Clinical predictors of residual breast disease after primary induction chemotherapy

Abstract

688 Background: The purpose of this study was to determine if pre-operative tumor characteristics can predict which patients are likely to have residual infiltrating breast cancer (RIBC) after undergoing primary induction chemotherapy. Methods: Data were collected prospectively for 85 patients with locally advanced breast cancer who underwent primary induction chemotherapy followed by surgery. Variables included age at diagnosis, hormone receptor status, pre-operative taxane, and cTNM tumor size, axillary node status, and stage. Recursive partitioning analysis was used to determine cutpoints in age or cTNM size that predict residual breast disease. The Chi-square test was used to identify significant univariable correlates of RIBC. A P-value of less than 0.05 was considered significant. Results: Age at diagnosis, cTNM axillary node status, and stage predicted RIBC in the pathologic specimen. RIBC was identified in 100% (n=19) of the surgical specimens of patients > 56 years of age compared to 79% (n=52 of 66) of patients ≤ 56 years (P=0.028). cTNM size, although not clinically significant (P=0.07), suggested increasing tumor size may correlate with RIBC: 72% (n=21 of 29) of tumors with clinical measurement of 1–6 cm had RIBC, compared to 88% (n=36 of 41) of those measuring 7–12 cm and 100% (n=11) of those measuring 13–20 cm. cTNM node status was available for 70 patients: 57% (n=4 of 7) of cN0 patients compared to 87% (n=55 of 63) of cN1 patients had RIBC (P=0.038). Higher cTNM stage predicted RIBC (P=0.026): 65% (n=11 of 17) of patients with cTNM stage IIA or IIB had RIBC compared to 80% (n=28 of 35) cTNM stage IIIA, 100% (n=17) of cTNM stage IIIB, and 94% (n=15 of 16) TNM stage IV. Hormone receptor status (P=0.76) and inclusion of a taxane derivative in the pre-operative chemotherapy course (P=0.39) were not significant predictors of residual disease. Conclusions: Patients diagnosed at more than 56 years of age, or who were cTNM node positive, or cTNM stage IIIB or IV were most likely to have RIBC. A trend was established that suggested cTNM size might predict RIBC. Neither hormone receptor status nor the addition of a taxane derivative to the induction chemotherapy regimen predicted RIBC. No significant financial relationships to disclose.