Derivation and validation of the AJCC9V pathological stage classification for HPV-positive oropharyngeal carcinoma: a multicentre registry analysis

Ho, Allen S., Huang, Shao Hui, O’Sullivan, Brian, Luu, Michael, Evans, Mererid, Ferris, Robert L., Lewis, James S., Seethala, Raja R., Yom, Sue S., Mehanna, Hisham, Branstetter, Barton F., Saba, Nabil F., Patel, Snehal G., Lydiatt, William M. and Zumsteg, Zachary S. 2025. Derivation and validation of the AJCC9V pathological stage classification for HPV-positive oropharyngeal carcinoma: a multicentre registry analysis. The Lancet Oncology 10.1016/S1470-2045(25)00281-5 Item availability restricted.

PDF - Accepted Post-Print Version Restricted to Repository staff only until 8 January 2026 due to copyright restrictions. Available under License Creative Commons Attribution. Download (289kB)

Abstract

Background Modernisation of the American Joint Committee on Cancer and Union for International Cancer Control (AJCC/UICC) staging for HPV-positive oropharyngeal carcinoma has strengthened its fundamental role to convey prognosis and guide treatment decisions. Implementation has nonetheless revealed imbalances. An AJCC committee reappraised HPV-positive oropharyngeal carcinoma pathological staging to advance prognostic accuracy and improve clinical applicability. Methods For this analysis, US registry data from the National Cancer Database were divided into derivation and validation cohorts. Eligible cases encompassed surgically treated adult patients aged 18 years or older with HPV-positive oropharyngeal carcinoma. The primary objective was to derive and validate an optimised HPV-positive oropharyngeal carcinoma pathological staging classification based on determinants of overall survival. Multivariable Cox regression models were used to assess lymph node characteristics for overall survival. Non-linear associations between metastatic lymph node number and survival were modelled with restricted cubic splines. Adjusted hazard ratios (AHRs) and recursive partitioning analysis methods were applied to generate optimal classification groups. Performance was evaluated with Groome's criteria. Findings Overall, 14 447 patients across 984 facilities in the USA met the criteria for inclusion, and were divided into a derivation cohort (n=7768) and validation cohort (n=6679). Patients were treated between 2010 and 2019; 12 276 (85·0%) were male, 2171 (15·0%) were female, 13 594 (94·1%) were White, and 4552 patients (31·5%) had pathological extranodal extension (pENE). Median follow-up was 52·4 months (95% CI 51·5–53·3). Mortality risk increased with each additional metastatic lymph node (HR 1·20 [95% CI 1·11–1·29], p<0·0001), up to an optimal cutoff at 4·3 lymph nodes. Multivariable analysis confirmed the association of pENE with increased mortality risk (HR 1·47 [95% CI 1·30–1·65], p<0·0001), but no significant prognostic changes were shown by the extent of pENE (minor vs major). AHR approaches derived and validated optimised pN and pTNM stage categories (N1a: 1 positive lymph node and pENE-negative; N1b: 2–4 positive lymph nodes and pENE-negative; N2: >4 positive lymph nodes and pENE-negative or 1–4 positive lymph nodes and ENE-positive; N3: >4 positive lymph nodes and ENE-positive; Stage I: T0-2N0-1M0; Stage II: T0-2N2-3M0 or T3N0-2M0; Stage III: T3N3M0 or T4N0-3MO); Stage IV: M1. AJCC9V showed superior hazard consistency, outcome prediction, and balance, but not hazard discrimination, compared to AJCC8E. The AJCC Expert Panel on HPV-positive oropharyngeal carcinoma endorsed the proposal by Delphi consensus.

Item Type:	Article
Date Type:	Published Online
Status:	In Press
Schools:	Schools > Medicine
Publisher:	Elsevier
ISSN:	1470-2045
Date of First Compliant Deposit:	29 July 2025
Date of Acceptance:	6 May 2025
Last Modified:	29 Jul 2025 15:15
URI:	https://orca.cardiff.ac.uk/id/eprint/179952

Actions (repository staff only)

Edit Item