Hospital-Based Registry Analysis of Staging Efficacy and Proposed Substaging for HPV Positive Oropharynx Cancers

Abstract

The American Joint Committee on Cancer (AJCC) introduced the 8th edition staging for cancer in 2017, which provided separate staging for HPV+ oropharyngeal cancer (OPC). This update simplified staging with the intent of restoring prognostic discrimination between stages with regard to disease-specific survival. However, it also altered implications for treatment as many patients formerly classified as group stage I-IVA in the 7th edition, treated with either single or multimodality therapy including chemoradiation, were now classified as stage I. We used the National Cancer Database (NCDB) to study the efficacy of AJCC staging in a modern cohort of HPV+ OPC in the years prior to the introduction of AJCC 8th edition, and explore substaging of the AJCC 8th edition in order to provide better guidance for treatment.We queried the NCDB for patients with HPV+ OPC as their first primary cancer diagnosis, with known radiotherapy, surgery, and chemotherapy treatment status and without mortality within 90 days of surgery. Clinical T, N, and M stages per AJCC 7th edition were converted to AJCC 8th edition staging. Kaplan-Meier estimates were generated for overall survival (OS) at 5 years and compared using log-rank testing. Cox-proportional hazards models were generated to test for significant predictors of overall survival which were then used to generate substages for AJCC 8th edition stage I patients.A total of 8249 patients diagnosed from 2010-2016 met search criteria. Median age was 57 years (22-90). By AJCC 7th edition, 6% (470) of patients were group stage I, 10% (839) stage II, 23% (1913) stage III, 58% (4755) stage IVA, and 3% (272) stage IVB/C. A Cox-proportional hazards model including AJCC 7th edition T and N stage as covariates was generated using 5933 patients with clinical T1-2N0-2b staging, showing significant association of cT2 and cN2b status with OS (P = < 0.001, HR of 1.7 [95% CI 1.4-2.1] and 1.4 [95% CI 1.2-1.8] respectively). These results were used for substaging of AJCC 8th edition group stage I: cT1N0-2a were designated stage IA, cT2N0-2a and cT1N2b as stage IB, and cT2N2b as stage IIA. Using this substaging, log-rank testing between substages showed significant differences between the respective curves for OS at 5 years (P < 0.01 for all comparisons).Substaging of AJCC 8th edition group staging for HPV+ OPC provided significant hazard stratification for patients currently classified as group stage I, with subclassification by cT2 and cN2b (7th ed stage) showing significant association with OS at 5 years. Substaging by these factors not only maintains prognostic risk stratification, but also may provide better guidance in treatment decision making. For example, stage IA patients could be treated with either with surgery or radiation alone, stage IB with either single or multimodality therapy, and stage IIA and above would be treated with multimodality therapy. Further validation of this staging is warranted.