The FSHD muscle–blood biomarker: a circulating transcriptomic biomarker for clinical severity in facioscapulohumeral muscular dystrophy

Abstract

AbstractFacioscapulohumeral muscular dystrophy (FSHD) is a prevalent, incurable skeletal myopathy. Clinical trials for FSHD are hindered by heterogeneous biomarkers poorly associated with clinical severity, requiring invasive muscle biopsy. Macroscopically, FSHD presents with slow fatty replacement of muscle, rapidly accelerated by inflammation. Mis-expression of the transcription factor DUX4 is currently accepted to underlie pathogenesis, and mechanisms including PAX7 target gene repression have been proposed. Here, we performed RNA-sequencing on MRI-guided inflamed and isogenic non-inflamed muscle biopsies from the same clinically characterized FSHD patients (n = 24), alongside isogenic peripheral blood mononucleated cells from a subset of patients (n = 13) and unaffected controls (n = 11). Multivariate models were employed to evaluate the clinical associations of five published FSHD transcriptomic biomarkers. We demonstrated that PAX7 target gene repression can discriminate control, inflamed and non-inflamed FSHD muscle independently of age and sex (P < 0.013), while the discriminatory power of DUX4 target genes was limited to distinguishing FSHD muscle from control. Importantly, the level of PAX7 target gene repression in non-inflamed muscle associated with clinical assessments of FSHD severity (P = 0.04). DUX4 target gene biomarkers in FSHD muscle showed associations with lower limb fat fraction and D4Z4 array length but not clinical assessment. Lastly, PAX7 target gene repression in FSHD muscle correlated with the level in isogenic peripheral blood mononucleated cells (P = 0.002). A refined PAX7 target gene biomarker comprising 143/601 PAX7 target genes computed in peripheral blood (the FSHD muscle–blood biomarker) associated with clinical severity in FSHD patients (P < 0.036). Our new circulating biomarker validates as a classifier of clinical severity in an independent data set of 54 FSHD patient and 29 matched control blood samples, with improved power in older patients (P = 0.03). In summary, we present the minimally invasive FSHD muscle–blood biomarker of FSHD clinical severity valid in patient muscle and blood, of potential use in routine disease monitoring and clinical trials.