Tumor tissue hnRNP M and HSP 90α as potential predictors of disease-specific mortality in patients with early-stage cutaneous head and neck melanoma: A proteomics-based study

Oncotarget. 2019 Nov 19;10(62):6713-6722. doi: 10.18632/oncotarget.27333.

Abstract

Background: Breslow tumor thickness and mitotic rate are standardly used for risk stratification of patients with malignant melanoma. However, their prognostic value is relatively limited and a need for improved prognostication has been advocated. We aimed to screen the tumor tissue proteome in a search for potentially useful prognostic factors in early-stage cutaneous head and neck melanoma.

Methodology and findings: Proteomic profiles of archival formalin-fixed tissue samples of 31 patients (age 23-90 years) with early-stage head and neck cutaneous malignant melanoma (American Joint Committee on Cancer, AJCC, stage I/II) were determined and expression intensities were compared to those of melanocytic nevi, yielding ratios used in data analysis. Medical charts were retrospectively reviewed to determine time elapsed since diagnosis to disease-specific death or censoring. In a multivariate recursive partitioning analysis (as per AJCC guidelines), higher expression levels of heterogeneous nuclear ribonucleoprotein M (hnRNP M) [n = 18, HR = 1.94 vs. the entire cohort; HR = 5.95 (95%CI 2.43-14.5) for "high" vs. "low" (n = 13)] and of heat shock protein 90 alpha (HSP 90α) [n = 17, HR = 2.09 vs. the entire cohort; HR = 4.59 (95%CI 1.87-11.2) for "high" vs. "low" (n = 14)] were each independently strongly associated with higher mortality (accounting for clinical and standard pathohistological features). Higher Breslow thickness and mitotic rate were associated with higher mortality only when proteomic data were disregarded.

Conclusions and significance: Data suggest that tumor tissue expression of hnRNP M and/or of HSP 90α deserve further investigation and clinical validation as potential novel risk stratification aids in patients with stage I-II cutaneous head and neck malignant melanoma.

Keywords: biomarkers; early stage; melanoma; proteomics; tissue.