Medullary carcinoma of the breast: a population-based perspective

Abstract

Prognostic factors specific to medullary carcinoma of the breast (MCB) are unknown. Our objective was to identify patient and tumor factors predictive of overall survival (OS) in a large cohort of MCB patients. The Surveillance, Epidemiology, and End Results database was used to identify patients with MCB diagnosed from 1988 to 2004. Patient, tumor, and treatment factors were compared by univariate analysis via the Kaplan–Meier method and survival differences detected using the log-rank test. A multivariate Cox proportional hazards model controlled for patient age, race, type of surgery, radiotherapy, tumor size, number of lymph node metastases (LNM), lymph node yield (LNY), estrogen receptor (ER) and progesterone receptor (PR) status, and extent of disease. On univariate analysis of 3,348 patients, factors influencing OS included age, race, tumor size, ER status, type of surgery, radiotherapy, LNM, LNY, and extent of disease (P<0.001). On multivariate analysis, advancing age (P<0.001), black race (P<0.001), regional metastases (P<0.001), distant metastases (P<0.001), increasing tumor size (P<0.001), ER positivity (P=0.003), and increasing LNM (P<0.001) were associated with decreased OS. An OS benefit was seen in PR-positive patients (P=0.002) and in those with increasing LNY (P<0.001). Even among node-negative patients, increasing LNY was associated with improved OS (P<0.001). Tumor size, LNM, regional and distant metastases, PR status, age, and race are important prognostic factors in MCB. ER positivity was associated with decreased OS, which may reflect inaccuracy in diagnosing MCB or a significant biologic variant. The improved OS seen with increasing LNY in node-negative patients suggests MCB may be currently understaged.