Development and Validation of an Extra Spindle Pole Bodies–like 1–Based Diagnostic and Prognostic Model for Hepatitis B Virus–Related Hepatocellular Carcinoma: Retrospective Cohort Study
Bibliographic record
Abstract
Background: Early diagnosis of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B virus (HBV) is challenging. Models that combine novel biomarkers with clinical features may improve both early diagnosis and risk stratification, but few have been systematically validated. Objective: This study aimed to develop and validate an extra spindle pole bodies-like 1 (ESPL1)-based model for diagnostic discrimination of HBV-related HCC and longitudinal risk stratification in patients with chronic HBV infection. Methods: Patients with chronic HBV were consecutively recruited from the First Affiliated Hospital of Guangxi Medical University (a single-center, tertiary hospital) between January 2012 and November 2023. Patients were divided into a training set and an independent hold-out testing set. A LASSO logistic regression model was constructed to identify independent predictors and then used to develop a risk score discriminating patients with HBV-related HCC from those with chronic hepatitis B or cirrhosis. Model performance was evaluated using discrimination (C-index), calibration, and decision curve analysis. Internal validation was performed with bootstrap resampling, and independent hold-out validation was conducted with the independent hold-out testing set. Longitudinal follow-up of patients with chronic hepatitis B or cirrhosis was subsequently used to examine cumulative incidence across risk groups, thereby assessing the model's ability to stratify future HBV-related HCC risk. A web-based calculator was developed to facilitate clinical application. Results: The study involved a cohort of 621 patients diagnosed with chronic HBV infection, divided into a training set of 373 (60.1%) patients and an independent hold-out testing set of 248 (39.9%) patients. Age (odds ratio [OR] 1.08, 95% CI 1.05-1.12), ESPL1 expression (OR 1.01, 95% CI 1.00-1.01), and log (alpha-fetoprotein) levels (OR 2.55, 95% CI 1.95-3.33) were identified as independent predictors of HBV-related HCC. The model demonstrated excellent diagnostic discrimination, with a C-index of 0.922 in the training set and 0.958 in the independent hold-out testing set, coupled with strong calibration. Decision curve analysis revealed that the model consistently provided a higher net benefit across clinically relevant threshold probabilities. Subgroup analyses further validated the model's high discriminative power, with C-index values ranging from 0.86 to 0.98, and no significant interactions were detected (all interaction P values > .10). Furthermore, the model demonstrated superior discriminatory power relative to 5 established HBV-related HCC risk scores, including REACH-B, GAG-HCC, CU-HCC, PAGE-B, mPAGE-B, and alpha-fetoprotein alone, with all pairwise comparisons yielding statistical significance (P<.001). For prognostic stratification, patients categorized as low risk, medium risk, and high risk had distinct 5-year cumulative HCC incidences of 5.1%, 21.1%, and 31.3%, respectively (P<.001). Conclusions: The ESPL1-based model may serve as both a diagnostic tool for differentiating patients with HCC from those with non-HCC and as a preliminary approach for risk stratification during follow-up. This dual role has the potential to support earlier detection and personalized monitoring. The web-based calculator improves accessibility and may facilitate future clinical integration.
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 | 0.015 | 0.015 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 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".