Changes in temporal patterns of local failure after breast-conserving therapy and their prognostic implications.

Purpose: The purpose of this analysis was to evaluate patterns and rates of ipsilateral breast tumor recurrence (IBTR) over time based on the type of failure (true recurrence/marginal miss [TR/MM] vs. elsewhere [E]) and to compare these to rates of contralateral failure in women with Stages I/II breast cancer treated with conservative surgery (CS) and radiation therapy (RT).
Methods and Materials: Between 1980 and 1997, 1448 patients with Stages I/II invasive breast carcinoma were treated with CS and RT to a median total dose of >60 Gy. IBTRs were analyzed in terms of their location as follows: TR/MM, referring to those occurring in the same quadrant, and E, referring to the development of a malignancy remote from the index lesion. Outcomes were characterized in terms of their temporal recurrence patterns and correlated with rates of contralateral breast failure (CL). The median follow-up was 8.5 years.
Results: A total of 79 ipsilateral and 98 contralateral failures were observed during the follow-up period corresponding to 5-, 10-, and 15-year actuarial rates of 2%, 7%, and 10% and 4%, 9%, and 12% respectively. Fifty-nine ipsilateral failures (74.7%) were designated as TR/MM and 20 (25.3%) as E. The corresponding 5-, 10-, and 15-year actuarial rates of TR/MM and E failures were 2%, 5%, and 8% and 0.1%, 2%, and 3%, respectively. The median times to total ipsilateral, TR/MM, E, and contralateral failure were 6.5, 5.7, 7.4, and 5.2 years, respectively. Between 0 and 5 years of follow-up, E failures represented 7% of the total ipsilateral failures observed. From 5-10 and 10-15 years, E failures represented 39% and 27% of the total, respectively (p = 0.01). Contralateral failure was significantly more likely to occur in the first 5 years of follow-up than IBTR, accounting for 50 (63.3%) of the 79 failures in that interval (p = 0.02). No significant differences in survival rates were detected between any of the various failure types. On multivariate analysis, only reduced time to tumor recurrence was found to be adversely associated with overall survival.
Conclusions: The rates and patterns of IBTR vary with time and, after 5 years, approach the rates of development of a contralateral breast cancer. E failures are, overall, less frequent than TR/MM but contribute increasingly to the IBTR rate after 5 years. Time to tumor recurrence is the most reliable predictor of prognosis after IBTR.