Introduction: Multi-modality treatment intensification is indicated for patients undergoing definitive surgical resection of head and neck cancer with high-risk pathologic features. Adjuvant radiation and chemotherapy following surgery or tri-modality therapy is recommended based on findings of extranodal extension (ENE) and positive margins for head and neck squamous cell carcinoma. However, tri-modality therapy for human papilloma virus (HPV) associated oropharyngeal squamous cell carcinoma (OPSCC) has not demonstrated significant improvement in survival compared to dual modality therapy such as surgery with adjuvant radiation or definitive chemoradiation. Avoidance of tri-modality therapy may reduce long-term functional side effects and toxicities in this cohort. National cancer registry data has demonstrated the number of pathologic lymph nodes is associated with the high-risk feature of ENE and may be predictive of tri-modality care. The objective of this study is to assess the use of regional lymph node count to minimize rates of tri-modality therapy for OPSCC at the University of Maryland Medical Center.
Methods: Patients at the University of Maryland Medical Center who were diagnosed with OPSCC from 2018 through 2021 were extracted using the appropriate diagnostic ICD-10 code. Exclusion criteria included patients presenting with OPSCC recurrence or any patient that did not receive definitive treatment during this time interval. Demographic, radiographic, pathologic and survival variables were collected. Statistical analysis conducted included descriptive statistics, univariate and multivariate logistic regression as well as subset analysis for patients who were diagnosed with HPV-positive OPSCC and specifically those patients who underwent transoral robotic surgery.
Results: 252 patients met criteria and became the analytic cohort. The majority of the cohort was male (87%), former smokers (54%), with HPV-positive disease (87%). Subsite date demonstrated that most OPSCC cancers involved the palatine tonsil (53%) and/or the base of tongue (42.5%). Overall our institutional tri-modality rate was 14%. Our single modality rate was 19.4% and included those who underwent definitive surgery or radiation alone. The dual modality rate was 66% which includes those who underwent definitive chemoradiation or surgery with adjuvant radiation. Of the surgery cohort, pathologic extranodal extension was present in 28.9% of patients and in the transoral robotic surgery subset specifically the ENE rate was 25.5%. Multivariate logistic regression has demonstrated that the presence of pretreatment variables such as clinical ENE and number of lymph nodes are more likely to undergo tri-modality treatment (0.008, 0.01).
Conclusion: The institutional practice at the University of Maryland Medical Center has traditionally been to use pretreatment variables to stratify patients to definitive surgery versus chemoradiation in an attempt to avoid tri-modality care. Our data demonstrate that patients with a multiple lymph nodes or radiographic clinical ENE are more likely to have pathologic ENE that would indicate adjuvant chemoradiation and thus tri-modality care. Overall this demonstrates that the use of radiographic variables such as number of lymph nodes and clinical ENE can be used to predict whether a patient should undergo definitive chemoradiation even if the patient is a surgical candidate.