Differential Proteomics Identifies Protein Biomarkers That Predict Local Relapse of Head and Neck Squamous Cell Carcinomas

Clin Cancer Res. 2009 Dec 15;15(24):7666-7675. doi: 10.1158/1078-0432.CCR-09-2134.

Abstract

PURPOSE: The 5-year survival rates of head and neck squamous cell carcinomas (HNSCC) remain disappointing. HNSCCs develop in precursor fields of genetically altered cells that are often not completely resected when the tumor is excised, causing local relapse. These precursor fields are mostly recognized as dysplasia, but histologic grading cannot reliably predict malignant transformation. Our aim was to discover and validate protein biomarkers that can detect precursor fields and predict local relapse in HNSCC using immunostaining of surgical margins. EXPERIMENTAL DESIGN: We compared paired and genetically characterized normal, precursor, and tumor tissues of eight patients by proteome analysis to identify differentially expressed proteins. The prognostic value of candidate protein biomarkers was evaluated by immunohistochemical analysis of 222 surgical margins of 46 HNSCC patients who developed local relapse or remained disease free. Significant associations were determined by Kaplan-Meier survival analysis and Cox-proportional hazards models. RESULTS: Forty proteins showed significant differential expression (false discovery rate-corrected P < 0.05). Most discriminative markers suited for immunostaining were keratin 4 and cornulin. Low expression in the surgical margins of keratin 4 (hazard ratio, 3.8; P = 0.002), cornulin (hazard ratio, 2.7; P = 0.025), and their combination (hazard ratio, 8.8; P = 0.0005) showed a highly significant association with the development of local relapse. Dysplasia grading had no prognostic relevance. CONCLUSIONS: Immunohistochemical assessment of keratin 4 and cornulin expression in surgical margins of HNSCC patients outperforms histopathologic grading in predicting the risk for local relapse. These markers can be used to initiate more frequent and lifelong surveillance of patients at high risk of local relapse, and enable selection for adjuvant treatment or tertiary prevention trials. (Clin Cancer Res 2009;15(24):7666-75).