Validation of the Performance of the Novel Prognostic Staging System for Overall Survival in De Novo Metastatic Breast Cancer and Demonstration of Performance for Cancer-Specific Outcomes.

INTRODUCTION

A prognostic staging system for de novo metastatic breast cancer was developed by Plichta et al. We aimed to validate this system within a more contemporary cohort; assess model performance with respect to disease-specific survival (DSS), progression-free survival (PFS), and distant PFS (DPFS); and evaluate the impact of additional data on all possible metastatic sites.

METHODS

A retrospective review was conducted of patients with stage IV breast cancer in our institutional cancer registry from 2010 to 2022. Kaplan-Meier curves with log-rank tests were used; model discrimination was estimated using the C-statistic. Multivariable Cox proportional hazards regression was used to compute adjusted hazard ratios (HRs). The analysis was repeated with modified definitions incorporating all possible metastatic sites.

RESULTS

Overall, 425 patients were included. The staging system showed fair discrimination at 3 years for overall survival (OS; C-statistic 0.64), DSS (0.65), PFS (0.60), and DPFS (0.60). With a median follow-up of 41 months, stage IVA-D OS was 84%, 79%, 59%, and 47% at 3 years, respectively. Use of modified definitions upstaged 13.1% of patients but did not change the model fit (C-statistic 0.64). Patients reclassified from A/B to C/D had worse OS than patients remaining A/B (adjusted HR 1.59) and better survival than patients who were C/D under both definitions (HR 0.61).

CONCLUSIONS

The novel prognostic staging system provided meaningful discrimination in OS as well as for DSS, PFS, and DPFS. Inclusion of all metastatic sites may improve model calibration. OS has improved in more recent years.