Practice variation in defining sentinel lymph nodes on lymphoscintigrams in oral cancer patients
PURPOSE To validate lymphatic mapping combined with sentinel lymph node biopsy as a staging procedure, and to evaluate the possible clinical implications of added oblique lymphoscintigraphy and/or tomography and test the intra- and interobserver reproducibility of lymphoscintigraphy. MATERIAL AND METHODS Forty patients (17 F and 23 M, aged 32-90) with 24 T1 and 16 T2 squamous cell carcinoma of the oral cavity. Planar lymphoscintigraphy, emission and transmission tomography were performed. Detection and excision of the sentinel nodes were guided by a gamma probe. The sentinel nodes were step-sectioning and stained with hematoxylin and eosin and cytokeratin (CK 1). Histology and follow-up were used as "gold standard". Tumor location, number of sentinel lymph nodes, metastasis, and recurrences were registered. Two observers evaluated the lymphoscintigraphic images to assess the inter-rater agreement. RESULTS Eleven (28%) patients were upstaged. The sentinel lymph node identification rate was 97.5%. Sentinel lymph node biopsy significantly differentiated between patients with or without lymph node metastasis (P = 0.001). Lymphatic mapping revealed 124 hotspots and 144 hot lymph nodes were removed by sentinel lymph node biopsy. Three patients developed a lymph node recurrence close to the primary tumor site during follow-up. Added oblique lymphoscintigraphic images and/or tomography revealed extra hotspots in 15/40 (38%) patients. In 4/40 (10%), extra contralateral hotspots were detected. CONCLUSION Sentinel lymph node biopsy upstaged 28% of the patients. Sentinel lymph nodes close to the primary tumor were difficult to find. Added oblique planar images and/or tomographic images revealed extra clinical relevant hotspots in 38% of patients. Reproducibility proved excellent.