Evaluation of PREDICT: a prognostic risk tool, after diagnosis of a second breast cancer
Menée à partir de données portant sur 363 933 femmes atteintes d'un premier ou d'un second cancer du sein, cette étude évalue la performance du modèle "PREDICT", basé sur des facteurs clinicopathologiques (âge au diagnostic, grade de la tumeur, nombre de ganglions atteints, statut ER, statut HER2, ...), pour prédire la mortalité spécifique à 5 ans après le diagnostic d'un second cancer du sein
Background : PREDICT is a clinical tool widely used to estimate the prognosis of early-stage breast cancer (BC). The performance of PREDICT for a second primary BC is unknown.
Method : Women 18 years and older, diagnosed with a first or second invasive BC between 2000-2013 and followed for at least 5 years were identified from the US Surveillance, Epidemiology, and End Results (SEER) database. Model calibration of PREDICT was evaluated by comparing predicted (P) and observed (O) 5-year BC-specific mortality separately by estrogen receptor (ER) status for first vs second BC. Receiver-operator curves and area under the curve (AUC) were used to assess model discrimination. Model performance was also evaluated for various races and ethnicities.
Results : The study population included 6,729 women diagnosed with a second BC and 357,204 women with a first BC. Overall, PREDICT demonstrated good discrimination for first and second BCs (AUCs ranging 0.73-0.82). PREDICT significantly underestimated 5-year BC mortality for second ER-positive BCs (P-O=-6.24%, 95%CI:-6.96%,-5.49%). Among women with a first ER-positive cancer, model calibration was good (P-O=-0.22%, 95%CI:-0.29%,-0.15%), except in Non-Hispanic Black (P-O=-2.33%, 95%CI:-2.65%,-2.01%) and women
≥
80 years of age (P-O=-3.75%, 95%CI:-4.12%,-3.41%). PREDICT performed well for second ER-negative cancers overall (P-O=-1.69%, 95%CI:-3.99%,0.16%), but underestimated mortality among those who previously received chemotherapy or had a first cancer with more aggressive tumor characteristics. In contrast, PREDICT overestimated mortality for first ER-negative cancers (P-O = 4.54%, 95%CI : 4.27%,4.86%).
Conclusion : PREDICT underestimated 5-year mortality after a second ER-positive BC and in certain subgroups of women with a second ER-negative BC.
JNCI Cancer Spectrum , article en libre accès, 2022