An immunohistochemical score to predict the outcome for oral squamous cell carcinoma

Abstract

Oral cancer is a major public health problem worldwide, with a poor survival. Our aim was to evaluate several protein markers in oral squamous cell carcinomas (OSCC) and analyse their prognostic value on patient's survival. We analysed the expression of EGFR, p53, p27, p16, cyclin D1, cyclin A2, COX-2, Ki-67, Bcl-2, VEGFR-1 and VEGFR-2, by immunohistochemistry on 67 primary OSCC. Cancer-specific survival (CSS) analysis was evaluated by the Cox regression model. Markers showed variable expression between 27.9% and 95.2%. In univariate analysis for CSS, we found that four of the tested markers, namely high expression of p53 (P = .001), EGFR (P = .003), cyclin A2 (P = .005) and low expression of p16 (P = .019), along with clinical stage (P < .001), tumour size (P < .001), presence of nodal metastasis (P < .001) and perineural permeation (P = .039) were related to decreased survival. On the basis of these results, we constructed an immunohistochemical score hinging on the possibility that any tumour could express none of these four markers (score 0), one or two markers (score 1) and three or more markers (score 2). In multivariable analysis, this immunohistochemical score revealed an independent prognostic value on cancer-specific survival (P = .001; HR: 3.7: 95%CI 1.7-7.9). Moreover, we confirmed that in early-stage tumours (stage I or II) this score maintained its independent prognostic value (P = .025; HR: 7.9, 95%CI 1.3-49.1) on CSS. The expression of the markers p53, p16, EGFR and cyclin A in OSCC, combined to give an immunohistochemical score, may identify high-risk subgroups for decreased survival and to further guide therapeutic decisions.