Sentinel node biopsy in melanoma: a single‐centre experience with 216 consecutive patients

DEAR EDITOR, Sentinel node biopsy (SNB) is an established staging procedure in malignant melanoma, with sentinel node status being the most important predictor of survival in clinically node-negative patients. SNB is associated with only minor morbidity and is praised for sparing a substantial number of patients from overtreatment with immediate complete lymph node dissection and its complications. As results from large randomized trials on the clinical benefit of SNB in melanoma are scarce, the aim of the present study was to add our singlecentre experience to the published evidence. In a situation where a procedure has gained widespread acceptance, albeit with decisive evidence lagging behind and being difficult to obtain, the collection of retrospective information is of particular importance. Our institution is a tertiary referral centre for patients with melanoma covering a population of about 600 000; up to 100 SNBs are performed each year. In a retrospective chart review we retrieved clinical data on patients with melanoma who underwent SNB at our institution between July 2009 and June 2012 (36 months). Throughout the study period the institutional criteria for performing SNB were tumour thickness (Breslow depth) ≥ 1 mm, or in melanomas < 1 mm the presence of either ulceration or mitoses ≥ 1 mm , or Clark level ≥ IV in rare cases. For sentinel node mapping we used colloidal 99Tc (Tc-99m as the pertechnetate) injected the day before and a blue dye (Patentblau V; Guerbet, Roissy CdG, France) injected immediately before surgery. Sentinel nodes were processed according to Starz et al., with semiserial-section (200 lm) histology, and nodes were considered positive on (immuno) histopathological detection of tumour cells (from isolated tumour cells to large aggregates). Statistical analysis was performed using GraphPad Prism version 5.1 for Windows (GraphPad Software, La Jolla, CA, U.S.A.). In total 216 patients (96 female, 43 4% of procedures) were identified, of whom five male patients underwent the procedure twice because of a second primary melanoma during the study period. The patients’ ages ranged from 9 to 86 years (median 64 9) and the age distribution differed between female [median 59 5 years, interquartile range (IQR) 41 9–71 9] and male patients (median 67 0 years, IQR 55 1– 74 9). Melanomas were located on the trunk (47%), the upper and lower extremities (19% each) and the head and neck (14%), with a significant sex-specific difference in distribution (relative predominance for the legs in female patients and trunk in male patients; v-test, P = 0 0014). Melanoma thickness ranged from 0 2 mm to 15 mm (median 1 5). Sixty-two procedures in our series involved T1 melanomas (thickness 0 2–1 0 mm, median 0 77). Overall 17 2% of tumours showed ulceration (female 14 6%, male 20 0%). In 204 of 221 SNB procedures (92 3%) a sentinel node was successfully retrieved. Among the 17 cases where detection or recovery of a sentinel node was not possible, 10 were located in the cervical region. In 25 melanomas (11 3%) located on the trunk, the radionuclide identified lymphatic drainage to two nodal sites, and in two cases to three sites. Sentinel node biopsy revealed metastases in 21 7% of all cases (48 procedures; 18 female, 38%). In patients with thin melanomas (thickness ≤ 1 mm, n = 62) sentinel nodes were positive in five cases (8% of all thin melanomas). Sentinel node involvement was significantly correlated with T stage