AHNS Abstract: B198

← Back to List


Program Number: B198
Session Name: Poster Session

Oral Cavity Squamous Cell Carcinoma and Tumor Depth of Invasion: Determining the Association with Lymph Node Metastasis and Oral Cavity Subsites using the National Cancer Database

Audrey M Abend, AB; Chase Hintelmann, BA; Katherine Dinh, BA; Craig Bollig, MD; Rutgers Robert Wood Johnson Medical School

Background: Oral cavity squamous cell carcinoma (OCSCC) constitutes over 90% of oral cancers and is associated with high morbidity and mortality, particularly when lymph node metastasis (LNM) is present. Depth of invasion (DOI) is a known predictor of LNM and thus survival in tongue cancers; however, this relationship is less well defined in other oral cavity subsites. This association has been incompletely analyzed in a national dataset. 

Objective(s): Our primary objective was to define the association between DOI with LNM and overall survival (OS) in patients with OCSCC using the National Cancer Database (NCDB). Our secondary objective was to explore oral cavity subsite trends with the association between DOI with LNM and OS.

Methods: We performed a retrospective analysis all patients of patients diagnosed with a T1 or T2 OCSCC who underwent resection between 2010-2017 with a reported DOI using the 2021 Patient User File of the NCDB. Patients with grossly positive margins, distant metastatic disease, and missing data were excluded. Logistic regression was used to assess the association of DOI with LNM as well as Cox proportional hazards models to evaluate the association between DOI with OS. Multivariable models were created to adjust for known confounders. Adjusted odds ratios (aOR) or hazard ratios (aHR) with associated 95% confidence intervals (CI) and hazard ratios (HRs) were generated. 

Results: 13,234 subjects fit inclusion criteria. On initial analysis, 6,676 patients (50.4%) had a recorded DOI <1.5mm, suggesting a miscoding in the NCBD dataset. After adjusting these values by a factor of 10, the DOI-LNM relationship aligned with previous literature. After adjustment, each 1mm increase in DOI was associated with a 15% increase in the odds of LNM (OR 1.15, 95% CI: 1.13-1.16) for all oral cavity cancers. This relationship was strongest in tongue cancers (OR 1.18, 95% CI: 1.16-1.20), followed by buccal mucosa (OR 1.12, 95% CI: 1.07-1.17), retromolar trigone (OR 1.11, 95% CI: 1.05-1.18) and floor of mouth (OR 1.10, 95% CI: 1.07-1.14). No significant association between DOI and LNM was seen in cancers of the gingiva, hard palate, or lip. Adjusting for age, comorbidity score, positive nodal disease, and positive surgical margins, DOI was associated with overall survival, with each 1mm increase in DOI resulting in a 7% higher odds of mortality (HR 1.07, 95% CI: 1.06-1.07). The increased odds of mortality was observed in cancers of the tongue (aHR 1.06, 95% CI: 1.05-1.08), buccal mucosa (HR 1.06, 95% CI: 1.03-1.09) and floor of mouth (aHR 1.03, 95% CI: 1.01-1.05). There was no significant association with DOI and OS in cancers of the gingiva, retromolar trigone, hard palate, or lip. 

Conclusion: This national analysis demonstrates the association of DOI with LNM and OS in OCSSC that is most prominent in tongue cancers, but is also present in buccal and floor of mouth malignancies. This relationship was not identified in cancers of the gingiva, hard palate, and lip. This analysis also demonstrates the potential for major coding errors with DOI in the NCDB.

 

 

← Back to List