Lymph node yield is an important prognostic factor in head and neck squamous cell carcinoma. Variability in neck dissection sampling techniques has not been studied as a determinant of lymph node yield.

This retrospective study used lymph node yield and average nodes per level to compare level-by-level and en bloc neck dissection sampling methods, in primary head and neck squamous cell carcinoma cases operated between March 2017 and February 2020.

From 123 patients, 182 neck dissections were analysed, of which 133 were selective and the rest were comprehensive: 55 had level-by-level sampling and 127 had undergone en bloc dissection. The level-by-level method yielded more nodes in all neck dissections combined (20 vs 17; p = 0.097), but the difference was significant only for the subcohort of selective neck dissection (18.5 vs 15; p = 0.011). However, the gain in average nodes per level achieved by level-by-level sampling was significant in both groups (4.2 vs 3.33 and 4.4 vs 3, respectively; both p < 0.001).

Sampling of cervical lymph nodes level-by-level yields more nodes than the en bloc technique. Further studies could verify whether neck dissection sampling technique has any impact on survival rates.

This study investigated the incidence and routes of submandibular gland involvement in oral cavity carcinoma to determine the feasibility of submandibular gland sparing neck dissection.

The records of 155 patients diagnosed with oral cavity squamous cell carcinoma, with a total of 183 neck specimens, including those involving level I, were reviewed retrospectively.

Submandibular gland involvement, via direct invasion from the anatomical proximity of T4a tumours, was evident in two patients. The floor of mouth location, either primarily or as an extension of the primary tumour, was the only risk factor for submandibular gland involvement in oral cavity carcinoma (p = 0.042). Tumour location, clinical and pathological tumour (T) and nodal (N) stages, and radiological suspicion of mandible invasion, were not found to be statistically relevant (p > 0.05).

The results suggest the feasibility of preserving the submandibular gland in early stage oral cavity carcinoma unless the tumour is located in, or extends to, the floor of mouth.