PET/CT Prediction of Head and Neck Melanoma Nodal Metastases

Abstract

<h3>Purpose/Objective(s)</h3> Lymph node dissection is an important oncologic treatment for many head and neck melanomas. Over time, surgeons have narrowed the basins covered by neck dissections based on the underlying disease. In melanoma, there is a need for diagnostic modalities that can effectively predict locoregional metastasis to tailor the extent of neck dissection. PET-CT is an imaging modality often employed in oncologic staging, but definitive data are lacking on basin-specific predictive capabilities of PET-CT for head and neck melanoma. The objective of this study is to determine the predictive value of PET-CT for detecting melanoma metastasis to head and neck lymph node basins. <h3>Materials/Methods</h3> This is a retrospective case series of adults who underwent lymphadenectomy for melanoma metastasis to the neck at a single tertiary academic medical center from 2000-2018. Exclusion criteria included: prior head and neck radiation or lymphadenectomy except for sentinel lymph node biopsies, lack of preoperative PET-CT images, incomplete pathology records, or less than one year of follow-up. Demographic data, clinical and pathologic staging, surgical details, and post-operative complications were collected. The eighth edition of the American Joint Committee on Cancer staging system was used. PET-CT was performed according to institutional protocol, and all PET-CT images were re-reviewed by a board-certified nuclear medicine physician blinded to pathology results. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of PET-CT for predicting nodal metastasis were calculated by comparing the results of the PET-CT to final pathology, which was considered the gold standard for this study. <h3>Results</h3> 108 subjects were included, 2 of whom had bilateral neck dissections, for a total of 110 neck dissections. Median follow-up was 2.0 years. The average age (SD) was 63 (14), 74% were male, and 93% identified as Caucasian. The most common primary sites were the face (n=30), scalp (n=29), and neck (n=23). The majority of tumors were classified as T2-T4 (77%) and as either cN1 (n=55) or cN0 (n=26) disease. The most frequent surgical complications included seroma (n=7) and hematoma (n=5). There were no reports of gross nerve injury or permanent neuropathies. When compared to final pathology, preoperative PET-CT had a 90% sensitivity (CI 79.5%-95.6%), 92% specificity (CI 80.7 %-97.4%), 93.1% PPV (CI 83.1%-97.8%), and 88.5% NPV (CI 76.7%-95%). In 86.5% of neck dissections, there was concordance on nodal basin metastases between PET-CT and final surgical pathology. <h3>Conclusion</h3> PET-CT demonstrates acceptable predictability of locoregional head and neck melanoma metastases with nodal basin concordance. This information could be incorporated into shared decision making regarding the extent of regional lymph node dissection as treatment paradigms shift.