Wicki, Andreas

Abstract Cancer outcomes vary widely between individual patients, each accumulating an irregular record of treatments, diagnoses, measurements, and complications. Current prognostic models reduce this complexity into a single snapshot, focus on narrow clinical settings, and rarely generalize across hospitals. Here we introduce Chronicle , an explainable transformer that learns from entire patient trajectories to predict diverse clinical outcomes throughout the disease course while capturing both short- and long-term temporal dependencies. Trained on 53.7 million longitudinal data points from 51,711 patients spanning 67 cancer types, Chronicle operates natively on irregular data without imputation and jointly predicts eight endpoints within a flexible framework adaptable to additional outcomes. Chronicle outperformed cross-sectional models for overall survival prediction (C-index 0.84 vs 0.76-0.79), stratified patients more accurately than established prognostic systems, including TNM stage, and predicted seven adverse event and transfusion endpoints (AUC 0.80-0.92). Applied without retraining to 69,341 patients in Germany, Switzerland, and the United States, Chronicle generalized across healthcare systems and improved further with local fine-tuning. Integrated explainability traced each risk update to patient-specific clinical factors, revealing distinct temporal persistence of prognostic information, with relevance half-lives ranging from weeks for therapies to nearly one year for baseline characteristics. These findings demonstrate that learning from hospital-wide patient trajectories enables interpretable and continuously updated predictions, providing a scalable framework to support individualized treatment decisions.

Background: Bimiralisib is a pan-PI3K/mTOR inhibitor demonstrating antitumor efficacy in preclinical models. The objectives of this study were to identify a maximum tolerated dose (MTD), pharmacokinetics (PK), a dosing schedule, and adverse events (AEs) in patients with advanced solid tumors. Patients and Methods: Patients received oral bimiralisib to determine the MTD of one continuous (once daily) and two intermittent schedules (A: Days 1, 2 weekly; B: Days 1, 4 weekly) until progression or unacceptable AEs occurred. Results: The MTD for the continuous schedule was 80 mg, with grade three fatigue as the dose-limiting toxicity (DLT). No MTD was reached with intermittent schedules, with only one DLT in schedule B. PK analysis suggested that 140 mg (schedule A) was within the biologically active dose range and was selected for further exploration. The most frequent treatment-emergent AEs were hyperglycemia (76.2%) in the continuous schedule, and nausea (56–62.5%) in schedules A and B. The most frequent treatment-emergent > grade three AE for all schedules combined was hyperglycemia (28.6%, continuous schedule; 12.0%, schedule A; 12.5%, schedule B). There was one partial response in a head and neck squamous cancer patient with a NOTCH1T1997M mutation. Conclusions: Bimiralisib demonstrated a manageable AE profile consistent with this compound class. Intermittent schedules had fewer > grade three AEs, while also maintaining favorable PK profiles. Intermittent schedule A is proposed for further development in biomarker-selected patient populations.