medRxivJClub
Individualised and Time-Dependent Prediction of Recurrence After Colorectal Cancer Liver Metastases Resection
medRxiv · · Preprint · Open access
Viljamaa, Baumgartner + 12 more
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
Recalibrated GAME, m-CS, Fong, Nordlinger, Nagashima, Konopke, and Helsinki scores for colorectal cancer liver metastases resection yield time-dependent recurrence-free survival predictions with AUC 0.63–0.74 at 3 years.
- Why it matters: Clinicians need externally validated, individualised recurrence probabilities to benchmark newer prognostic models, because existing clinical risk scores are hard to apply and usually provide only categorical strata.
- What they did: Using recurrence-free survival in 709 patients from the KaroLiver cohort (2013–2021; follow-up to 2025) and external validation in 198 patients from OSLO-COMET (2012–2016; follow-up to 2022), the authors assessed seven established scores and recalibrated them into probabilistic, time-dependent models.
- The result: The recalibrated models showed good predicted–observed agreement and broadly comparable net benefit in OSLO-COMET, and were implemented in PLATO to support patient communication, risk-adapted surveillance, and transparent benchmarking.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.
- Open access