ABSTRACT
PURPOSE: The aim of this population-based study was to estimate short-term and long-term survival of interval breast cancers and to compare them to clinically detected cancers, taking into account prognostic features. METHODS: This study included all interval cancers and clinically detected cancers diagnosed in the Loire-Atlantique population-based cancer registry from 2000 to 2010 in women aged 50-76 years. We used the Pohar-Perme method to estimate 5- and 10 year net survival rates and a flexible parametric model to compare interval cancer and clinically detected cancer prognosis with and without adjustment for the main prognostic factors (age, stage, histological grade, and phenotype). RESULTS: This study included 813 interval cancers and 1,354 clinically detected cancers. Interval cancers were diagnosed at a significantly less advanced stage than clinically detected cancers, but more often with a triple-negative phenotype. Interval cancer age-standardised net survival was 88.0% at 5 years (95% CI 84.9-91.2) and 81.7% at 10 years (95% CI 76.9-86.9), whereas clinically detected cancer age-standardised net survival was 77.8% (95% CI 75.1-80.6) and 64.6% (95% CI 60.7-68.7), respectively. After adjustment for covariates, survival no longer differed between interval cancers and clinically detected cancers at 5 and 10 years. CONCLUSION: Although the interval cancer net survival rate was higher, interval cancers had a similar short-term and long-term prognosis than clinically detected cancers after taking into account the main prognostic factors.
