Factors associated with long-term survival of invasive breast cancer patients in Belgium: a population-based cohort study.

Background: Most women diagnosed with breast cancer (BC) survive treatment and become long-term survivors. This study examines the association between long-term survival of female BC patients and clinical factors, socio-economic determinants, healthcare utilization, and drug use.

Methods: This retrospective population-based survival study uses linked cancer registry data, claims, and social security data at the individual level for all Belgian women diagnosed with invasive BC in 2010 (n = 9982). Ten-year survival probabilities were computed by using Kaplan-Meier and relative survival curves. A Weibull mixture cure model was employed to perform multivariable analysis.

Results: The population consisted of women with early-stage (75.0%), locally advanced (13.1%), and metastatic BC (5.9%). The overall 10-year unadjusted survival was 68.2% [95% confidence interval (CI), 67.1%-69.3%], with decreasing survival in older women. However, survival was lower in women age <40 years compared with those aged 40-60 years. Multivariate analysis revealed that age was associated with long-term survivorship (odds ratio, 0.919; 95% CI, 0.918-0.919). Women receiving multimodal treatment (surgery followed by adjuvant radiotherapy and systemic treatment) were more likely to become long-term survivors compared with women receiving other treatments. Lastly, beneficiaries of increased reimbursement had lower odds of long-term survivorship.

Conclusion: Our study highlights the association between long-term survivorship in women diagnosed with invasive BC and factors such as age at diagnosis, treatment scheme, and entitlement to increased reimbursement. It is crucial to inform clinicians and policymakers on the appropriate use of mixture cure models, as their results can substantially impact healthcare decision-making. Furthermore, tailored BC screening strategies are essential to reduce social inequalities and mortality disparities.