| Objective:Acute graft versus host disease (aGVHD) is one of the complications after allogeneic hematopoietic stem cell transplantation (allo-HSCT), with an incidence of10to50%. The fatality rate of aGVHD is as high as75%, seriously affecting the effective of transplantation and postoperative survival rate. Currently, the diagnosis of acute graft-versus-host disease (aGVHD) is mainly based on clinical symptoms and biopsy results. This study is to explore the clinical significance of serum biomarkers in aGVHD and to find a validated diagnostic or predictive model for aGVHD by serum cytokines and biochemical tests, microRNAs and peptides.Methods:The sera and clinical parameter of aGVHD (n=20) and non-GVHD (n=25) were collected for clinical data set. In first part, we detectedthe concentrations of eight cytokines (IL-8, IL-10, sCD40L, TNF-α, MCP-1, MIP-1α, RANTES and LIGHT) and nineteenconventional biochemical indicatorsin the sera of aGVHD and non-GVHD patients through the whole process of allogeneic hematopoietic stem cell transplantation (allo-HSCT) and onset of aGVHD. In second part, we profiled miRNAs in serum pools from patients with aGVHD (grades Ⅱ-Ⅳ) and non-aGVHD controls by real-time qPCR (quantitative PCR)-based TaqMan MicroRNA arrays and3miRNAs which could predict the happening of aGVHD had had been validated by QPCR. In the third part, we compared the difference in polypeptide expression before and after transplantation by using CLINPROT technology, and attempted to predict aGVHD by serum protein and polypeptide profiling. In the last part, we tried to establish predition model for aGVHD combining with the four kinds of serum parameters.Results:Our data in first part showed serum cytokines and biochemical indicators are significantly different betweenaGVHD patients and non-GVHD group (two-tail p<0.05), either before the transplantation practice or before the onset of aGVHD. Our models established by binary logistic regression at the day+7and+14showed a significant absolute capacity of predicting grade Ⅱ～Ⅳ aGVHD, with a positive and negative predictive value of at least90%. In second part, we found that miR-411were significantly down-regulated when aGVHD attacked comparing with the day+7after HSCT. Through screening by miRNA array in sera of discovery cohort (n=18) and validating in sera of validation cohort (n=38), we have developed a predictive model including2miRNAs (miR-26b and miR-374a) that can predict the probability of aGVHD with AUC of0.931, and a diagnostic model including3miRNAs (miR-28-5p, miR-489and miR-671-3p) that had an AUC of0.961. In the third part, the data of mass spectrum showed serum peptides polifing between aGVHD and non-GVHD was different. The accuracy of predictive models for aGVHD(Ⅰ～Ⅳ) using MS spectral data before and after transplant were77.8%and90.6%, respectively. Meanwhile, the accuracy of predictive models for aGVHD(Ⅱ～Ⅳ) before and after transplant were increased to88.57%and92.13%, respectively. In the last part, we combined the four kinds of serum parameters chose before and established predicitve model for aGVHD at the day+7. Two parameters (miR-26b, peptide mass2866Da) entered the model and improved the predictive value of model, with a sensitivity, specificity and accuracy of94.70%,89.50%amd92.10%, respectively.Conclusions:Our study showed the levels of serum parameters, including biochemical indicators, cytokines, microRNAs and peptides, would be affected by initial aGVHD, making it possible to pre warn aGVHD before clinical symptom appeared. Our models combined with multi biomarkers would predict grade Ⅱ-Ⅳ aGVHD outbreak one or two weeks ahead, in which the positive and negative predictive values are all exceed90%. Meanwhile, the analysis of aGVHD associated serum biomarkers provide for a noninvasive, fast and accurate method for clinical diagnosing aGVHD. |
PDF Full Text Request