KMGen is introduced, the first end-to-end framework that fully automates KM curve extraction at accuracy competitive with human-guided tools, and generates synthetic per-patient AE trajectories from public trial registry records.
Abstract
Individual patient data (IPD) from clinical trials is the substrate for survival modeling, meta-analysis, and safety research, yet IPD is rarely released. Prior work has addressed only half of this gap: reconstructing Kaplan-Meier (KM) curves from published plots -- typically requiring manual digitization or human-in-the-loop correction -- while offering no mechanism for generating the adverse-event (AE) streams that constitute the other half of a patient record. We introduce KMGen, the first end-to-end framework that (i) fully automates KM curve extraction at accuracy competitive with human-guided tools, and (ii) generates synthetic per-patient AE trajectories from public trial registry records. The extraction stage is a fully automated agentic pipeline -- an agent generates code to extract each step in the KM curve -- achieving a mean Integrated Absolute Error (IAE) of 0.0151 on a 32-plot benchmark spanning clean, edge-case, and adversarial conditions. The IPD generation stage decouples patient archetype extraction from statistical sampling: an LLM distills the trial record into arm-specific statistics, adverse events, patient demographics, and risk multipliers. A mechanistic sampler generates patient events via clinical archetypes, bootstrap rank-correlation coupling to the empirical KM curve (preserving the marginal survival distribution exactly), and cycle-based AE scheduling with an induction/maintenance split. Across three held-out oncology trials spanning an order of magnitude in cohort size and 30 independent regenerations per trial, KMGen achieves mean integrated KM absolute difference $\Delta_{\text{KM}}\,{\leq}\,0.051$, sex/ECOG JSD ${\leq}\,0.013$ on 5 of 6 demographic slots, and recovers ${\geq}\,71\%$ of the top-15 AEs by exact MedDRA term under a single fixed parameter set. The pipeline is released as open source at https://github.com/chufangao/kmgen.
CoMedBench is introduced, a reproducible benchmark that evaluates a family of generators under a common clinical-validity framework and one shared training and evaluation engine, spanning static tabular and temporal downstream tasks on established critical-care datasets.
This tutorial presents OneEHR, an open-source toolkit that defines a unified experiment contract for modern EHR modeling and enables head-to-head comparison among conventional, neural, LLM-based, and agentic methods through a single configuration-driven interface.
Yinghao Zhu, Zixiang Wang, Lei Gu et al.· Proceedings of the 32nd ACM...· 0 citations
Objective: Health systems and biopharma face a gap between holding patient data and acting on it: records are fragmented, manually mapped, and structured for storage rather than decisions, so every application re-derives patient state. We present PRomop, an open-source longitudinal record that closes this gap. Materials and Methods: PRomop builds on the OMOP Common Data Model (CDM 5.4) with oncology extensions and adds PatientRecord, a flattened projection collapsing each patient's longitudinal history into a single decision-ready 304-column row. State derivations - lines of therapy, disease status, normalized biomarkers - are computed once at projection time, so analytics, trial matching, and standard-of-care evaluation read one substrate. Results: PRomop is deployed by two oncology organizations - the independently governed HealthTree Foundation (~14,000 patients) and CancerBot (~3,500), a HealthKey-owned deployment - matching against 19,500 recruiting trials across five cancer types. A 20-criterion eligibility search requiring 27-39 joins over raw OMOP reduces to zero against the projection. On a synthetic 1000-patient breast-cancer cohort, eligibility screening averaged 0.30 ms via PatientRecord versus 11.0 ms from raw OMOP, a ~36.8x speedup. Discussion: The projection's significance is as a foundation for other applications: it lowers each one's marginal cost by computing error-prone clinical derivation once and removing it from every consumer. Line-of-therapy inference showed decision-readiness demands embedded clinical reasoning, and that the projection is a living artifact requiring maintenance. Conclusion: A flattened, decision-ready projection over a standards-based longitudinal record is a deployed pattern for turning fragmented data into actionable infrastructure, while remaining OMOP-conformant. Benchmarks measured a ~36.8x eligibility-screening speedup.
A. Blum, Louis Ferger-Andrews, Steven Labkoff· 0 citations
As precision oncology drives development toward narrower biomarker-defined populations, patient-level outcome prediction models can improve trial planning and decision-making. Methods that integrate both real world data (RWD) and recent trial evidence can produce better-calibrated predictions and enable clinical applications such as synthetic control arms, comparative effectiveness, and trial design optimization.
We developed a calibrated pan-cancer foundation model integrating patient-level RWD with summary-level clinical trial outcomes. The model uses a transformer-based architecture to capture joint distributions across thousands of clinical and genomic features. It was trained on over 300,000 tumor biopsy records with linked clinical data, primarily from RWD sources. The model generates synthetic patient-level cohorts conditional on user-specified I/E criteria and predicts outcomes under specified treatments. An information-geometric calibration procedure aligns these predictions with published trial baseline characteristics and outcome landmarks.
We validated the model across three Phase III settings: (1) To evaluate subgroup-level prediction from population-level calibration, we generated a synthetic cohort matching mNSCLC trial POSEIDON baseline characteristics and calibrated to its published control-arm OS. We then predicted OS across PD-L1 strata, histology, and KEAP1/STK11/KRAS mutation status and recapitulated published results: 90.9% (40/44) of median and 2 to 5-year OS estimates fell within 95% CIs, with median absolute deviation 2.7%. (2) To assess out-of-sample prediction in a genetic subgroup, we simulated a BRAF V600E mCRC cohort matching BREAKWATER baseline characteristics. The model was calibrated on prior unselected mCRC trials (XELOX, TRIBE) and applied without calibration to BREAKWATER or any BRAF-selected trial. Despite only five training-data patients meeting BREAKWATER I/E criteria, model-predicted OS matched observed values at 6, 12, and 18 months. (3) To demonstrate indirect head-to-head comparison without a randomized trial, we compared nab-paclitaxel plus gemcitabine (NG) and FOLFIRINOX in mPDAC. These regimens were evaluated in MPACT and PRODIGE4 respectively, with differing populations. We simulated the MPACT arm, then used entropy balancing to conform baseline characteristics to PRODIGE4, estimating NG outcomes in a healthier PRODIGE4-like population. This decomposed the published 80-day median OS gap: ∼14% was attributable to baseline demographics, with a residual 73-day FOLFIRINOX advantage.
We present a framework that calibrates patient-level OS predictions to published clinical trial evidence. Pan-cancer pretraining enables transfer learning across data sources, improving prediction in narrow populations with sparse data. The model can be further fine-tuned on data to support tailored predictions across biomarkers, indications, and treatments. Together, these capabilities provide a data-efficient approach to generating patient-level evidence for clinical applications.
Daniele Bertolini, Franklin Fuller, Jason Christopher, Jonathan Walsh, Samantha I . Liang, Aaron Smith. Patient-level prediction of trial outcomes with a calibrated pan-cancer foundation model [abstract]. In: Proceedings of AACR Drug Discovery and Development (AACR D3) Conference; 2026 Jul 21-24; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(14_Suppl):Abstract nr A030.
D. Bertolini, F. Fuller, J. Christopher et al.· Clinical Cancer Research· 0 citations
Background: Disease severity is a multidimensional construct difficult to capture with rule-based approaches in Electronic Healthcare Records (EHR). Agentic large language model (LLM) systems could synthesise clinical evidence and reason over EHRs, but remain unevaluated for this task. Methods: MOSAIC is a two-phase agentic LLM framework for severity phenotyping, using type 2 diabetes (T2D) as a proof-of-concept. MOSAIC was evaluated on a synthetic cohort (SyntheticMass; open-weight N = 4,886; closed-weight N = 200) against three algorithmic ground truths (DCSI, DiSSCo, Cooper) and against all-cause mortality and incident complications. Open-weight (locally deployable) and proprietary pipelines were also compared. Results: The generated framework spanned domains absent from the comparators, including biomarker-based glycaemic staging, beta-cell function, and social determinants of health. Open-weight MOSAIC matched the proprietary pipeline (closed- vs open-weight weighted kappa = 0.773) and reached moderate agreement with Cooper (kappa = 0.597) and DCSI (kappa = 0.534) and fair agreement with DiSSCo (kappa = 0.320). Agent-based (Type 1) tiers showed significant separation of all-cause mortality (log-rank p<0.001; crude hazard ratios 1.6-2.4 for non-Baseline tiers), with non-monotonic separation at the upper tiers, and an inverse gradient for incident complications (log-rank p<0.001) consistent with depletion of susceptibles. Agentic classification also diverged from deterministic execution of the same rubric (MOSAIC Frozen; kappa = 0.428), indicating reasoning beyond fixed rules. Conclusion: MOSAIC shows agentic LLM systems can generate and apply clinically meaningful severity phenotypes from structured EHR data in T2D. Extending it to other diseases with similarly multidimensional severity warrants further research.
Manuel Suero, Arnault-Quentin Vermillet, Nicole Sonne Heckmann et al.· 0 citations
Getting clinical data from different sources to “talk” to each other within the OMOP Common Data Model (CDM) is arguably the most tedious part of multi-center research. While this integration is essential, the transformation process is frequently a manual grind, requiring a rare overlap of deep clinical knowledge and technical expertise. In this paper, we present a framework designed to alleviate some of the burden on the researcher by automating data harmonization through two distinct steps: structural schema mapping and terminological standardization. For the structural piece, we moved away from “black box” logic in favor of a stateful workflow managed by large language models (LLMs) and directed acyclic graphs. By profiling EHR data at the source, our system generates context-aware dictionaries that offer ranked mapping suggestions alongside confidence scores. While our benchmarking showed a 97.5% agreement rate at the schema level and an 84% agreement rate at the value level when compared with human experts, the system appears most effective when treated as a “co-pilot” rather than a total replacement for human oversight. To handle value-level standardization, we implemented a hybrid search strategy that pairs the semantic depth of SapBERT embeddings with the literal precision of fuzzy string matching. By using FAISS for rapid similarity retrieval, the engine attempts to resolve messy or “noisy” clinical descriptions to standard OMOP concepts. This approach seems particularly promising for handling the non-standardized labels that often plague smaller, local datasets. Ultimately, our results suggest that this guided approach can shift the timeline for OHDSI-compliant warehousing from weeks of manual curation to a more manageable and scalable pipeline, potentially lowering the barrier to entry for smaller research teams.
Nishu Nehra, Rohit Swami, Dharani Dadi et al.· bioRxiv· 0 citations