Year: 2026
Type: article
Abstract: Routine laboratory panels drawn during cancer treatment constitute longitudinal physiological recordings of organ function, yet their temporal structure is discarded by single-timepoint prognostic tools. A transformer trained on 2,777,595 laboratory measurements from 3905 patients with multiple myeloma or ovarian cancer predicted which of 162 treatment-associated complications would be newly recorded within two years, spanning eight clinical categories, achieving 1.5- to 6.1-fold enrichment above prevalence. Across individual endpoints it exceeded non-sequential baselines, including engineered-trajectory models, by +0.024 to +0.070 in mean AUROC. Predictions generalised across both cancers, divergence concentrating in disease-specific complications, and biomarker masking recovered signatures consistent with established pathophysiology. External validation on MIMIC-IV and MMRF CoMMpass showed transferability, strongest for renal and metabolic endpoints, attenuated for infection endpoints. On CoMMpass it added prognostic value to ISS staging. Routine oncological laboratory data carry complication-specific physiological signal months before a complication is recorded, and require no testing beyond standard oncological care.
Source: npj Digital Medicine
Authors: Jannik Lübberstedt, Krischan Braitsch, Jacqueline Lammert, Christof Winter, Florian Gabriel, Tristan Lemke, Christopher Zirn, Markus Graf, Friedrich Puttkammer, Hartmut Häntze, Johannes Moll, Anirudh Narayanan, Andrei Zhukov, Fabian Drexel, Zeineb Ben Chaaben, Sebastian Ziegelmayer, Su Hwan Kim, Marion Högner, Jan Kirschke, Florian Bassermann, Marcus Makowski, Christian Wachinger, Lisa Adams, Keno Bressem
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $4,090
Award: G9001395076
OpenAlex ID: W7163984111
