Expression of Serum Micrornas Are Dysregulated during Acute Graft Versus Host Disease

Introduction

MicroRNAs are expressed in body fluids and have recently been associated with the etiology of acute graft versus host disease (GvHD), but global expression profiling in a haematopoietic stem cell transplant (HSCT) setting is lacking. An improved understanding of the molecular pathogenesis of aGvHD may allow for improved therapeutic options, or guide personalised prophylactic protocols.

Methods

Mature microRNA serum expression (n=799) was assessed using nCounter technology (NanoString) in a training cohort of diagnostic aGvHD samples (n=12). Analysis assessed for microRNAs that were dysregulated in patients who developed aGvHD compared to no aGvHD. Signature microRNAs were validated in an independent cohort of serum samples taken at aGvHD diagnosis (n=42) and prior to disease onset (day 14 (D14) post-HSCT, n=47).

Results

NanoString profiling identified 61 microRNAs that were differentially expressed at aGvHD onset, of which n=27 were downregulated (fold change (FC) ^-6.94 - ^-1.75; p<0.01 - p=0.048) in aGvHD while n=34 were upregulated (FC 1.35 - 5.41; p<0.01 - p=0.046). 10 microRNAs (miR-146a, miR-30b, miR-374, miR-20a, miR-15a, miR-181a, miR-18a, miR-19a, miR-19b and miR-451a) were selected for further assessment in an independent diagnostic cohort (n=42), based on high FC or those reported in the literature to be implicated in GvHD, T-cell function or the inflammatory response. MiR-146a (p=0.03), miR-30b-5p (p=0.007), miR-374-5p (p=0.02) and miR-181a (p=0.03) were significantly downregulated, whilst miR-20a (p=0.03) and miR-15a (p=0.03) were significantly upregulated in aGvHD. MiR-30b (AUC=0.75, p=0.007), miR-374-5p (AUC=0.74, p=0.01) and miR-15a (AUC=0.70, p=0.04) had diagnostic ability for aGvHD as assessed by receiver operator characteristic (ROC) analysis, whilst miR-181 (AUC=0.68, p=0.06), miR-146a (AUC=0.66, p=0.09) and miR-20a (AUC=0.68, p=0.06) were approaching significance. There was no significant difference in microRNA expression between clinical factors including transplant relationship, patient gender or conditioning regimen, indicating the effect to be related to aGvHD incidence.

MicroRNA expression was also assessed at D14 post-HSCT in an independent cohort (n=47) to investigate their prognostic marker potential. MiR-146a (p=0.01), miR-20a (p=0.03), miR-18a (p=0.03), miR-19a (p=0.03), miR-19b (p=0.02) and miR-451 (p=0.01) were expressed at significantly higher levels in patients who developed aGvHD vs. no aGvHD. In ROC analysis, miR-146a (A=0.68, p=0.03), miR-19b (A=0.70, p=0.02) and miR-451 (A=0.69, p=0.03) had diagnostic ability with regards to aGvHD incidence, whilst miR-18a (A=0.65, p=0.09), miR-19a (A=0.65, p=0.08) and miR-20a (A=0.67, p=0.06) were approaching significance.

Conclusions

Circulating microRNAs have the capacity to act as prognostic and diagnostic biomarkers for aGvHD and might assist in developing personalised therapeutic approaches. Their dysregulated expression suggests a role in aGvHD pathology, which warrants further investigation.