A three-phase trans-ethnic study reveals B7-H3 expression is a significant and independent biomarker associated with colon cancer overall survival

Abstract

Background: There have been inconsistent results and conflicting conclusions among the existing prognostic studies of B7-H3 expression in colon cancer patients. Therefore, the association between B7-H3 expression and colon cancer survival has remained largely unclear.Methods: We performed a three-phase and trans-ethnic prognostic study of B7-H3 expression in colon cancer patients involving perhaps the largest population to date. In the discovery phase, we utilized a Cox proportional hazards model adjusted for covariates to test the association between B7-H3 expression and colon cancer overall survival (OS) time in a European population from The Cancer Genome Atlas (TCGA) cohort (n=433). In the validation phase I, the association was replicated in a European population from Gene Expression Omnibus (GEO) cohort (n=811). In the validation phase II, we again confirmed the significant association in an Asian population from Fujian Medical University Union Hospital (UNION) cohort (n=179).Furthermore, a series of Kaplan-Meier analysis, bioinformatics analysis of tumor immune microenvironment (TIME), and immune checkpoint prognostic prediction analysis, as well as sensitivity analysis, were also conducted.Results: Highly expressed B7-H3 was a significant and robust biomarker to colon cancer survival, with a large hazard ratio (HR) [HR TCGA =4.60, 95% confidence interval (CI): 2.15 to 9.83, P=8.37×10 −05 ; HR GEO =1.47, 95% CI: 1.12 to 1.94, P=0.0056; HR UNION =1.63, 95% CI: 1.36 to 1.95, P=7.91×10 −08 ]. We detected an involvement of B7-H3 in the tumor immune microenvironment (TIME). Meanwhile, B7-H3 was significantly and weakly correlated with 6 out of 27 well-recognized immune checkpoint genes. Even after adjusting for effects of other immune checkpoint genes, B7-H3 still exhibited a harmful effect on colon cancer survival using samples from TCGA and GEO cohorts (HR =1.47, 95% CI: 1.07 to 2.02, P=0.0184), indicating that it was an independent prognostic factor of colon cancer. We also proposed an immune checkpoint prognostic risk score which possessed the capability to identify colon cancers with high risk of mortality. Conclusions:The expression of B7-H3 is a significant, robust, and independent prognostic factor to colon cancer OS.

Tissue-Based Markers as a Tool to Assess Response to Neoadjuvant Radiotherapy in Rectal Cancer—Systematic Review

“…Some other studies found that B7-H3, B7-H4 and B7-H7 expression in colorectal cancer was significantly upregulated as compared with normal tissues. The high expression of these molecules was correlated with a poor outcome in patients with colorectal cancer [ 109 , 110 , 111 , 112 ]. In 2020, Yasui et al published a study whose aim was to investigate the effect of chemoradiotherapy on the immunological status of rectal cancer patients who were treated with preoperative chemoradiotherapy.…”

B7-H3 enhances colorectal cancer progression by regulating HB-EGF via HIF-1α

“…B7-H3 is a vital factor in CRC which not only functions as a marker of CRC prognosis and clinicopathological stage ( 9 - 11 ), but also promotes CRC in many ways ( 7 , 8 , 12 ). In CRC, B7-H3 serves as a significant biomarker for survival ( 10 ). As an immune checkpoint protein, B7-H3 can inhibit the function of anti-tumor immune cells in many ways ( 30 ).…”

“…As a transmembrane protein, B7-H3 is expressed in the pro-B7-H3 form on the membrane of various tumors, and further cleaved by matrix metalloproteinase (MMP) to yield soluble B7-H3 that performs cellular functions ( 6 ). In CRC, B7-H3 promotes the tumor progression ( 7 , 8 ) and is associated with recurrence and overall survival ( 9 , 10 ). Highly B7-H3 expression is associated with enhanced lymph node metastasis and poor tumor differentiation ( 11 ).…”

Mining novel biomarkers for prognosis of skin cutaneous melanoma with proteomic analysis

“…We analyzed the correlation between the expression values of prognosis-related proteins involved in the construction of prognostic models and various clinical phenotypes (including patient age, gender, stage, etc. ), calculated the correlation coe cients and P-values between them, and drew correlation boxplots to visualize the results with statistical differences 29,30 .…”