Incidence and patterns of retropharyngeal lymph node involvement in oropharyngeal carcinoma.

Abstract

The aim was to evaluate in oropharyngeal carcinoma the: (1) incidence and predictors of retropharyngeal (RP) lymph node (LN) involvement, (2) pattern of ipsilateral/bilateral/contralateral-only RP LNs (3) location of RP LNs in relation to contouring guidelines. Single centre retrospective analysis of 402 patients with oropharyngeal carcinoma treated non-surgically between 2010 and 2017. All patients had a baseline FDG PET-CT and contrast-enhanced MRI and/or CT. All cases with reported abnormal RP LNs underwent radiology review. Abnormal RP LNs were identified in 40/402 (10%) of patients. On multivariate analysis, RP LN involvement was associated with posterior pharyngeal wall/soft palate primaries (OR 10.13 (95% CI 2.29-19.08), p = 0.002) and contralateral cervical LN involvement (OR 2.26 (95% CI 1.05-4.86), p = 0.036). T stage, largest LN size, levels of ipsilateral LN level involvement, HPV and smoking status did not predict risk. 5/402 (1.2%) patients had bilateral RP involvement. 3/402 patients (0.7%) had contralateral-only RP LNs. All patients with contralateral RP LNs had contralateral neck nodes or primary cancers extending across midline. In 5/40 (12.5%) cases with involved RP LNs, the RP LNs were superior to hard palate/upper edge of body of C1 vertebra. RP LNs were identified in 10% of oropharyngeal carcinoma patients, and were associated with contralateral neck disease and/or posterior pharyngeal wall/soft palate primary. Contralateral RP LN involvement was rare and associated with contralateral neck disease and/or primary crossing midline, suggesting potential for omission from target volumes for selected patients. Involvement of RP LNs close to the skull base highlights the need for generous elective outlining.