Prognostic classification of ipsilateral breast tumor recurrence (IBTR).

Abstract

137 Background: IBTR after breast conservation encompasses true recurrence (TR) and new primary cancer (NP). No clear criteria distinguish TR from NP, but there is agreement that NP tumors have better outcomes than TR. Prior studies have used distance of IBTR from index cancer (IC), time to IBTR, histological, immunohistochemical (IHC) and genetic differences, but all data are not often available. We have examined IBTR patterns with the goal of identifying the simplest, most robust determinants of outcomes following IBTR. Methods: We reviewed records of breast cancer patients diagnosed with IBTR at the Lynn Sage Breast Center from 8/1992 to 6/2010. Data for the IBTR and IC were reviewed for histology, IHC, location, time between IC and IBTR, follow-up status, and cause of death. Parameters were scored as 1 if IBTR and IC were similar, and 0 if different (location=1 if ≤3cm; IHC=1 if hormone receptors and HER2 similar; interval=1 if ≤ 4 years). Univariate and multivariate proportional hazard models were used to determine impact on overall survival (OS), disease-specific survival (DSS), recurrence-free survival (RFS), distant recurrence free survival (DRFS) and local recurrence (LR). The multivariate model included significant univariate parameters. Results: We identified 161 patients with IBTR and complete data on ≥3 parameters; post-IBTR median follow up was 25 months. Data were missing on location in 13%, histology in 9%, IHC in 26%, and time interval in 0%. In univariate analysis, short interval to IBTR significantly decreased OS (HR 2.56, p=0.04), DSS (HR 4.31, p=0.009), RFS (HR 2.25, p=0.01), DRFS (HR 2.53, p=0.02), LR (HR 2.28, p=0.02); close location of IBTR decreased OS (HR 2.68, p=0.04). Histology and receptor status had no significant impact on the outcomes. Multivariate analysis included time and location, time ≤ 4 years was shown to decrease DSS (HR 4.00, p= 0.04), RFS (HR 2.32, p=0.03) and LR (HR 2.41, p=0.03). Conclusions: A short time interval between IC and IBTR is the most important prognostic parameter; location of IBTR within 3 cm of the IC also increases HR of subsequent events. These are the most easily available parameters when evaluating patients with IBTR, and therefore the most useful for distinction of TR versus new primary.