Tailored follow-up for early breast cancer patients: A prognostic index that predicts locoregional recurrence

Aims: After treatment, early breast cancer patients undergo follow-up according to standard regimens. After the first year, the main goal is particularly to detect locoregional recurrences (LRR). Our aim was to develop a simple prognostic index to predict LRR to tailor the follow-up programme. Methods: We used data from four large international clinical randomised trials and constructed the prognostic index using Cox proportional hazards regression. The bootstrap (a resampling method) was used for internal validation. Results: A total of 6516 patients treated according to current guidelines with complete covariable information were used for analysis. Covariables important for LRR in patients treated with breast conserving therapy were age, pathological tumour status, boost and surgical margins. The same variables were important for patients treated with a mastectomy, however, instead of the boost, the pathological nodal status was important. The index is composed to consist of three groups based on LRR risk after 10-years. Conclusions: We constructed a simple prognostic index that can be used to estimate risks of LRR in patients with early breast cancer. The prognostic index enables patients to be stratified into three subgroups with different outcomes with regard to LRR. (C) 2010 Elsevier Ltd. All rights reserved.