Circulating microRNAs as non-invasive biomarkers for canine Cushing’s syndrome

Canine Cushing's syndrome (hypercortisolism) can be caused by a pituitary tumor (pituitary-dependent hypercortisolism; PDH) or a cortisol-secreting adrenocortical tumor (csACT). For both cases, non-invasive biomarkers that could pre-operatively predict the risk of recurrence after surgery would greatly impact clinical decision making. The aim of this study was to determine whether circulating microRNAs (miRNAs) can be used as diagnostic (presence of PDH or csACT) and/or prognostic (disease recurrence, histological grade) non-invasive biomarkers for canine Cushing's syndrome. After a pilot study with 40 miRNAs in blood samples of healthy dogs (n = 3), dogs with PDH (n = 3) and dogs with a csACT (n = 4), we selected a total of 20 miRNAs for the definitive study. In the definitive study, these 20 miRNAs were analyzed in blood samples of healthy dogs (n = 6), dogs with PDH (n = 19, pre- and post-operative samples) and dogs with a csACT (n = 26, pre-operative samples). In dogs with PDH, six miRNAs (miR-122-5p, miR-126-5p, miR-141-3p, miR-222-3p, miR-375-3p and miR-483-3p) were differentially expressed compared to healthy dogs. Of one miRNA, miR-122-5p, the expression levels did not overlap between healthy dogs and dogs with PDH (p = 2.9x10(-4)), significantly decreased after hypophysectomy (p = 0.013), and were significantly higher (p = 0.017) in dogs with recurrence (n = 3) than in dogs without recurrence for at least one year after hypophysectomy (n = 7). In dogs with csACTs, two miRNAs (miR-483-3p and miR-223-3p) were differentially expressed compared to healthy dogs. Additionally, miR-141-3p was expressed significantly lower (p = 0.009) in dogs with csACTs that had a histopathological Utrecht score of >= 11 compared to those with a score of