient with low or intermediate scores (p = 0.01). Conclusions: Modified Magee equations applied to pretreatment core biopsies seem to predict pathologic response to neoadjuvant chemotherapy. Use of these equations to assign risk scores may be a useful tool in deciding which ER positive breast cancer patients are likely to benefit from preoperative chemotherapy for cytoreduction, and who should go directly to surgery. These findings need to be validated in larger studies. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P1-08-37.