Novel biomarkers are needed to improve current imperfect risk prediction models for cancer-associated thrombosis (CAT). We recently identified an RNA-sequencing profile that associates with CAT in... Show moreNovel biomarkers are needed to improve current imperfect risk prediction models for cancer-associated thrombosis (CAT). We recently identified an RNA-sequencing profile that associates with CAT in colorectal cancer (CRC) patients, with REG4, SPINK4, and SERPINA1 as the top-3 upregulated genes at mRNA level. In the current study, we investigated whether protein expression of REG4, SPINK4 and alpha-1 antitrypsin (A1AT, encoded by SERPINA1) in the tumor associated with CAT in an independent cohort of CRC patients. From 418 patients with resected CRC, 18 patients who developed CAT were age, sex, and tumor stage-matched to 18 CRC patients without CAT. Protein expression was detected by immunohistochemical staining and scored blindly by assessing the H-score (percentage positive cells*scoring intensity). The association with CAT was assessed by means of logistic regression, using patients with an H-score below 33 as reference group. The odds ratios (ORs) for developing CAT for patients with A1AThigh, REG4high, SPINK4high tumors were 3.5 (95%CI 0.8–14.5), 2.0 (95%CI 0.5–7.6) and 2.0 (95%CI 0.5–7.4) when compared to A1ATlow, REG4low, SPINK4low, respectively. The OR was increased to 24.0 (95%CI 1.1–505.1) when two proteins were combined (A1AThigh/REG4high). This nested case–control study shows that combined protein expression of A1AT and REG4 associate with CAT in patients with colorectal cancer. Therefore, REG4/A1AT are potential biomarkers to improve the identification of patients with CRC who may benefit from thromboprophylaxis. Show less
