Prognosis of locally recurrent node-negative breast cancer according to type of primary surgery long-term results of a matched pair analysis

Objective: Most studies have found survival after local recurrence of breas t cancer to be poorer after mastectomy than after breast-conserving surgery . We studied whether this is the case because patients undergoing mastectom y have a more unfavorable risk profile and whether it applies also to patie nts with early stage disease. Methods: We conducted a matched-pair analysis (breast-conserving surgery vs . mastectomy) of 62 patients with local recurrence of early stage (pT1-pT2 NO) breast cancer without evidence of systemic disease. Patients were match ed for tumor size, histologic grade and age. Prognostic parameters at the t ime of primary diagnosis and at recurrence were evaluated with univariate a nd multivariate analyses with respect to survival after recurrence. The med ian follow-up was 7.5 years. Results: There were no significant differences between the two groups in th e size of the primary tumor, histologic grading, and size of recurrent tumo r. The disease-free interval between primary diagnosis and local recurrence was 8 months shorter in patients after mastectomy than in those after brea st-sparing surgery (p=0.35). Survival after recurrence was shorter in the p atients after mastectomy but this difference was not statistically signific ant (98 vs. 208 months, respectively; p=0.11). The disease-free interval be fore local recurrence (but not the type of primary surgery) was the only in dependent prognostic factor for subsequent survival in both the univariate (p = 0.02) and the multivariate (p=0.01) analyses. Conclusion: The difference in survival after local recurrence after mastect omy compared with breast-sparing surgery for early breast cancer is small a nd appears to be associated with different disease-free intervals before lo cal recurrence.