Abstract A053: A pre-operative Artificial Intelligence-based model to estimate the risk of cancer specific mortality in patients with non-metastatic kidney cancer
Bibliographic record
Abstract
Abstract Surgical resection is the standard treatment for patients with non-metastatic clear cell renal cell carcinoma (ccRCC). Recurrence occurs in around 30% of patients, necessitating personalized risk-based strategies. Current prognostic models, e.g. TNM staging, lack precision, and cannot be applied preoperatively. We developed and validated a preoperative Machine Learning (ML)-based but interpretable TRIPOD type III model for RCC mortality using Real-World Data (RWD). We made it clinically available as a Web-based app for further external validation (TRIPOD type IV). The model was developed using 206 RWD variables from 3081 ccRCC patients (1987-2020) at our institution, rigorously externally validated on an independent cohort of 720 patients from another Italian hospital and benchmarked against a fine-tuned version of the GRANT model. A hybrid approach combining black-box (Random Survival Forest) and white-box models was employed: the black-box model identified hidden patterns in RWD, which were used to develop interpretable white-box models, ensuring transparency and clinical relevance. Model performance was evaluated using C-index and Brier scores. Patients were stratified into three risk groups based on preoperative mortality probabilities. The pre-operative model (survival tree) outperformed the GRANT model on the external cohort: C-index 0×88, cumulative AUC 0×89, Brier’s score 0×02. It requires only 8 preoperative features: tumor size, lymph node involvement, platelet count, hemoglobin, glomerular filtration rate, age, BMI, and performance status. A web-based version of the model ensures its broader applicability. ML explainability highlighted tumor size, performance status, lymph node involvement as the most important covariates. Lower hemoglobin and higher platelet counts were confirmed as poor prognostic indicators. Compared to the GRANT, our model had remarkable performances within the first year after surgery. Citation Format: Alberto Traverso, Alessandro Larcher, Patrick Scuri, Antonio Esposito, Carlo Tacchetti, Andrea Salonia. A pre-operative Artificial Intelligence-based model to estimate the risk of cancer specific mortality in patients with non-metastatic kidney cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Artificial Intelligence and Machine Learning; 2025 Jul 10-12; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(13_Suppl):Abstract nr A053.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".