Level IB nodal involvement in oropharyngeal carcinoma: implications for submandibular gland-sparing intensity-modulated radiotherapy.

OBJECTIVES/HYPOTHESIS: Submandibular gland-sparing intensity-modulated radiotherapy (SMG-sparing IMRT) has been proposed to reduce xerostomia following head and neck irradiation. However, the safety of this practice has been questioned. Data from a large surgical series of oropharyngeal carcinoma patients were extracted to identify clinicopathological correlates for submandibular involvement and to create a risk stratification scheme to guide decision making to refine selection guidelines for SMG-sparing IMRT.

STUDY

DESIGN

Clinicopathologic analysis.

METHODS

The medical records of 153 consecutive patients with squamous cell carcinoma of the oropharynx treated by primary surgery and neck dissection were reviewed. Univariate and multivariate analysis was performed with logistic regression to identify factors predictive of submandibular involvement. Recursive partitioning was used to develop risk stratification schemas based on preoperative data alone and in combination with pathologic data to guide treatment decisions in the definitive and postoperative settings, respectively.

RESULTS

Submandibular (level IB) nodal dissection was performed in 119 heminecks (85 ipsilateral and 17 contralateral). The incidence of submandibular involvement was 18%. Young age, T3-4 disease, N2b-3 disease, and perineural invasion were identified as risk factors for submandibular nodal involvement on multivariate analysis (P < .01). Three distinct risk groups for submandibular involvement were identified: age >60 years and N0-2a disease (low risk, 2%), age ≤60 years and T1-2N2b-3 (intermediate risk, 16%), age ≤60 years and T3-4N2b-3 disease (high risk, 57%).

CONCLUSIONS

These data provide assurances that SMG-sparing IMRT can reasonably be offered to appropriately selected patients. Risk stratification schemas were successfully developed for SMG-sparing IMRT in both the definitive and adjuvant settings.