A withdrawal design not fully realized
Based on the title of Schwarz Pharma’s February 15th press release: “Positive Results with Lacosamide in Diabetic Neuropatic Pain in Phase III trial” I was expecting to read of a favorable result of a typical double-blind, placebo-controlled trial of an antiepileptic drug used in patients with painful diabetic neuropathy. I was pleasantly surprised to see the subtitle: “Preliminary results from a phase III withdrawal-design trial with lacosamide in patients with painful diabetic neuropathy show that lacosamide was statistically significant over placebo in reducing pain scores. SCHWARZ PHARMA plans for regulatory submission during the second half of 2007.”
“Cool,” I thought, ”A little creativity used to test a notoriously stubborn subjective endpoint.” Here’s the release’s capsule summary:
106 patients with painful diabetic neuropathy were treated in this multi-center, double-blind, placebo-controlled phase III study. All patients began the trial receiving their optimal dose of lacosamide (< 400mg/day). During the 16 weeks treatment duration, lacosamide was withdrawn and re-introduced in a blinded fashion at pre-defined time points. Overall, patients showed consistently higher average pain scores while taking placebo than when re-exposed to lacosamide.
The potential benefits of withdrawal designs are several-fold and include the following: They allow all study subjects to receive a promising new treatment for at least part of the study period. They can be used to identify and enrich the re-treatment period with responders to reduce the likelihood of futile long-term treatment. When endpoints are reversible and tolerable, they can provide a robust test of cause and effect in individuals.
The lacosamide trial offered all subjects the opportunity to receive lacosamide. That’s it.
Subjects were all given lacosamide prior to withdrawal; there’s was no reported attempt to distinguish responders from non-responders prior to or during withdrawal. Therefore, there was no no enrichment in responders in the re-treatment phase, a lost opportunity to increase efficiency in that phase. Also, since withdrawal itself was not randomized (i.e. genuine vs. mock withdrawal), there was a missed opportunity to perform controlled tests of the withdrawal/re-introduction of lacosamide. It’s a high hurdle to pass, for sure, but it’s a very convincing case for the clinical use of lacosamide over other drugs (like Lyrica or Cymbalta, for example) among patients with characteristics similar to those of the lacosamide-responder study population who show a subjective response to the drug in the “real world”. The higher cost of such a randomized withdrawal design can be offset by enrichment in the re-treatment phase.
As it turned out, then, this was just another Phase 3, double-blind, placebo-controlled study of an antiepileptic in diabetic neuropathy. But at least it looks like this antiepileptic from Schwarz Pharma has a decent chance to see the light of day.
