Human papillomavirus (HPV)-associated oropharyngeal cancer (OPC) is an epidemic: incidence has risen 70% in the last decade alone and 70% of OPCs are now attributable to HPV. In 2017, an updated staging system for HPV-positive OPC was issued by the American Joint Committee on Cancer (AJCC) that stratifies neck disease according to node number, location and size – not extracapsular extension (ECE). Conversely, HPV-negative mucosal squamous cell carcinoma neck disease staging is stratified according to the presence of absence of ECE given its adverse impact on survival.
To evaluate the impact of: 1) ECE on mortality in patients with HPV+ OPC with nodal metastases and 2) treatment type (surgery alone versus surgery with postoperative radiation versus surgery with postoperative chemoradiation) in patients with ECE.
The study design was a retrospective cohort analysis of HPV+ OPC patients in the National Cancer Database undergoing primary surgery from 2010-2014. Patients with second primary cancers, distant metastatic disease, and those receiving palliative treatment were excluded. A total of 3,565 patients with HPV+ OPC met all inclusion criteria, of which 3,158 were node-positive.
Median follow-up was 16.4 (IQR 8-26.3) months. Out of the entire cohort with positive nodes, ECE status and AJCC 8thedition stage was: no ECE: n=2,018 (64%), microscopic ECE: n=574 (18%), macroscopic ECE: n=143 (5%), ECE extent not specified (NS): n=423 (13%); stage I: n=2,568 (81%), stage II: n=524 (17%), stage III: n=66 (2%). The presence of ECE was a strong predictor of treatment paradigm in node-positive patients (p<0.0001). Surgery alone was performed in 21%, 9% ,11%, and 9% of patients with no ECE, micro ECE, macro ECE, and ECE NS respectively; PORT alone was delivered in 46%, 20%%, 17%, and 19%, respectively; and adjuvant CRT was used in 33%, 71%, 72%, and 71% of patients with macro ECE. Node-positive patients with ECE experienced significantly inferior survival in comparison to those without: 95% vs. 92% at 3 years (p=0.0035). Among all node-positive patients, after multivariable adjustment for stage, lymphovascular invasion, comorbidity and treatment (age was not significant and was excluded from the model), presence of any ECE was significantly associated with overall mortality (HR 1.607, 95% CI 1.076-2.399, p=0.0204). In a subset analysis of node-positive patients without ECE, there was no mortality difference between PORT (reference) or adjuvant CRT (HR 0.891, 95% CI 0.472-1.680, p=0.7215). However, among node-positive patients with ECE, the delivery of adjuvant PORT alone was associated with a significantly increased risk of death (HR 2.070, 95% CI 1.039-4.124, p=0.0387). This result was consistent when restricting the analysis to patients with microscopic ECE (HR 3.060, 95% CI 1.101-8.506, p=0.0320).
ECE remained a significant adverse prognostic factor after adjusting for AJCC 8 stage and treatment paradigm. In contrast to some single-institution studies, patients with both microscopic and macroscopic ECE experienced a survival benefit with adjuvant chemoradiotherapy over radiotherapy alone. Inability to adjust for adverse risk factors including perineural invasion and tobacco use status in this database analysis highlight the importance of ongoing and completed de-escalation trials.