Belka, Claus

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.

AbstractBackgroundAs a step towards clinical cardiac xenotransplantation, our experimental heterotopic intrathoracic xenotransplantation model offers a beating and ejecting donor heart while retaining the recipient′s native organ as a backup in case of graft failure. Clinically applicable immunosuppressive regimens (IS) were investigated first, then treatments known to be effective in hypersensitized patients or those with recalcitrant rejection reactions.MethodsConsecutive experiments were carried out between 2009 and 2013. Twenty‐one genetically modified pigs (GGTA1‐knockout/hCD46/± thrombomodulin, in one caseHLA‐E instead) were used as donors. In all experiments, two cycles of immunoabsorption reduced preformed antibodies. Recipient baboons were divided into two groups according toISregimen:In group one (n = 10), pre‐treatment started either one (anti‐CD20) or four weeks (anti‐CD20 plus the proteasome inhibitor bortezomib) prior to transplantation. The extended conventional (as for allotransplantation) immunosuppressive maintenance regimen included anti‐thymocyte globuline, tacrolimus, mycophenolate mofetil, methylprednisolone and weekly anti‐CD20.In group two (n = 11), myeloablative pre‐treatment as in multiple myeloma patients (long and short regimens) was added to extended conventionalIS; postoperative total thoracic and abdominal lymphoid irradiation (TLI; single dose of 600cGY) was used to further reduce antibody‐producing cells.ResultsIn the perioperative course, the surgical technique was safely applied: 19 baboons were weaned off extracorporeal circulation and 17 extubated. Nine animals were lost in the early postoperative course due to causes unrelated to surgical technique orISregimen.Excluding these early failures, median graft survival times of group 1 and 2 were 18.5 (12–50) days and 16 (7–35) days. Necropsy examination of group 1 donor organs revealed hypertrophy of the left ventricular wall in the six longer‐lasting grafts; myocardial histology confirmed pre‐clinical suspicion of humoral rejection, which was not inhibited by the extended conventional IS including intensified treatments, and signs of thrombotic microangiopathy.Grafts of group 2 presented with only mild‐to‐moderate features of humoral rejection and thrombotic microangiopathy, except in one case of delayed rejection on day 17. The other experiments in this group were terminated because of untreatable pulmonary oedema, recurring ventricular fibrillation, Aspergillus sepsis, as well as a combination of a large donor organ and late toxic side effects due toTLI.ConclusionsLonger‐term results were difficult to achieve in this model due to theISregimens used. However, we conclude that heterotopic intrathoracic heart transplantation may be an option for clinical xenotransplantation.

The heterotopic thoracic pig‐to‐baboon heart transplantation has been established by our group as a safe preclinical model. Since the recipient's own heart remains in place, it is possible to evaluate immunological reactions of various types and the symptoms of the thrombotic microangiopathy under working heart conditions. However, these experiments are complex and costly. In order to limit the number of transplants, we tested our multi‐transgenic porcine donor hearts in an ex‐vivo perfusion model first.Ex‐vivo model: Both beating ventricles were perfused with heparinized freshly drawn human whole blood using a centrifugal pump and a membrane oxygenator. During 3 h of observation, cardiac function parameters were obtained continuously; specimens from the perfusate, coronary sinus blood, and myocardial biopsies were assessed. Various genetically modified porcine donor hearts were tested; we are currently evaluating the impact of human thrombomodulin on the porcine microcirculation; the efficacy of complement regulation is another work package.Pig‐to‐baboon heterotopic cardiac xenotransplantation: In our latest group, seven baboons received five double (Gal‐KO/hCD46), and two triple (Gal‐KO/hCD46/hTM resp./HLA‐E) transgenic pig hearts. Our immunosuppressive regimen included preoperative anti‐CD20‐antibody, bortezomib, dexamethasone and cyclophosphamide; postoperatively, ATG, tacrolimus, MMF, anti‐CD20‐antibody, bortezomib and dexamethasone were administered. Total lymph node irradiation (6 Gy) was applied on postoperative day five.The triple transgenic hearts survived 35 and 37 days respectively. Both hearts maintained function throughout the experiment. Both recipients succumbed to fungal infections. Humoral rejection was seen only once. The next barrier appears to be the complex event of thrombotic microangiopathy that needs to be addressed with the additional expression of human thrombomodulin strategies.