Search papers, labs, and topics across Lattice.
This paper introduces CACSurv, a novel framework for cancer survival prediction that leverages large language models (LLMs) to analyze patient reports, addressing critical mismatches in survival evaluation. By reformulating survival modeling as mini-cohort comparative reasoning, CACSurv utilizes concordance-aligned rewards to effectively rank patients based on prognostic orderings, even in the presence of censored data. The method outperforms existing models, achieving the highest C-index across six cancer cohorts, demonstrating a significant advancement in report-centric survival prediction.
CACSurv redefines cancer survival prediction by turning patient reports into a powerful tool for ranking prognostic outcomes, achieving a 6.5 percentage point improvement over the best existing model.
Cancer survival prediction supports treatment planning, risk stratification, and follow-up management. Existing methods use structured clinical variables, whole-slide images, genomic profiles, or multimodal inputs, while patient reports remain underexplored. We study report-centric survival prediction using reports that organize pathological, clinical, and molecular evidence. Large language models (LLMs) can reason over such reports, but case-wise time regression introduces two mismatches. First, a formulation mismatch arises because survival evaluation depends on ordering comparable patients, whereas independent time predictions do not enforce ranking consistency. Second, a supervision mismatch arises because a censored patient's observed time indicates survival beyond that point and cannot serve as an exact regression target, although it still implies orderings relative to patients who died earlier. To address these mismatches, we propose CACSurv, a Concordance-Aligned Comparative framework for report-centric survival prediction. CACSurv reformulates survival modeling as mini-cohort comparative reasoning, where an LLM predicts relative prognostic orderings. We introduce concordance-aligned rewards derived from comparable relations under right censoring, enabling censored outcomes to provide ranking supervision without exact event-time targets. At inference, Monte Carlo Reference Aggregation compares each patient with sampled references and aggregates positions into a cohort-level ranking. We establish TCGA-SurvReport, a benchmark covering six TCGA cancer cohorts. CACSurv achieves the highest C-index on all six cohorts and an average C-index of 0.722, outperforming the strongest published survival model by 6.5 percentage points and the strongest LLM time-regression baseline by 4.2 percentage points. Our code, models, and dataset will be available at https://github.com/xmed-lab/CACSurv.
