Objective To perform a mechanistic study on the effect of interleukin (IL)-1 blockade by anakinra in patients with refractory myositis and to explore possible predictive biomarkers.

Methods Fifteen patients with refractory myositis were treated with anakinra for 12 months. Clinical response was assessed by the six-item core set measures of disease activity International Myositis Assessment and Clinical Studies (IMACS) and functional index (FI). Repeated muscle biopsies were investigated for cellular infiltrates, IL-1α, IL-1β, IL-1Ra and major histocompatibility complex-class I by immunohistochemistry. Serum levels of IL-1Ra and granulocyte colony-stimulating factor (G-CSF) were measured by ELISA. T cell phenotype and functional assays were investigated by multicolour flow cytometry.

Results Seven patients had clinical response according to IMACS, four of them also showed improved FI. Responders had higher baseline extramuscular score compared with non-responders. In muscle biopsies, baseline CD163 macrophages and IL-1α expression were inversely correlated with muscle performance after 6 months treatment; all responders had IL-1Ra expression in the post-treatment biopsies but only 3/8 non-responders. In serum, IL-1Ra levels were increased and G-CSF was decreased after 6 months treatment, but their levels and changes were not related to clinical response. For T cells, an inverse correlation between baseline frequency of CD4 activated/memory T cells and decreased creatine kinase levels was observed. Five of six patients demonstrated less IL-17A and more IFN-γ secreting CD4 T cells after 6 months treatment. Moreover, anakinra reduced IL-17A secretion in vitro.

Conclusions Patients with myositis may respond to anakinra. Extramuscular score, muscle CD163 macrophages and IL-1α expression, blood CD4 activated/memory T cells might associate with anakinra treatment response. Blocking the IL-1 receptor disfavoured Th17 cell differentiation both in vivo and in vitro.