Sentinel lymph node biopsy is unreliable in predicting melanoma mortality for both younger and older patients

Abstract

AbstractBackgroundMelanoma disease patterns vary with patient age.AimTo evaluate sentinel lymph node biopsy (SLNB) in managing melanoma at differing patient ages.MethodsOnline prediction tools were applied to compare SLNB positivity (SLNB+) and survival risk at patient ages 20–80. Tübingen melanoma data were used to determine variations in the hazard ratio of SLNB+ for mortality at different patient ages.ResultsRegardless of tumour thickness, predicted SLNB+ rates were markedly higher than mortality rates for 20‐year‐old patients. For 80‐year‐old patients, it is the opposite.DiscussionIf 1000 20‐year‐olds with a 0.4 mm thickness non‐ulcerated melanoma underwent SLNB, 100 would likely be positive. If all 100 were to be offered adjuvant drug therapy (ADT), fewer than three more melanoma deaths in those 1000 patients would be avoided. In total, 97 patients would have received medication they may never have needed. If 1000 80‐year‐olds with a 3 mm thickness non‐ulcerated melanoma underwent SLNB, only 40 would likely be positive. In total, 274 patients would be predicted to die of melanoma, 245 being SLNB negative and 29 SLNB+. ADT linked to SLNB+ could deny treatment to 89% of these high‐risk patients.LimitationsThe authors relied on published risk data.ConclusionSLNB has poor specificity at predicting mortality in young melanoma patients and poor sensitivity in older patients. SLNB is not indicated in managing cutaneous melanoma for patients under 40 or over 60 years of age. Many such patients could be managed with wide local excision alone in their clinician's office‐based practice. For all cutaneous melanoma patients at all ages, linking ADT to BAUSSS biomarker, (an algorithm of Breslow thickness, age, ulceration, subtype, sex and Site) rather than SLNB+ is likely more appropriate. BAUSSS provides a more accurate melanoma‐specific mortality risk assessment for patients without burdening them with added surgery, hospitalization, costs or morbidity risk.