In Reply

Abstract

In Reply: In their reply, Drs Walters and Hadley make points that can be questioned just by quoting the published text of our correspondence with them. They credit themselves with detective work in finding that the US Food and Drug Administration (FDA) has designated GM-1 as an orphan drug in acute spinal cord injury (SCI) but has not approved it. No detective work was required. This was exactly what our letter stated: “The actual approval status is that the US FDA on December 3, 2012, designated GM-1 as an orphan drug in acute SCI treatment, and GM-1 is now in the US FDA IDE NDA (New Drug Application) and approval process.” We clearly said that the NDA and approval were “in process.” They go on to describe us as “apparently suggesting” that GM-1 be used off-label. Reading of our brief letter shows no such suggestion. We actually did not advocate that it be used at all. What we repeatedly argued was that it should be considered a “treatment option” in a scientific sense. The fact that GM-1 is not yet an approved drug for acute SCI in the United States does not affect this contention; when the Guidelines Committee issued their 2002 Guidelines, they too described GM-1 as a treatment option without recognizing any obligation themselves to provide the drug, then unavailable in the United States. As we noted, the process toward regulatory approval is being pursued. Aside from these inaccuracies, the basic argument that Drs Walters and Hadley present is puzzling. A key premise is their statement: “When the previous guidelines were published, their final follow-up study of patients in their last GM-1 trial was just available for inclusion.” So the follow-up was “available,” but only “just available.” They possibly might have wished to be more diligent in rapidly evaluating those data, or they might have thought it prudent to wait to publish their conclusions until they could be as thorough as they wanted to be. They say now that they chose then to publish an interim conclusion—without having warned their readers then. This time they wish us to understand that they have been more careful. They state “We were determined, in this iteration, to be as true as possible to the evidence while removing such perceived contradictions.” Were they not previously determined to be true to the evidence? In both the 2002 and the 2013 Guidelines, the committee describes our papers from both the single-center and the multicenter GM-1 studies as “Evidence Class I.” In their latest reply, Drs Hadley and Walters go to some length and effort to show that authorities then and since consider that they themselves did not have adequate evidence of their conclusions. They claim a determination now to be more careful in the new Guidelines. Aside from paradoxically working to undermine their own report, their deference to the authorities that they cite is tacitly selective. They quote the Cochrane Review on not recommending GM-1, but their 2013 Guidelines mention and then freely disregard the same Cochrane article's approbation for methylprednisolone. They quote Dr Haines out of context to suggest disapproval of GM-1 treatment and a need for substantial revision. In fact, he asked for a change in wording: “It would seem best to couch options in terms such as ‘may,’ ‘suggest,’ and ‘consider.’” What Dr Tator described as “not particularly helpful” was in fact not the listing of GM-1 as a treatment option, but the tendency of the 2002 Guidelines to “merely reiterate” our published reports. In any case, if Drs Hadley and Walters thought that the comments published with the 2002 Guidelines invalidated their conclusions, then why did they not revise their recommendations at the time? The evidence that Drs Walters and Hadley ignore but that is mentioned in our reply is this: Two trials were carried out under FDA IDE supervision. The efficacy of GM-1 in acute SCI, started 36 to 72 hours post-injury and continued for 30 or 60 days during the recovery phase, was demonstrated in a pivotal trial, and results including earlier recovery for incomplete injuries were confirmed in a multicenter trial. This may not prove the unqualified efficacy of GM-1 in the accepted scientific sense, and we have never anywhere argued that it does. In the absence of a safety concern and in an orphan drug situation in which no other pharmaceutical treatment is approved, it does show that use of GM-1 in acute, incomplete SCI, if eventually approved by the FDA, could be rational and plausible: in other words, that it can be regarded as an option. Disclosure The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.