Accuracy of the non-sentinel node risk score (N-SNORE) in patients with cutaneous melanoma and positive sentinel lymph nodes: a retrospective study.

BACKGROUND

Sentinel node (SLN) biopsy in patients with melanoma permits identification of those at risk for further metastases in non-sentinel lymph nodes (NSLN). However, a mere 20% of SLN-positive patients have metastases in NSLN. Therefore we need criteria to predict NSLN-positivity. A new score system known as the non-sentinel risk score, (N-SNORE) based on five clinical and pathological characteristics (gender, regression in primary melanoma, proportion of SNs containing melanoma, perinodal lymphatic invasion, and SN tumor burden), was first published in 2010. In this study, the accuracy of N-SNORE was validated in melanoma patients with positive SLN.

METHODS

A total of 106 melanoma patients with positive SLN, who had undergone complete lymph node dissection (CLND) subsequently, were included in the study. The N-SNORE was calculated in all patients, and the risk was compared to the frequency of NSLN metastases. Statistical analysis of the data was performed.

RESULTS

Thirteen patients were at very low risk for NSN metastasis (score 0), 63 patients at low risk (score 1-3), 19 at intermediate risk (score 4-5), 6 at high risk (score 6-7), and 5 at very high risk (score >8). NSLN positivity rates for these 5 risk groups were 7.7%, 18.2%, 21.1%, 33.3%, and 80%, respectively. According to Fisher's exact test, the contingency coefficient was .322; the p-value was .025.

CONCLUSION

An increasing N-SNORE was clearly correlated with a higher risk of NSLN positivity. Based on the p-value and the contingency coefficient, the overall accuracy of the N-SNORE was proven on statistical calculation.