Joint- and Muscle-Dominant Sarcoidosis: Choosing the First Steroid-Sparing Agent
A single patient whose sarcoidosis has never really been a lung disease. The disagreement isn’t about whether she needs a steroid-sparing agent — it’s about which one actually has evidence behind the organ system doing the damage.
Renata M., a 52-year-old woman, shares a two-bedroom apartment with her twenty-year-old daughter, home for the summer between college semesters and, this year, doing most of the grocery-carrying and stair-climbing her mother can no longer manage without stopping halfway up. Three years ago a routine pre-employment chest X-ray incidentally found bilateral hilar lymphadenopathy; a mediastinal lymph node biopsy confirmed noncaseating granulomas, and she was staged as pulmonary sarcoidosis, Stage I. Her lungs, though, have never really been the problem — her most recent pulmonary function testing put her FVC at 94% of predicted, essentially unchanged since diagnosis. What has actually shaped the last two years is a symmetric ankle-and-knee polyarthritis and a proximal myopathy severe enough that she braces against the countertop to stand from a kitchen chair, both of which reliably worsen whenever her prednisone dose drops below 15mg.
She has been on some dose of prednisone continuously since diagnosis, unable to taper past that floor without a flare within two weeks. Her serum ACE level, drawn last month, runs at 2.4 times the upper limit of normal — nonspecific on its own, but consistent with the granuloma burden actually driving her joint and muscle disease rather than any change in her chest imaging, which is unchanged from three years ago. Her hands, at least, have been spared — the polyarthritis has stayed confined to her ankles and knees, though it is enough on its own to make the stairs to her daughter's old bedroom a genuine daily obstacle rather than a minor inconvenience. The cost of that prednisone floor is now showing up in her own labs: a DEXA six months ago found early osteopenia in her lumbar spine, and her fasting glucose has climbed from 92 to 118 over the same interval. She needs an agent that can actually replace the prednisone she’s stuck on, not just supplement it, and needs it to work against joint and muscle disease specifically — a different target than the pulmonary fibrosis most sarcoidosis drug trials are actually built to measure.
Deciding what replaces the prednisone
Methotrexate should be the first steroid-sparing agent here. The 2021 European Respiratory Society treatment guideline puts methotrexate ahead of biologic therapy for chronic sarcoidosis needing a second agent — conditionally, on very low quality evidence, which is frankly the ceiling in a disease with this little randomized data — and retrospective series specifically report real response in articular and myopathic sarcoidosis, not just pulmonary disease. It’s a familiar, monitorable toxicity profile — CBC and liver enzymes every few months, folic acid alongside it — and it’s the guideline’s own default position for exactly this situation.
You’re right that methotrexate is the guideline’s stated default — but two years of an unsuccessful steroid taper is its own kind of evidence, and infliximab is the only sarcoidosis therapy with real randomized controlled trial evidence behind it rather than retrospective case series. Baughman et al., 2006, ran a placebo-controlled trial of infliximab in chronic pulmonary sarcoidosis and found a statistically significant improvement in percent-predicted forced vital capacity at 24 weeks — 2.5 points over placebo — which is modest, but it is more than methotrexate has ever been asked to demonstrate in a trial of that quality.
I’d rather start with the best-evidenced agent available in a disease this thin on RCTs than work through a lower-evidence-tier drug first and lose more time to a floor she’s already been stuck on for two years.
The RCT argument only holds if the trial’s population actually looks like her, and it doesn’t.
Baughman’s trial required an FVC between 50 and 85% of predicted to enroll at all, and the mean it actually enrolled was 69%. Nor were those patients refractory — they had chronic, stable disease on stable background steroids or immunosuppressants; failing prior therapy was never an entry requirement. It was a population defined by pulmonary severity. Her FVC is 94% predicted and has never moved: she sits above that trial’s own ceiling, not inside it. And the result you’re leaning on was 2.5 points of predicted FVC, with no significant separation from placebo on any of the major patient-oriented endpoints at 24 weeks — not quality of life, not dyspnea score, not six-minute walk distance. That trial doesn’t describe her disease; it describes a different organ system at a different stage, and it moved a number there rather than a symptom. Methotrexate is the correct first agent here not just because the guideline says so, but because it’s the drug with actual response data in the organ system that’s driving her disability. Infliximab also carries real infection and latent-TB-reactivation risk that shouldn’t be taken on for a mismatch. Start methotrexate; hold infliximab in reserve if her joints and muscles don’t respond in three to six months, or if her lungs ever drop into the 50-to-85% window where Baughman’s patients actually lived.
Agreed: start methotrexate now, continue prednisone at its current floor as a bridge until methotrexate has had time to work, and repeat serum ACE and ESR alongside a joint and strength exam at three months to judge response before deciding anything further.
Not agreed, and left as a real open branch rather than smoothed over:
Send latent tuberculosis screening now, so there’s no delay if infliximab becomes necessary later.
Hold that screen until methotrexate’s response is known, rather than ordering a test whose result may be stale by the time, if ever, it’s actually acted on.