Schadendorf, Dirk

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.

9037 Background: Therapies for aBCC, which includes metastatic (m) and locally advanced (la) BCC, are limited. Abnormal Hedgehog pathway signaling is a key driver in BCC pathogenesis. Primary analysis of the pivotal ERIVANCE BCC trial of vismodegib, an oral hedgehog pathway inhibitor (HPI), demonstrated an objective response rate (ORR) of 30% and 43%, in mBCC and laBCC patients, respectively, with a median duration of response (DOR) of 7.6 months. We present safety and investigator (INV) assessed efficacy results 18 months (29 May 2012) after primary analysis (26 Nov 2010). Methods: Multicenter, international, nonrandomized study in patients (N=104) with radiographically measurable mBCC or laBCC (surgery inappropriate due to multiple recurrence, or substantial morbidity or deformity anticipated) receiving 150 mg oral vismodegib daily until disease progression or intolerable toxicity. Key secondary endpoints included INV-assessed ORR, progression-free survival (PFS), DOR, overall survival (OS), and safety. Results: At data cutoff, 21 patients continued to undergo protocol-specified assessments and 56 patients were in survival follow-up. The median dose intensity was comparable with primary analysis. ORR was 48.5%, mBCC; 60.3%, laBCC, comparable with primary analysis. However, median DOR improved (mBCC=14.7; laBCC=20.3 months). The median OS for mBCC was 30.9 months but not estimable in laBCC. Adverse events remained consistent, with muscle spasm, alopecia, dysgeusia, weight decrease, and fatigue most frequently reported. Eleven more deaths were reported in the update period after primary analysis; these occurred in survival follow-up and were not drug-related. Conclusions: Vismodegib is the first FDA-approved HPI; thus, long-term efficacy and safety data are particularly relevant. 18-month update data confirmed prolonged responses and consistent safety in vismodegib-treated aBCC patients. Clinical trial information: NCT00833417.