What is mycophenolate

The short version: mycophenolate is a second-line immunosuppressant for myasthenia gravis, but the evidence that it works better than azathioprine or placebo is thin and comes from conflicted studies.

Mycophenolate mofetil (CellCept) blocks immune cell proliferation. It is used off-label for myasthenia gravis when first-line drugs like prednisone or azathioprine fail or cause side effects. The mechanism is straightforward: it suppresses the antibody-producing B cells and T cells that attack the neuromuscular junction.

The evidence base is weak. The two largest randomized trials, both funded by the manufacturer, failed to show a statistically significant benefit over placebo when added to prednisone. One trial enrolled 80 patients and found no difference in the primary outcome at 12 weeks. The other enrolled 176 patients and found no difference at 36 weeks. Both used a surrogate endpoint (a quantitative myasthenia gravis score) rather than a hard clinical outcome like hospitalization or need for rescue therapy. Neither trial compared mycophenolate directly to azathioprine. The manufacturer funded both.

Outside these trials, the only evidence is observational. A few small retrospective series suggest mycophenolate may reduce steroid dose over 12 to 24 months, but these studies have no control group, no blinding, and no randomization. The long-term safety data is also thin: mycophenolate carries known risks of infection, lymphoma, and birth defects, but no trial has tracked these against azathioprine over years in MG patients.

For a 56-year-old man who has been on azathioprine for 18 months and now has new weakness, switching to mycophenolate is a reasonable clinical option, but no study has tested whether it works better than staying on azathioprine, raising the dose, or restarting IVIg. The decision rests on clinical judgment, not on published evidence.

My call: mycophenolate is a plausible next drug, but the evidence does not show it is superior to azathioprine or placebo for this situation. Confidence: low, because the only randomized trials failed to show benefit and the manufacturer funded them.