Abstract 176: ICG-fluorescence Imaging With Lymphoscintigraphy For Sentinel Node Biopsy In Melanoma: Reducing The False Negative Rate

Abstract

Purpose: In melanoma, sentinel lymph node (SLN) status has shown to be the best predictor of disease-free survival and identifying the correct SLN is of paramount importance to enable appropriate adjuvant work-up and treatment. Despite the advances in cutaneous melanoma management, the false negative rate (FNR) of sentinel lymph node biopsy (SLNB) is still reported to be as high as 29.8%. The majority of studies examining ICG-guided SLNB in melanoma are retrospective studies with inadequate sample size and follow-up time, making calculation of FNR difficult. Amongst studies with large cohorts of patients (>300) and at least two-years mean or median follow-up, the FNR ranges from 13.4-29.8%. Indocyanine green (ICG) fluorescence imaging has improved identification of SLNs in numerous cancers. The goal of this study was to analyze the largest cohort of patients with primary cutaneous melanoma who underwent a SLNB with the combination of lymphoscintography and ICG-based fluorescence to determine the FNR. Methods: Consecutive primary cutaneous melanoma patients who underwent radioisotope lymphocintigraphy and ICG-based fluorescence imaging for SLNB by the senior author from 2012-2018 were prospectively enrolled. All patients were staged according to AJCC version 8. The FNR was calculated based on the formula FN/TP+FN. Results: 594 melanomas were analyzed, of which there were 130 T1a, 114 T1b, 148 T2a, 34 T2b, 46 T3a, 56 T3b, 22 T4a and 43 T4b. At least one SLN was identified in every patient. 1827 nodes were sampled. 1556 (85.2%) were identified by radioactivity/fluorescence, 255 (14%) by radioactivity only and 16 (0.9%) with fluorescence only. There were 163 positive sentinel nodes. 147 (90.2%) were identified by radioactivity/fluorescence, 13 (8%) by radioactivity only and 3 (0.6%) with fluorescence only. Of the 128 true positive patients, 116 (90.6%) had at least one positive node identified by radioactivity/fluorescence, 8 (6.3%) by radioactivity only and 4 (3.4%) with fluorescence only. There were 128 true positive, 454 true negative and 12 false negative patients. The FNR was 8.6%. Mean follow up was 1030.9 days. Conclusion: In our cohort, while the majority of positive nodes were identified by both radioactivity and fluorescence, for 12 patients, their only positive node was identified by only radioactivity or fluorescence. Thus, without dual modality, these nodes may have been missed and patients may not have been accurately staged and allocated adjuvant therapy. The FNR of our study at 8.6% is lower than that of any large cohort study with at least two-year follow-up, including the Sunbelt Melanoma trial, the largest prospective study examining nodal recurrence in melanoma patients, which demonstrated a FNR of 10.8%. In conclusion, in the study of the largest cohort of patient with primary cutaneous melanoma who underwent a SLNB with radioisotope lymphocintigraphy and ICG-based technology, we demonstrate the lowest reported FNR amongst large studies with at least two-year follow-up. This has important implications for melanoma patients as the widespread adoption of this technique with subsequent accurate staging, adjuvant work-up and treatment may improve survival outcomes.