Answers are grounded in this paper and cite their sources.
Tumor-node-metastasis (TNM) staging standardizes the description of disease extent in oral squamous cell carcinoma. The eighth edition of the American Joint Committee on Cancer (AJCC) staging system incorporated depth of invasion into T classification and extranodal extension into N classification. This narrative review examines the evidence underlying these revisions, their validation, and priorities for further refinement. Multicenter cohorts, registry analyses, and systematic reviews generally support improved prognostic stratification, although gains in discrimination vary and are often modest. Measurement variability and heterogeneous outcomes within categories remain important limitations. We propose an evaluation roadmap that links each candidate to a defined classification weakness, reproducible measurement, independent incremental value, and feasibility. For patients treated with upfront surgery, positive-node count combined with extranodal extension status warrants first priority for evaluation, followed by standardized bone-invasion descriptors. Quantitative extranodal extension requires further standardization. Additional pathological features and prediction models may complement individualized risk assessment. Current evidence neither establishes a ceiling for anatomical staging nor demonstrates that further anatomical subdivision alone will resolve residual risk variation. Retaining TNM as the core staging system and supplementing it with validated models is a practical approach. Hybrid stage groupings require separate evaluation. Clinical, pathological, and post-treatment settings should remain distinct. These priorities reflect evidence synthesis and author interpretation, not guideline-level recommendations. Greater prognostic accuracy alone does not establish treatment benefit.
No takes yet. Share an insight, caveat, or question.
Shan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: