Significant prognostic value of cell-cycle proteins in early-stage small cell carcinoma of cervix

Small cell carcinoma of cervix (SCCC) was a highly aggressive tumor with dismal prognosis. Current treatment strategies manifested poor survival outcomes and novel treatment options were needed exploration. We aimed to investigate several prognostic biomarkers for SCCC and conducted a novel risk-score system to predict cancer specific survival (CSS) in early-stage SCCC. Seven cell-cycle proteins were detected by immunohistochemistry in 88 SCCCs. Univariate and multivariate analysis were performed to identify prognostic proteins and establish a predicting model. Total patients were divided into two groups by the median risk-score: the high-risk group and the low-risk group. Logistic regression and Wilcoxon test were used to investigate the association between clinical variables and the risk-score system. Seven cell cycle proteins were overexpressed in SCCC. The expression of CDC20, MAD2L1, MCM2 and BUBR1 were correlated to survival outcomes with P < 0.05. A novel risk-score system consisting of CDC20, MAD2L1 and BUBR1 was significantly an independent prognostic factor for CSS and the high-risk group possessed worse survival (P < 0.001). The c-indexes for clinical model, risk-score system and the combined model were 0.668, 0.718 and 0.727, respectively. The AUCs for these three models were 0.730, 0.775 and 0.823, respectively. Furthermore, we discovered that patients with high-risk scores were inclined to possessing older age, parametrial invasion and higher FIGO stage (IIA vs IA/IB) with P < 0.05. This risk-score system consisting of CDC20, MAD2L1 and BUBR1 presented good discrimination and predictability for SCCC. Novel biomarkers in this study might have some merits in providing guidance of novel treatment strategies for SCCC.