Abstract
OBJECTIVES: Oral squamous cell carcinoma (OSCC) exhibits heterogeneous outcomes based on nodal status, complicating personalized prognosis. This study aimed to identify nodal-specific prognostic factors in OSCC by integrating metabolic metrics from (18)F-fluorodeoxyglucose Positron Emission Tomography/Computed Tomography, body composition (L3 skeletal muscle index, L3-SMI), and key pathological features, to refine risk stratification. METHODS: We conducted a retrospective cohort study of 147 OSCC patients (74 pN0, 73 pN+) who underwent curative resection and neck dissection. Associations between metabolic metrics, L3-SMI, pathological factors, and overall (OS) or progression-free survival (PFS) were evaluated using Cox regression. Internal bootstrap validation (1000 repetitions) was performed to assess the stability and potential bias of the prognostic factors. RESULTS: In pN0 patients, tumor maximum standardized uptake value (T-SUVmax) >13.27 emerged as an independent predictor of poor OS (HR = 10.24, P = 0.003) and PFS (HR = 8.23, P = 0.002), which was validated by bootstrapping. Among pN+ patients, ≥3 positive lymph nodes significantly predicted worse outcomes (OS HR = 4.15, P < 0.001; PFS HR = 1.97, P = 0.049), while higher L3-SMI served as a protective factor for survival (OS HR = 0.13, P = 0.047; PFS HR = 0.18, P = 0.024); both findings were supported by internal validation. CONCLUSIONS: We propose a nodal status-stratified approach for risk assessment in OSCC. For pN0 necks, high risk is characterized by elevated T-SUVmax, whereas in pN+ necks, high nodal burden and sarcopenia define increased risk.