Individualized prediction of distant metastases risk in oral cavity carcinoma: A validated predictive-score model.
Bibliographic record
Abstract
6037 Background: We aimed to develop and validate a risk-scoring system for distant metastases (DM) in oral cavity carcinoma (OCC). Methods: In this IRB-approved retrospective study, OCC patients treated at 4 tertiary cancer institutions with curative surgery +/- postoperative radiation/chemo-radiation (PORT/PO-CRT) were divided into discovery and validation cohorts (randomly selected in 3:2 ratio). Staging was reviewed based on TNM 8 th edition. Predictors of DM identified on multivariable analysis in discovery cohort were used to develop DM risk-score model to classify patients into risk groups using Contal and O’Quigley method for cut-off optimization. The utility of risk classification was subsequently evaluated in validation cohort. C-index was used to assess predictive ability of the continuous risk score. Results: Overall 2749 patients were analyzed (Table). Predictors (risk score coefficient) of DM in discovery cohort were: pT3-4 (0.4), pN+ (N1:0.8; N2:1.0; N3:1.5), histologic grade 3 (G3, 0.7) and lymphovascular invasion (LVI, 0.4). The DM risk groups were defined by cumulative sum of risk score coefficients: high risk (sum >2), intermediate risk (sum=1-2), and standard risk (sum<1). In the discovery cohort, 5-yr DM for high vs intermediate vs standard risk groups was 33% vs 19% vs 6%, p<0.001 (C-index=0.79). Similarly, in the validation cohort, 5-yr DM for high vs intermediate vs standard risk groups was 36% vs 23% vs 7%, p<0.001 (C-index=0.77). When applied to entire study population, this predictive model showed excellent discriminative ability in predicting DM only without locoregional failure (29% vs 18% vs 3%, p<0.001), late (>2 yr) DM (11% vs 5% vs 3%; p<0.001), DM in patients treated with surgery only (26% vs 11% vs 6%, p<0.001), PORT (37% vs 23% vs 7%, p<0.001), and PO-CRT (42% vs 29% vs 9%, p<0.001). Finally, 5-yr OS for high vs intermediate vs standard risk groups in the overall cohort was 24% vs 38% vs 66%, p<0.001. Conclusions: A predictive-score model for DM utilizing pT3-4, pN1/2/3, G3 and LVI demonstrated a validated utility in identifying patients at higher risk of DM who may be evaluated for individualized risk-adaptive treatment escalation and/or surveillance strategies. Study cohorts and predictors of distant metastases in discovery cohort. Discovery (n=1650)N (%) Validation (n=1099)N (%) p Median follow up 4.6 yr 4.5 yr 0.11 pT3-4 895 (54) 565 (51) 0.16 pN1/pN2/pN3 156(9)/ 258 (16)/ 280 (17) 99 (9)/ 159 (14)/ 196 (18) 0.78 G3 210 (13) 132 (12) 0.62 LVI 349 (22) 271 (25) 0.05 PORT/PO-CRT 873 (54)/ 220 (13) 567 (52)/ 150 (14) 0.48 5-yr DM (95% CI) 14% (12%-17%) 12% (11%-14%) 0.07 5-yr OS (95% CI) 55% (52%-59%) 53% (51%-56%) 0.38 Predictors of DM @ * pT3-4 (p=0.04)* pN+ (p<0.001)* G3 (p< 0.001)* LVI (p<0.01) - - @ variables included in multivariable analysis: age, gender, smoking history, subsite, pT, pN, grade, LVI, PNI, margin status, pN+ at level IV/VB, PORT and PO-CRT.
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 | 0.003 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 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".