338P - Breast cancer follow-up: A population-based cohort study

Background

Breast cancer (BC) survivorship guidelines recommend follow-up (F/U) with a schedule of physician visits and annual mammography. However, literature reflects an increasing reliance on advanced imaging. Our goal was to analyze BC F/U practice with its roots and effects at a population level.

Methods

We included women ≥18 years with early BC undergoing surgery between 2012-2014. We accessed the data warehouse of Lombardy Region in Italy. After calculating the annual frequency of each imaging per patient, we included the derived covariates in unsupervised clustering to set three statistically different groups with increasing F/U intensity, ie minimal, moderate and maximal. We used the Kaplan-Meier method for overall survival (OS) estimates and the Cox model for multivariable analysis.

Results

Overall 28,775 women entered the study, median age was 61 years (IQR 49-72), median time since diagnosis 87.6 months (IQR 77.7-97.9), 69.7% had breast-conservative surgery. Of the 862,955 exams prescribed in the 5 years after surgery, 12.8% were mammograms. Ultrasound (64.7%) was the most common non recommended imaging, followed by computed tomography (11.7%), magnetic resonance (6.2%), and bone scan (4.6%). Intensive F/U was more likely after triple negative (30.1%) and HER2 positive (27.8%) than luminal-like BC (13.1%). Surveillance intensity decreased over time, resulting in 56% of BC survivors having minimal F/U from the third year after surgery onwords. At a 12 month landmark analysis, OS was 80.8%, 87.4% and 66.2% in women having minimal, medium, and maximal intensive F/U. The survival benefit of medium intensive F/U was significant in luminal-like; whereas, the negative effect of maximal intensive F/U extended to all subtypes and was independent of age, surgery, adjuvant and radiotherapy (HR 2.16, 1.99-2.36).

Conclusions

Use of non recommended imaging is common after BC surgery and increases with increasing risk of relapse. Due to the retrospective nature of our study, survival findings are only hypotheses generating. While intensive F/U possibly included women with poor prognostic factors not accounted for by the adjustment, medium intensive F/U might have allowed for timely/curative intervention of recurrent BC. Data overall offers an opportunity to discuss current F/U strategies.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Fondazione IRCCS Istituto Nazionale dei Tumori.

Funding

This study is part of the project “I controlli periodici (follow-up) dopo la diagnosi e le terapie in pazienti liberi da malattia e asintomatici: verso una personalizzazione delle strategie di follow-up” supported by Grant n. RR33 of Lombardy Region, Italy.

Disclosure

S. Di Cosimo: Financial Interests, Personal, Other, Ad hoc project/protocol reviewer: MEDSIR; Financial Interests, Personal, Advisory Board, Treatment of HER2-positive breast cancer: Pier Fabre; Financial Interests, Personal, Other, Medical meeting material setup and distribution; Speaker Bureau: IQVIA. All other authors have declared no conflicts of interest.