Predictors of local recurrence after conservative surgery and whole-breast irradiation

Purpose. To identify independent factors associated with increased risk of local recurrence (LR) in patients with breast cancer treated with conservative surgery and radiotherapy with or without systemic therapy. Methods and materials. Between January 1997 and December 2001, 969 women were treated at the Radiation Oncology Department in Chieti. We retrospectively analyzed 802 of them who were treated with conservative surgery and whole breast irradiation with or without systemic therapy. Tangential fields delivering 50 Gy to the whole breast were used and a boost was added for a total dose of 60 Gy. chi(2)-test or Fisher's exact test were used to identify independent significant factors that are predictive for LR. Kaplan - Meier method was used to calculate the 8-year rates of recurrence according to age, histologic findings, tumor size, number of positive nodes, margin status, receptor status and systemic therapy use: log-rank test was used to compare these curves. Cox proportional hazard model was used to obtain hazard ratios and 95% CI of LR for each covariate. Results. Median follow-up time was 63.1 months. LR occurred in 33 (4.1%) of 802 patients. Percentage of LR was greater in < 50 year-olds compared with 50 - 64 year-olds and 65 year-olds (9.8% versus 4.1 and 2.0%, respectively). LR was 18.8% in women with a tumor size > 3 cm versus 3.5, 4.0, 5.5% in women with a tumor size of 0.1 - 1, 1.1 - 2, 2.1 - 3 cm, respectively. The 8-year LR rate calculated with Kaplan-Meier method was 6.54 +/- 1.51. Multivariate Cox regression analysis showed that independent significant factors that are predictive for LR were: age < 50, tumor size > 3 cm, positive margin or unknown status, and hormonal therapy alone versus chemotherapy or combined therapy. Conclusions. Age and tumor size were the most important and statistically significant factors that correlated independently with higher rates of LR. Women < 50 years old and with a tumor size > 3 cm had a higher risk of LR. Also margin status and systemic therapy could influence LR risk.