Validation of miR-181A-5p as a biomarker of acute rejection after heart transplantation. Multicenter study

Abstract

Abstract Background Acute Cellular Rejection (ACR) remains a major complication in heart transplantation (HT) being endomyocardial biopsy (EMB) the gold standard for ACR early detection. However, this invasive procedure has several limitations so last years new non-invasive ACR biomarkers have been evaluated. Previous studies have shown miR-181a-5p (miR-181) is overexpressed in serum samples from patients with moderate/severe ACR. However, techniques performed in these studies (mainly RT-qPCR) express miRs results as relative quantification and this leads to a lack of harmonization of both the technique and clinical decision-making. This would be solved with a new technique for absolute quantification of miRs, the digital PCR (dPCR). The aim of this study is the validation of miR-181a-5p as ACR biomarker in a multicentric collaboration. The results obtained by the conventional RT-qPCR technique will also be compared with those obtained by the new and promising dPCR. Methods four reference centers in HT from Spain were included in the study named "A,B,C, D". All centers performed a retrospective search in their serum libraries to find sera from patients with ACR≥2R. The same number of ACR=0R samples (control group) were also sent. All samples were processed at one center by RT-qPCR and dPCR. With the collected data we studied the correlation between both techniques, the expression of miR-181 in each ACR group (2R vs 0R) as well as in each center. Finally, the diagnostic accuracy was studied using ROC curves. A two-sided p<0.05 was considered statistically significant. Results ACR≥2R samples involved in the study were from: Center A, 41(42%), Center B, 32 (33%), Center C, 18(19%) and Center D 6 (6%). All centers sent the same number of samples with no rejection 0R. Correlation obtained between both techniques was not good (Spearman correlation coeficient, r = 0,60; [0,50-0,70]; p<0,0001). In both techniques, miR-181 is overexpressed in ACR≥2R group, although only RT-qPCR performed it significantly (p<0,0005). There was no significant difference between centers in miR-181 expression. Diagnostic accuracy achieved by RT-qPCR was similar to that of previous studies (AUC=0,65; [0,57-0,73], p=0,0006). Conclusion Our findings confirm previous results obtained by miR-181a-5p as ACR biomarker using conventional techniques as RT-qPCR. In addition, robustness of miR-181a-5p has been shown using serum samples from 4 different centers. Digital PCR shows promising results but we still optimize it for the quantification of miRs.miR181 expression