Inclusion Body Myositis: A Bounded Drug Trial Against Supportive Care Alone
A single patient with three years of progressive, biopsy-confirmed inclusion body myositis. The disagreement is whether an individual trial of immunosuppression is worth its own real costs given the weight of negative group-level evidence.
George K., 71, has spent most of his free time at a woodworking bench in his garage for thirty years, and the first thing he actually noticed, three years before this visit, wasn't his legs — it was dropping chisels he'd held steady his whole life. The weakness has since spread and worsened: pronounced quadriceps weakness now causing two falls in the past six months, early difficulty swallowing solids, and the classic asymmetric pattern — quadriceps and long finger flexors disproportionately affected — that led to a muscle biopsy showing rimmed vacuoles and endomysial inflammation. He is anti-cN1A antibody positive, consistent with sporadic inclusion body myositis, with a CK only mildly elevated in the 600s, well below the levels typical of an acute necrotizing process.
The therapeutic literature here is unusually one-sided for an inflammatory myopathy. RESILIENT, the largest randomized trial of a targeted biologic in IBM, tested bimagrumab, a human monoclonal antibody that blocks activin type II receptors — binding them more avidly than activin and myostatin themselves — and produced a genuine, measurable increase in thigh muscle volume and lean body mass. It still failed its primary functional endpoint, six-minute walk distance at 52 weeks, and development of the drug for this indication has since been discontinued. That result matters beyond bimagrumab itself: it is the clearest evidence that an intervention doing something real and measurable at the muscle-biology level in IBM can still fail to translate into functional benefit, which is exactly the risk a trial of prednisone would run, in a disease where steroid myopathy could compound the very weakness being treated. Against that backdrop, the one intervention with a positive controlled signal in IBM is not a drug at all. A Danish randomized trial of blood-flow-restricted resistance training, twice weekly for twelve weeks against a non-exercising control group, likewise failed to show improvement in measured or self-reported function — but knee-extensor strength held steady in the exercising group, up 5.8%, while the controls lost 9.2% over the same three months. That is a maintenance claim rather than an improvement claim, and it is the most any intervention in this disease has honestly earned.
In clinic, three years in and two falls this year
I understand IBM overall behaves very differently from polymyositis or dermatomyositis, and I'm not arguing otherwise. But he's three years in and still ambulatory — earlier in the disease course than a lot of trial populations, RESILIENT's included. A bounded trial of prednisone or IVIG, three months, with an explicit stop rule if there's no measurable functional gain, is a low-cost way to find out if he's an individual responder.
An individual trial isn't unreasonable in principle. But RESILIENT is worth naming specifically: bimagrumab produced a real, measurable increase in thigh muscle volume and still failed its primary functional endpoint, six-minute walk distance. That's the clearest data point we have showing an intervention that clearly does something biologically in IBM can still fail to help function at all. And immunosuppression itself has already been tried under randomization: Badrising's 48-week trial of oral methotrexate in 44 IBM patients lowered creatine kinase and left muscle strength on myometry no better than placebo.
Prednisone risks the same trap from a worse angle — steroid myopathy can directly worsen the exact weakness we're trying to treat, in a 71-year-old who has already fallen twice this year and whose bone health a steroid course would put at further risk. That's a real, quantifiable cost, not a hypothetical one.
I'd rather not frame this as drug versus nothing. Structured resistance exercise has the one IBM-specific controlled signal worth having: in the Danish blood-flow-restriction trial, knee-extensor strength was essentially unchanged over twelve weeks in the exercising arm — up 5.8% — while the non-exercising controls lost 9.2%. I want to be precise that this is maintenance, not improvement; that trial missed its functional endpoints too. But it is genuine support, and it carries none of prednisone's toxicity.
Whatever the group decides about a drug trial, his two falls this year and his new dysphagia to solids need direct attention now — formal gait and fall-prevention evaluation, and a swallowing assessment given IBM's known aspiration risk. That's the evidence-based plan available today, not something to hold while we debate a drug.
Agreed: a bounded twelve-week prednisone trial, with the explicit stop rule the rheumatologist required, run in parallel with starting supervised resistance and endurance exercise and a formal swallowing evaluation — both begun regardless of the drug trial's outcome.
Not agreed: the neuromuscular specialist remains more optimistic that an individual response is plausible than the rheumatologist, who agreed to the trial only because it is tightly bounded and would have preferred moving straight to the non-pharmacologic plan alone.