LAUSR

BACKGROUND Patients with stage II melanoma have a considerable risk for recurrence. Current guidelines are imprecise as to optimal follow-up. We hypothesized that by examining recurrence patterns, we could help to better inform guidelines. STUDY DESIGN We queried IRB-approved melanoma databases of Thomas Jefferson University and University of North Carolina, identifying 581 patients with stage II melanoma between 1996 and 2015 with at least 1 year of follow-up. Data included location of first recurrence and how recurrence was detected (ie patient symptom, physician examination, or routine surveillance imaging). Cox regression with backward elimination was used for multivariable analysis. RESULTS One hundred and seventy-one patients had a recurrence (29.4%), the incidence increased considerably by stage sub-group. Significant predictors of recurrence included male sex (p = 0.003), ulceration (p = 0.03), and stage (p < 0.001). On multivariable analysis, male sex and stage continued to be significant (p < 0.01). For overall survival, regression, ulceration, stage, and age were significant predictors of survival. Stage, regression, and age remained significant by multivariable analysis. Patient symptoms were the most frequent mode of detection (40%), followed by physician examination (30%) and surveillance imaging (26%)-this did not differ significantly by stage. Regional nodes were the most common site of recurrence (30%), followed by lung (27%) and in-transit (18%). CONCLUSIONS The majority of recurrences in stage II melanoma are detected by patients and their physicians and rarely by routine imaging. As such, clinical follow-up and patient education are critical factors in detection of recurrence. With the prevalence of regional nodal recurrences, ultrasound might prove to be an important strategy in early recurrence detection.