Validation of the Toronto recurrence inference using machine-learning for post-transplant hepatocellular carcinoma model
Bibliographic record
Abstract
BACKGROUND: Organ shortages require prioritizing hepatocellular carcinoma (HCC) patients with the highest survival benefit for allografts. While traditional models like AFP, MORAL, and HALT-HCC are commonly used for recurrence risk prediction, the TRIUMPH model, which uses machine learning, has shown superior performance. This study aims to externally validate the model. METHODS: The cohort included 2844 HCC patients who underwent liver transplantation at six international centers from 2000-2022. The TRIUMPH model utilized a regularized Cox proportional hazards approach with a penalty term for coefficient adjustment. Discrimination was assessed using the c-index, and clinical utility was evaluated via decision curve analysis. RESULTS: The most common liver diseases are hepatitis C (49%) and hepatitis B (27%). At listing, 84% meets the Milan criteria, and 91% are within criteria at transplant. Median model for end-stage liver disease score is 10 (IQR:8-14), alpha-fetoprotein level 8 ng/mL (IQR:4-25), and tumor size 2 cm (IQR:1.1-3.0). Living donor grafts are used in 24% of cases. Recurrence rate is 9.1% with a median time to recurrence of 17.5 months. Recurrence-free survival rates at 1/3/5 years are 95.7%/89.5%/87.7%, respectively. The TRIUMPH model achieves the highest c-index (0.71), outperforming MORAL (0.61, p = 0.049) and AFP (0.61, p = 0.04), though not significantly better than HALT-HCC (0.67, p = 0.28). TRIUMPH shows superior clinical utility up to a threshold of 0.6. CONCLUSIONS: The TRIUMPH model demonstrates good accuracy and clinical utility in predicting post-transplant HCC recurrence. Its integration into organ allocation could improve transplantation outcomes.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".