Upregulation of CRISP3 and its clinical values in adult sepsis: a comprehensive analysis based on microarrays and a two-retrospective-cohort study.

Abstract

Current lines of evidence indicate that cysteine-rich secretory protein 3 (CRISP3) is an immunoregulatory factor. Nevertheless, no study has explored the relationships between the values of CRISP3 and sepsis. We conducted a comprehensive literature search and meta-analysis from the Gene Expression Omnibus (GEO) and ArrayExpress to determine the expression of CRISP3 in sepsis patients. Then, we explored whether plasma CRISP3 could serve as a potential biomarker to predict the risk of sepsis via two retrospective trauma cohorts. We evaluated the prediction power using the area under the curve (AUC). A total of 23 datasets were recruited for the comprehensive meta-analysis, and the combined standardized mean difference (SMD) of CRISP3 was 0.90 (0.50-1.30) (p < 0.001), suggesting that CRISP3 was overexpressed in sepsis patients. Meanwhile, sepsis patients had higher CRISP3 concentrations than non-sepsis patients in 54 trauma patients (p < 0.001). Plasma CRISP3 on admission was significantly associated with the incidence of sepsis [OR = 1.004 (1.002-1.006), p < 0.001]. As a predictive biomarker, CRISP3 obtained a better AUC [0.811 (0.681-0.905)] than C-reactive protein (CRP) [0.605 (0.463-0.735)], procalcitonin (PCT) [0.554 (0.412-0.689)], and Sequential Organ Failure Assessment (SOFA) [0.754 (0.618-0.861)]. Additionally, the clinical relationships between plasma CRISP3 and sepsis were verified in another trauma cohort with 166 patients [OR = 1.002 (1.001-1.003), p < 0.001]. The AUC of CRISP3 was 0.772 (0.701-0.834), which was better than that of CRP [0.521 (0.442-0.599)] and PCT [0.531 (0.452-0.609)], but not SOFA [0.791 (0.717-0.853)]. Our study indicated and validated that CRISP3 was highly expressed in sepsis. More importantly, CRISP3 may serve as a latent biomarker to predict the risk of sepsis.