Validation of Invasive Area for Predicting Sentinel Node Status and Survival in Primary Cutaneous Melanoma

Two-dimensional histologic features have demonstrated independent prognostic value for survival in primary cutaneous melanoma, but their predictive value for sentinel node (SN) status has yet to be validated. We aimed to demonstrate the predictive value of the previously described calculated tumor area (CTA), and the novel Simplified Breslow Area (SBA), for SN metastasis and survival.

A total of 177 primary melanomas were assessed for standard histological characteristics, maximum invasive width (IW) of the primary tumor and CTA. We simplified CTA measurement by transforming IW with Breslow thickness (BT) [ln(IW) + ln (BT)], yielding SBA. Multivariate analysis was undertaken to assess the performance of CTA and SBA, respectively, as independent predictors of both SN status and survival outcomes.

The SN + rate was 18.1% (32/177). The median CTA for SN-patients was 3.2 mm² (IQR 1.2-10.9) compared with 6.7 mm² (IQR 4.2-26.9) for SN + patients (p < 0.01). Maximum threshold analysis identified an optimal CTA cutoff point of 6.3 mm² for disease-specific (DSS) [HR 1.01 (1.00-1.02); p = 0.008], distant metastasis-free [HR 1.01 (1.00-1.02); p = 0.005], and disease-free survival [HR 1.01 (1.00-1.02); p = 0.005]. The 5-year DSS for low-risk CTA tumors was 91.2% versus 61.3% for high-risk tumors. Cox regression showed CTA [HR 3.5 (1.21-10.81); p = 0.021] and ulceration status (US) were independent predictors of DSS. Similar results were obtained for SBA, which, on multivariate analysis, was the single most important predictor of SN status outperforming lymphovascular invasion, US, and BT.

The two-dimensional histologic features CTA and SBA are independently prognostic for survival in primary cutaneous melanoma, and SBA may be a better predictor of SN status than BT.