The predictive value of selected serum microRNAs for acute GVHD by TaqMan MicroRNA arrays

Abstract

Currently, the diagnosis of acute graft-versus-host disease (aGVHD) is mainly based on clinical symptoms and biopsy results. This study was designed to further explore new no noninvasive biomarkers for aGVHD prediction/diagnosis. We profiled miRNAs in serum pools from patients with aGVHD (grades II-IV) (n = 9) and non-aGVHD controls (n = 9) by real-time qPCR-based TaqMan MicroRNA arrays. Then, predictive models were established using related miRNAs (n = 38) and verified by a double-blind trial (n = 54). We found that miR-411 was significantly down regulated when aGVHD developed and recovered when aGVHD was controlled, which demonstrated that miR-411 has potential as an indicator for aGVHD monitoring. We developed and validated a predictive model and a diagnostic model for aGVHD. The predictive model included two miRNAs (miR-26b and miR-374a), which could predict an increased risk for aGVHD 1 or 2 weeks in advance, with an AUC, Positive Predictive Value (PPV), and Negative Predictive Value (NPV) of 0.722, 76.19 %, and 69.70 %, respectively. The diagnostic model included three miRNAs (miR-28-5p, miR-489, and miR-671-3p) with an AUC, PPV, and NPV of 0.841, 85.71 % and 83.33 %, respectively. Our results show that circulating miRNAs (miR-26b and miR-374a, miR-28-5p, miR-489 and miR-671-3p) may serve as biomarkers for the prediction and diagnosis of grades II-IV aGVHD.

MicroRNAs as Potential Diagnostic, Prognostic, and Predictive Biomarkers for Acute Graft-versus-Host Disease

“…The diagnostic model for miR-489, miR-671-3p, and miR-28-5p was 0.841, 85.71%, and 83.33% for the area under the receiver operating characteristic curve, positive predictive value, and negative predictive value, respectively. In addition, the predictive model for miR-374a and miR-26b, which could predict an increased risk for 1 to 2 weeks before the onset of aGVHD, was 0.722, 76.19%, and 69.70% for area under the receiver operating characteristic curve, positive predictive value, and negative predictive value, respectively [60]. Thus, miRNA panels (miR-671-3p, miR-489, miR-28-5p, miR-374a, and miR-26b) may be used as diagnostic and predictive biomarkers for grade II to IV aGVHD.…”

“…Zhang et al [60] also demonstrated that miR-671-3p, miR-489, and miR-28-5p have important roles in tissue damage, tissue repair, and inflammation. Myeloid differentiation primary response gene 88 (Myd88) is an important cytosolic adaptor protein in the TLR signaling pathway, which acts as a target of miR-489 [65].…”

“…These microRNAs have important roles in tissue damage, tissue repair, and inflammation. [60] miR-29a 19 Small RNA sequencing Serum miR-29a was significantly upregulated in aGVHD onset compared with non-aGVHD. Serum miR-29a was also elevated as early as 2 weeks before time of diagnosis of aGVHD compared with timematched control subjects.…”

“…Thus, the identification of biomarkers is necessary to predict the incidence, development, and diagnosis of GVHD after transplantation. Zhang et al [60] identified a significant reduction in miR-411 at the onset of aGVHD. Moreover, with aGVHD recovery, there was no significant difference in the expression of this miRNA.…”

Dynamic changes in serum cytokine levels and their clinical significance in predicting acute GVHD

“…Ferrara et al[ 2 ] created an algorithm that included TNFR1, ST2, and Reg3α in a prognostic score for GVHD. We and others have reported clinical findings for the prediction and diagnosis of GVHD, such as the use of mass spectrometric (MS) profiling of urine[ 3 ], serum samples [ 4 , 5 ], microRNA of peripheral blood [ 6 , 7 ] and T cells [ 8 ].…”