medRxivJClub
Reconstructing real-world metastatic lines of therapy enables progression risk stratification in breast cancer
medRxiv · · Preprint · Open access
Zhao, Niederhauser + 4 more
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
In the MSK-CHORD breast cancer cohort, reconstructing 8,791 metastatic lines of therapy from 2,881 patients enabled progression-risk prognostication with Antolini’s C = 0.681 ± 0.006.
- Why it matters: Progression-free survival by each metastatic line of therapy is valuable for precision oncology but is rarely available as paired labels, limiting risk stratification.
- What they did: The study built an evidence enrichment pipeline to reconstruct mLoTs and PFS from longitudinal EHR, then trained a progression-risk model evaluated in nested cross-validation.
- The result: The model reached integrated Brier score (IBS) = 0.124 ± 0.004 and, after GENIE BPC validation, matched curated regimens at 86.2%–82.4% with PFS event agreement 93.2%–94.6%.
The findingWhy it mattersWhat they didThe result
- Open access