Abstract
2701
Introduction: Initial staging in patients diagnosed with metastatic melanoma includes whole body imaging with PET-CT to evaluate for the extent of lymph node involvement and distant metastatic disease. Physical exam has proven unreliable, with 20% of clinically node-negative patients having metastatic involvement.1 Lymphatic mapping followed by sentinel lymph node biopsy (SLNB) are currently standard of care for metastatic melanoma. Growing evidence about sentinel lymph node involvement in breast cancer suggests that certain imaging features of nodes may help predict nodal metastatic positivity and reduce the need for lymph node dissection.2,3 This study investigates if SPECT/CT lymphoscintigraphy can help predict diseased sentinel nodes in patients with malignant melanoma.
Methods: We performed a retrospective, cross-sectional chart review of patients known to have a lymphoscintigraphy scan from 2019 to 2021 at a single center in a major metropolitan area of the United States. Inclusion criteria consisted of a diagnosis of melanoma, history of a SPECT/CT lymphoscintigraphy, and subsequent sentinel lymph node biopsy. Of the 181 patients known to have a lymphoscintigraphy scan within the time frame, 15 patients were excluded because a sentinel node biopsy was not performed, 8 because a sentinel node was not identified, 8 because a sentinel node identified on SPECT imaging was not definitively visualized on the corresponding CT scan, and 3 because no SPECT/CT imaging was performed. The final sample included 147 cases.
Results: Among the 147 patients who underwent a SPECT/CT lymphoscintigraphy with SLNB, a positive lymph node was found in 21 patients (14.3%), comparable to the estimated incidence of nodal involvement.4 In 13 patients, the node was positive and had suspicious imaging features. Specifically, in 11 patients the positive node had a nodular morphology and a different shape than the contralateral node. In 2 patients, the positive node was normal in size and shape but a corresponding contralateral node was absent. The remaining 8 positive nodes appeared benign on imaging, with size and shape similar to the contralateral node. There were 6 false positive cases where the node appeared nodular, but the biopsy was negative. The overall sensitivity and specificity of using SPECT/CT lymphoscintigraphy to predict diseased nodes were 61.9% and 95.2% respectively.
Conclusions: Using SPECT/CT lymphoscintigraphy, radiologists were able to accurately identify sentinel nodes involved by malignant melanoma in 90.5% of cases. These results suggest that high-quality SPECT/CT lymphoscintigraphy can reduce the need for diagnostic surgical intervention in initial staging for patients diagnosed with malignant melanoma. Future studies with a larger sample size are warranted to establish proof of concept.