Skip to main content
IFHNOS 2026
Non-smoking, non-drinking oral cavity squamous cell carcinoma: a distinct clinicopathologic phenotype?
Verbal Presentation

Verbal Presentation

9:10 am

29 August 2026

Plaza P2

Concurrent Session: Oral Cavity Abstracts

Themes

Default

Talk Description

Institution: Adelaide university - South Australia , Australia

Aims:To characterise the clinicopathologic phenotype of NSND oral cavity squamous cell carcinoma (OCSCC) compared with tobacco/alcohol-exposed (smoking/drinking; SD) disease focusing on histopathological risk features as well as clinical outcomes. Methodology: Of 403 consecutive OCSCC cases screened at a tertiary head and neck oncology centre (2016–2024), 241 met eligibility criteria. Patients were stratified as NSND (n=39, 16.2%) or SD (n=202, 83.8%). Univariate regressions, Fisher's exact tests compared pathologic features, margin status, treatment, and survival. Dose-response analyses examined six tobacco/alcohol exposure strata. Results: The NSND cohort demonstrated striking demographic heterogeneity spanning three age groups: young patients (45 or under, 19.6%), middle-aged (46 to 79 years, 47.1%), and elderly (80 >years, 33.3%), each with divergent pathologic profiles. We also characterised five independent histopathological risk features in both cohorts, an approach not previously reported, SD patients demonstrated poorer differentiation (OR 1.98, P=0.033), 4.7-fold greater odds of extranodal extension (ENE; OR 4.66, P=0.046), borderline increases in perineural invasion (OR 2.25, P=0.055) and lymphovascular invasion (OR 2.40, P=0.052), and more metastatic nodes (P=0.031). Macroscopic residual disease was confined to SD patients (6.9% versus 0%). These exposure-stratified pathologic variables were incorporated into a preliminary machine learning framework exploring prediction of high-risk outcomes. Overall survival did not differ significantly. Conclusion: This study provides a simultaneous clinical and histopathological characterisation of NSND OCSCC, identifying a coherent lower-risk phenotype most critically demonstrated by a four-fold reduction in ENE, a finding not previously reported. These findings support the case for phenotype-stratified MDT planning in OCSCC and highlight the need for prospective multi-centre validation in this field.
Presenters
Authors
Authors

Dr Jonathan Daniel - , Dr. Kevin Fenix - , Dr. Andrew Foreman - , Prof Alkis Psaltis - , A Prof Jc Hodge -