Individual risk profiling for breast cancer recurrence: towards tailored follow-up schemes

Background: Breast cancer follow-up is not tailored to the risk of locoregional recurrences (LRRs) in individual patients or as a function of time. The objective of this study was to identify prognostic factors and to estimate individual and time-dependent LRR risk rates. Methods: Prognostic factors for LRR were identified by a scoping literature review, expert consultation, and stepwise multivariate regression analysis based on 5 years of data from women diagnosed with breast cancer in the Netherlands in 2005 or 2006 (n = 17 762). Inter-patient variability was elucidated by examples of 5-year risk profiles of average-, medium-, and high-risk patients, whereby 6-month interval risks were derived from regression estimates. Results: Eight prognostic factors were identified: age, tumour size, multifocality, gradation, adjuvant chemo-, adjuvant radiation-, hormonal therapy, and triple-negative receptor status. Risk profiles of the low-, average-, and high-risk example patients showed non-uniform distribution of recurrence risks (2.9, 7.6, and 9.2%, respectively, over a 5-year period). Conclusion: Individual risk profiles differ substantially in subgroups of patients defined by prognostic factors for recurrence and over time as defined in 6-month time intervals. To tailor follow-up schedules and to optimise allocation of scarce resources, risk factors, frequency, and duration of follow-up should be taken into account.