Low-Dose Interleukin-2 for Cancer-Associated Myositis: a Prospective Multicentre Proof-Of-Concept Cohort Study

Abstract

Background: Cancer-associated myositis (CAM) is associated with substantial morbidity and mortality, and treatment is complicated by concern that non-selective immunosuppression could impair antitumour immunity. Low-dose interleukin-2 (Ld-IL2) selectively expands regulatory T cells. We assessed the clinical activity, immunological effects, and short-term oncological safety of Ld-IL2 in patients with CAM. Methods: In this prospective, multicentre, proof-of-concept cohort study, adults with CAM and active myositis despite standard treatment, or for whom further immunosuppressive escalation was considered unsuitable, received subcutaneous Ld-IL2 (1 million IU every other day) in addition to stable background treatment for 24 weeks at four tertiary centres in China. The primary outcome was clinical improvement at week 24 according to the 2016 American College of Rheumatology–European Alliance of Associations for Rheumatology Total Improvement Score (TIS). Clinical, immunological, oncological, and safety outcomes were assessed in the treated cohort. An exploratory comparison of overall survival was made with a non-randomised, contemporaneous historical cohort receiving conventional treatment. Findings: Between Sep 1, 2021 and Jan 30, 2026, 101 patients were screened across four participating centres and 48 were included: 17 in the prospective Ld-IL2 cohort and 31 in the historical-control cohort. All 17 treated patients completed 24 weeks of follow-up. At week 24, 15 (88%) of 17 patients had at least minimal improvement (TIS ≥20), including five (29%) with moderate improvement and four (24%) with major improvement. Median Manual Muscle Testing-8 (MMT-8) score increased from 134·0 (IQR 128·0–142·0) to 144·0 (138·0–144·0; p Interpretation: Add-on Ld-IL2 was associated with improvement across several myositis domains, glucocorticoid sparing, and expansion of circulating regulatory T cells over 24 weeks, without an observed short-term signal of tumour progression. The small uncontrolled cohort, heterogeneous cancers and anticancer treatments, and non-randomised historical comparison preclude conclusions about comparative efficacy, long-term oncological safety, or survival. Controlled studies with prespecified oncological outcomes and longer follow-up are warranted.