Hu antigen R (HuR) heterogeneous expression quantification as a prognostic marker of melanoma.

Abstract

Prognostic markers for melanoma, particularly for stage II disease, are needed for the risk-benefit evaluation of future adjuvant therapies. The mainly nuclear RNA-binding protein human antigen R (HuR) regulates the protein expression of thousands of mRNAs, its own heterogeneous expression could therefore reflect tumor heterogeneity and plasticity. Here, we evaluate its quantification in primary melanoma as a marker of metastatic outcome. We conducted an immunohistochemistry-based automated quantification of HuR nuclear expression heterogeneity in primary melanomas, most with Breslow thickness ≥ 1 mm and calculated the dimensionless fourth moment, that is, the kurtosis of HuR (HuR K) expression distribution. Twelve tumors from patients with no metastatic disease were compared to a similar number of tumors from patients who had metastatic disease at 2 years follow up. HuR K value appeared significantly higher in the non-metastatic group comparatively to the metastatic group (P = 2.84 × 10-3 , 1-tailed Wilcoxon rank-sum test). Moreover, compared to the Breslow thickness, HuR K value appeared as a more robust marker of metastatic outcome (respective areas under receiver operating characteristic curves 0.84 and 0.87). Our data need confirmation on a large cohort, however strongly suggest that HuR expression heterogeneity quantification using kurtosis, could be used as a prognostic marker in melanoma.