Science 23 November 2007:
Vol. 318. no. 5854, p. 1227

CLINICAL RESEARCH:
ALS Trial Raises Questions About Promising Drug
Jennifer Couzin

Bombshell. Patients on therapy declined more rapidly than those on a placebo.

An antibiotic long thought to hold promise for treating neurologic diseases failed dramatically in a recent clinical trial; this news has sparked a debate about why the drug flopped and whether the results should be interpreted broadly. In a study of 412 people with amyotrophic lateral sclerosis (ALS), published earlier this month, scientists were startled to find that patients on the drug, minocycline, declined more quickly than those on a placebo. Now clinics are urgently tracking down ALS patients whose physicians have prescribed minocycline off label. Many have been taking the drug--approved in the 1970s to treat certain infections--because animal studies have shown it can slow the course of brain diseases, and some small human trials have suggested that it's safe. That did not appear to be the case here.
The disappointing outcome may affect new or ongoing trials of minocycline for other conditions, including multiple sclerosis (MS) and Huntington's disease. Researchers leading these trials generally reject the idea that the drug's showing in ALS, published online 1 November in The Lancet Neurology, has much bearing. "We don't really believe that the ALS data adds something, but when it's out there, you have to deal with it," says neurologist Luanne Metz of the University of Calgary in Canada, who is beginning a study of minocycline in 200 people at risk for MS. She argues that MS is very different from ALS--more autoimmune than neurodegenerative--and unlikely to respond in the same way. Still, Metz's trial has added extra safety monitoring.

In contrast, the authors of The Lancet Neurology paper are quite concerned, arguing that their trial "generates the need to reexamine … the justification for other trials of minocycline in patients with neurological disorders." Over 9 months of treatment, the ALS patients on minocycline declined 25% faster, as measured by a functional rating scale, than did those taking a placebo. But there was no significant difference in death rates between the two groups, and the patients rated quality of life equally, suggesting that the decline was subtle.

Exactly what went wrong remains a mystery. Merit Cudkowicz, a neurologist at Massachusetts General Hospital in Boston who is running a minocycline trial in 100 patients with Huntington's disease, believes the drug doses in the ALS study were too high, as much as double what she and Metz are using. But Paul Gordon, who led the ALS trial and works as a neurologist at Columbia University and at Hôpital de la Pitie-Salpêtrière in Paris, notes that patients on the higher doses did no worse than those on lower ones, nor did side effects appear to correlate with deterioration. "We're sort of left with scratching our heads," he says.

The results are especially confusing because published animal studies suggested that the drug can suppress neuroinflammation and inhibit cell death. But Gordon notes that mouse studies didn't really reflect the way ALS patients are treated. For example, the mice received minocycline before they began showing symptoms.

Animal experiments offer only a rough guide to human testing, says neurologist Nigel Leigh of King's College London, who still hopes that minocycline will prove useful for ALS. Leigh says it's possible that high doses of the antibiotic were neurotoxic for patients, whereas lower doses might be neuroprotective. He'd been planning with colleagues to launch a minocycline study in 1000 ALS patients; now he's seeking approval for a revised proposal to identify an ideal minocycline dose in a smaller group of patients.

"It would be hasty to stop every trial" testing the drug, agrees John Marler, an associate director for clinical trials at the U.S. National Institute of Neurological Disorders and Stroke in Bethesda, Maryland. Even very brief trials of minocycline, though, will consider whether to modify their protocols, says Marler, such as a 3-day acute stroke trial recently funded by the institute. Adds Marler: "We're not going to overlook any possible relief " for these patients.