Performance of 8 Smoking Metrics for Modeling Survival in Head and Neck Squamous Cell Carcinoma
Bibliographic record
Abstract
Importance: Cigarette smoking is a strong risk factor for mortality in patients diagnosed with head and neck squamous cell carcinoma (HNSCC). However, little evidence supports which smoking metric best models the association between smoking and survival in HNSCC. Objective: To determine which smoking metric best models a linear association between smoking exposure and overall survival (OS) in patients with HNSCC. Design, Setting, and Participants: A retrospective multicenter cohort study of 6 clinical epidemiological studies was performed. Five were part of the Human Papillomavirus, Oral and Oropharyngeal Cancer Genomic Research (VOYAGER) consortium. Participants included patients 18 years and older with pathologically confirmed HNSCC. Data were collected from January 2002 to December 2019, and data were analyzed between January 2022 to November 2024. Main Outcomes and Measures: The primary outcome was OS. The performance of 8 smoking metrics, including pack-years, duration, and log cig-years (calculated as log10[cigarettes smoked per day + 1] × number of years smoked) for modeling OS were compared. Metric performance was measured by the strength of association in Cox proportional hazard models, linearity based on P for linear trend, Akaike information criterion (AIC; lower value indicates better model fit), and visual assessment of spline curves. Secondary outcomes included modeling OS in clinicodemographic subgroups and HNSCC anatomic subsites. Exploratory outcomes included cancer-specific survival and noncancer survival. Results: In total, 8875 patients with HNSCC (2114 [24%] female; median [IQR] age, 61 [54-69] years) were included. Of 8 smoking metrics evaluated, smoking duration (adjusted hazard ratio [aHR], 1.11 [95% CI, 1.03-1.19]) and log cig-years (aHR, 1.11 [95% CI, 1.04-1.18]) had the highest aHRs; both had a statistically significant linear association with OS. Log cig-years had the lowest AIC linear value and the most visually linear spline curve when modeling OS. Duration and log cig-years outperformed pack-years for modeling OS regardless of age, smoking status, and cancer stage. Both performed well in lip and oral cavity, laryngeal (only duration was significant), and human papillomavirus-negative oropharyngeal subsites. In an exploratory analysis, duration had the highest aHR (1.15 [95% CI, 1.02-1.29]), and log cig-years had the lowest AIC linear value when modeling noncancer survival. Conclusions and Relevance: In this cohort study, smoking duration and log cig-years best modeled a linear relationship with OS for patients with HNSCC. Both metrics maintained robust performance within specific clinicodemographic subgroups and anatomic subsites. Although most HNSCC survival models control for smoking exposure using smoking status or pack-years, duration and log cig-years may be superior metrics to account for the effects of smoking on survival.
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 | 0.011 | 0.021 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 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".