INTRODUCTION: Extracorporeal photoapheresis (ECP) is an immunomodulatory therapy for patients with acute or chronic graft-versus-host disease (aGVHD/cGVHD). So far, few studies have explored the molecular regulation of GVHD, and to date no studies have addressed how specific is miRNA expression change during ECP, and whether selected miRNA profiles might be predictive of clinical responses to ECP.

OBJECTIVE: The aim of the study was to analyze the expression profile of miRNAs in plasma of patients with GVHD candidates for ECP, and their changes in responding and non-responding patients to this therapy.

PATIENTS AND METHODS: Patients with GVHD underwent ECP therapy by off-line methods according to internal protocols. Peripheral blood samples were drawn pre-ECP and after 6 months of treatment. Data on patient characteristics, medical therapies and responses were obtained from medical records. We included the following study cohorts: 1) Initial cohort of 10 GVHD patients (7 cGVHD, 3 aGVHD) and 3 controls; 2) Internal validation cohort with 21 GVHD patients (14 cGVHD, 7 aGVHD) and 10 controls; and 3) External validation cohort (Policlinico S. Matteo, Pavia) composed of 24 GVHD patients (17 cGVHD, 7 aGVHD) and 12 controls. Additionally, samples from 12 patients undergoing ECP due to lung transplantation were also included. Plasma miRNAs were purified with NucleoSpin miRNA Plasma (Macherey-Nagel). In the initial cohort, we analyzed 178 miRNAs, using the Plasma focus miRNAs PCR array (Exiqon). In the validation cohorts we quantified, by qRT-PCR, candidate miRNAs using miRcury LNA RT miRNA PCR (Exiqon) and specific Exiqon primers.

RESULTS: In the initial cohort, 4 miRNAs (miR-22-5p, miR-34a-5p, miR-148a-3p, and miR-505-3p) showed higher expression in patients with GVHD compared to controls (p<0.05), which significantly decreased to values similar to healthy individuals at 6 months post-ECP (Fig. 1A). In the validation study with the groups of both the internal and external cohorts, only miR-34-5p and miR-148a-3p were significantly increased in GVHD patients before ECP treatment compared to controls, and those levels significantly decreased with this therapy. This effect was not an intrinsic change associated with ECP, since this pattern was not observed in patients undergoing ECP for lung transplantation (data not shown). We explored the ability of these 2 miRNAs to predict clinical responses to ECP at 6 months: responders showed higher pre-ECP levels of miR-34a-5p and lower miR-148a-3p compared to non-responders (p=0.009, and p=0.003, respectively) (Fig. 1B). The ROC curves showed an area under the curve (AUC) of 0.82 for miR-34a-5p and 0.89 for miR-148a-3p. The ability to identify patients who respond to ECP was improved (p<0.01) by combining both markers (AUC=0.96) (Fig. 1C).

CONCLUSION: This study identifies miR-34a-5p and miR-148a-3p as potential biomarkers to predict responses to ECP in patients with GVHD. The combinatorial function of these miRNAs can provide a more informative outlook on the pathophysiology of the disease by identifying potential target genes being inhibited and the pathways involved, both in the development of GVHD and following responses to ECP. Overall, this information would enable the implementation of more personalized patient treatment strategies, and likely lead to significant advances in the management of GVHD patients.


Lozano:Amgen: Consultancy; Terumo S.A.: Consultancy; Grifols S.A.: Consultancy; Novartis: Consultancy; Macopharma: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution