Outcomes of Biopsy Techniques Prior to Sentinel Lymph Node Biopsy for Primary Cutaneous Melanoma of the Head and Neck

Presentation: AHNS040
Topic: Skin Cancer
Type: Oral
Date: Thursday, April 19, 2018
Session: 7:45 AM - 8:35 AM Skin Cancer
Authors: Matthew May, MD, Jeffrey Janus, MD
Institution(s): Mayo Clinic Rochester MN

Background: Sentinel lymph node biopsy (SLNB) is one of the most prognostic factors in respect to survival in patients with primary melanoma. Previous wide local excision (WLE) prior to SLNB may have the potential to disrupt lymphatic channels, thus incorrectly identifying the sentinel node.  The purpose of this study is to assess the accuracy of WLE compared to other biopsy techniques in the identification of the SLN and survival outcomes in patients with melanoma of the head and neck.

Methods: Between the years 2000-2016, records of 391 cases of SLNB were reviewed in patients with primary cutaneous melanoma of the head and neck.  SLNB was performed with lymphoscintigraphy and dye injection with a median tumor thickness of 1.9 mm and median follow up time of 30 months.  Biopsy practices prior to SLNB (shave, punch, wide local excision, and excisional/Mohs), clinicopathologic features , location in the head and neck, SLN identification, nodal disease, incidence and sites of re-occurrence, and associations with time regional relapse and time to death from melanoma were evaluated using chi squared, Fisher exact tests, and Cox proportional hazard regression models.

Results: Of the 391 patients identified, biopsy patterns were as follows: 77 (19%) unknown biopsy, 30 (8%) prior WLE, 105 (27%) excisional biopsy, 69 (18%) punch biopsy, and 110 (28%) shave biopsy. SLNB was successfully identified in all 30 patients whom had a prior WLE. The median depth for WLE and excisional biopsy was significantly different compared to punch and shave biopsy (p< .001).  Age, sex, mitotic index, ulceration, positive nodal status, and time from diagnosis to SLNB were not significantly different between biopsy practices (Table 1). At last follow up, there were 50 regional recurrences in the neck and 27 local recurrences with the median (IQR) at 1.2 years and 1.0 years, respectively following SLNB. Regional recurrence free survival rates (95% CI; number still at risk) at 2,4,6,8 and 10 years following SLNB were 87% (83-91; 186), 83% (78-88; 117), 78% (72-85; 63), 77% (71-84; 33), and 77% (71-84; 19), respectively. Univariable and multivariable associations of type of prior biopsy, depth of invasion, and nodal status with time to regional recurrence, local recurrence, death from any cause, and death from malignant melanoma are summarized in Table 2. Four of the 30 patients with WLE prior biopsies experienced regional recurrence at a median of 0.7 years following SLNB; 20 of the 109 patients with Mohs/excisional prior biopsies experience regional recurrence at a median of 1.3 years following SLNB; 19 of the 179 patients with punch/shave prior biopsies experienced regional recurrence at a median of 1.2 years following SLNB. There was no significant difference between type of biopsy and regional or local recurrence.

Conclusions: SLNBs of the head and neck can be successfully performed in patients with primary cutaneous melanoma after previous wide local excision. We continue to recommend performing a wide local excision and sentinel lymph node biopsy in a single stage procedure to minimalize mortality and morbidity of staging procedures for malignant melanoma of the head and neck.