A Molecular Prognostic Model Predicts Esophageal Squamous Cell Carcinoma Prognosis
et al. (2014) A Molecular Prognostic Model Predicts Esophageal Squamous Cell Carcinoma
Prognosis. PLoS ONE 9(8): e106007. doi:10.1371/journal.pone.0106007
A Molecular Prognostic Model Predicts Esophageal Squamous Cell Carcinoma Prognosis
Hui-Hui Cao 0
Chun-Peng Zheng 0
Shao-Hong Wang 0
Jian-Yi Wu 0
Jin-Hui Shen 0
Xiu-E Xu 0
Jun-Hui Fu 0
Zhi-Yong Wu 0
En-Min Li 0
Li-Yan Xu 0
Jo rg D. Hoheisel, Deutsches Krebsforschungszentrum, Germany
0 1 The Key Laboratory of Molecular Biology for High Cancer Incidence Coastal Chaoshan Area, Shantou University Medical College , Shantou, Guangdong , China , 2 Institute of Oncologic Pathology, Shantou University Medical College , Shantou, Guangdong , China , 3 Department of Biochemistry and Molecular Biology, Shantou University Medical College , Shantou, Guangdong , China , 4 Departments of Oncology Surgery, Shantou Central Hospital, Affiliated Shantou Hospital of Sun Yat-sen University , Shantou, Guangdong , China , 5 Departments of Pathology, Shantou Central Hospital, Affiliated Shantou Hospital of Sun Yat-sen University , Shantou, Guangdong , China
Background: Esophageal squamous cell carcinoma (ESCC) has the highest mortality rates in China. The 5-year survival rate of ESCC remains dismal despite improvements in treatments such as surgical resection and adjuvant chemoradiation, and current clinical staging approaches are limited in their ability to effectively stratify patients for treatment options. The aim of the present study, therefore, was to develop an immunohistochemistry-based prognostic model to improve clinical risk assessment for patients with ESCC. Methods: We developed a molecular prognostic model based on the combined expression of axis of epidermal growth factor receptor (EGFR), phosphorylated Specificity protein 1 (p-Sp1), and Fascin proteins. The presence of this prognostic model and associated clinical outcomes were analyzed for 130 formalin-fixed, paraffin-embedded esophageal curative resection specimens (generation dataset) and validated using an independent cohort of 185 specimens (validation dataset). Results: The expression of these three genes at the protein level was used to build a molecular prognostic model that was highly predictive of ESCC survival in both generation and validation datasets (P = 0.001). Regression analysis showed that this molecular prognostic model was strongly and independently predictive of overall survival (hazard ratio = 2.358 [95% CI, 1.391-3.996], P = 0.001 in generation dataset; hazard ratio = 1.990 [95% CI, 1.256-3.154], P = 0.003 in validation dataset). Furthermore, the predictive ability of these 3 biomarkers in combination was more robust than that of each individual biomarker. Conclusions: This technically simple immunohistochemistry-based molecular model accurately predicts ESCC patient survival and thus could serve as a complement to current clinical risk stratification approaches.
-
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its
Supporting Information files.
Competing Interests: The authors have declared that no competing interests exist.
. These authors contributed equally to this work.
Among all types of cancer, esophageal cancer (EC) has the
eighth and sixth highest incidence and mortality rates worldwide,
respectively [1]. Although esophageal adenocarcinoma (EAC) has
become the predominant histological subtype in some western
countries, esophageal squamous cell carcinoma (ESCC) remains
dominant in China, with almost 90% of newly diagnosed patients
exhibiting this cancer subtype [2]. The 5-year survival rate for
ESCC remains dismal, despite improvements in treatments such
as surgical resection and adjuvant chemoradiation. In current
clinical practice, pathological tumor-node-metastasis (pTNM)
stage is considered the optimal prognostic indicator. However,
this clinical staging approach is limited in its ability to precisely
stratify patients for treatment options due to wide variation in
survival rates, such as that observed among T3N1 patients [3].
Clearly, identifying effective biomarkers to complement current
clinical staging approaches is highly important. According to
national guidelines [4,5], biomarkers should be sensitive, specific,
cost-effective, fast, robust against variability, and more accurate
than current clinical stages. A single biomarker, however, may be
unlikely to fulfill all of these requirements.
In recent decades, the identification of combinations of
biomarkers instead of single biomarkers has become a popular
research endeavor. Multi-gene signatures of breast cancer,
colorectal cancer, esophageal and gastroesophageal junction
adenocarcinoma, and other cancer types have served as successful
prognostic indicators [3,69]. The ability of these gene signatures
to accurately predict survival provides a foundation on which to
build molecular classification systems and (...truncated)