Prognostic Impact of Time to Ipsilateral Breast Tumor Recurrence after Breast Conserving Surgery

BACKGROUND: The poor prognosis of patients who experience ipsilateral breast tumor recurrence (IBTR) after breast conserving surgery (BCS) is established. A short time between primary cancer and IBTR is a prognostic factor but no clinically relevant threshold was determined. Classification of IBTR may help tailor treatment strategies. PURPOSE: We determined a specific time frame, which differentiates IBTR into early and late recurrence, and identified prognostic factors for patients with IBTR at time of the recurrence. METHODS: We analyzed 2209 patients with IBTR after BCS. We applied the optimal cut-points method for survival data to determine the cut-off times to IBTR. A subgroup analysis was performed by hormone receptor (HR) status. Survival analyses were performed using a Cox proportional hazard model to determine clinical features associated with distant-disease-free survival (DDFS) after IBTR. We therefor built decision trees. RESULTS: On the 828 metastatic events observed, the majority occurred within the first 3 months after IBTR: 157 in the HR positive group, 98 in the HR negative group. We found different prognostic times to IBTR: 49 months in the HR positive group, 33 in the HR negative group. After multivariate analysis, time to IBTR was the first discriminant prognostic factor in both groups (HR 0.65 CI95% [0.54–0.79] and 0.42 [0.30–0.57] respectively). The other following variables were significantly correlated with the DDFS: the initial number of positive lymph nodes for both groups, the initial tumor size and grade for HR positive tumors. CONCLUSION: A short interval time to IBTR is the strongest factor of poor prognosis and reflects occult distant disease. It would appear that prognosis after IBTR depends more on clinical and histological parameters than on surgical treatment. A prospective trial in a low-risk group of patients to validate the safety of salvage BCS instead of mastectomy in IBTR is needed.