ISHL10 Abstract T013

Plasma vesicle-associated miRNAs as therapy response biomarkers in Hodgkin Lymphoma

Cure rates for classical Hodgkin Lymphoma (CHL) patients are generally high. CHL patients that are cured remain at risk for long-term side effects from the aggressive radiation/chemotherapy regimens and relapses occur. Upcoming treatment strategies aim towards personalized therapy by early assessment of the effectiveness of active treatment by measuring vital tumor tissue with FDG-PET/CT imaging. Apart from radiation burden, imaging is costly and reported positive predictive values are suboptimal. Studies suggest that tumors actively secrete small extracellular vesicles (EVs) that encapsulate tumor-derived nucleic acids, including 22nt non-coding microRNAs (miRNAs). Here we identified with RNAseq, CHL-associated miRNAs in circulating EVs as defined biosource, using size-exclusion chromatography (SEC) as isolation method. Apart from miR-21-5p and miR-155-5p oncomiRs already implicated in lymphomagenesis, we identified let7-a-5p, miR-24-3p and miR-127-3p as circulating CHL biomarkers. Notably, a high abundance of these miRNAs in circulation of CHL patients pre-therapy, correlated with a high concentration of 100nm EVs, as determined with tunable-resistive pulse sensing (TRPS). Using logistic regression modeling we determined that EV-associated miR-127-3p levels alone can distinguish primary and relapse CHL patients from healthy controls. Importantly, the calculated AUC for miR127-3p in vesicles outperform ROC calculations using unfractionated (total) plasma. Finally, in serial samples of individual CHL patients, during- and after treatment but also in long-term follow up, we measured robust and consistent decreases in EV-miRNA levels that corresponded with FDG-PET/CT outcome . Thus defined EV-miRNAs in circulation reflect the presence of vital tumor tissue. This liquid biopsy method may proof suitable for early respsonse and relapse monitoring, MRD detection and guiding therapy decisions. Upcoming studies in serial samples are ongoing to determine whether RNAseq-generated small RNA signatures are useful for diagnosis, prognosis and prediction in CHL.