I am a little peeved at a new attack on the push for FDA’s Accelerated Approval Program for the ALS treatment called GM6 (known as GM604 in clinical trials). This attack basically follows the same line as the one from the ALS Association (and in fact incorporated it by reference). Both are sloppy and disingenuous in style and content. They attempt false comparison to other previously-failed trials without including the reasons for those previous failures and they also try a very clumsy smear by comparison to an act of intentional clinical fraud. I am disappointed that such comparisons are presented to a public looking for facts as well as hope.
First let me discuss the false comparison to previous trial:
- The comparisons are false primarily because the old trials exclusively used the ALS Functional Rating Scale (ALSFRS) which is a horrible metric by even the most generous interpretation. This is the only way a call for large trials can be justified, because the numb insensitivity of the ALSFRS (even the Revised version) requires such to make any kind of meaningful conclusion in a heterogeneous disease like ALS. However, the GM6 trial used several biomarker candidates and other objective clinical measurements as surrogate end-points (keep that phrase in mind) which correlated which the subjective end-points such as the ALSFRS-R. The biomarkers used were suggested and evaluated by Dr. Robert Bowser, a 2015 winner of the Sheila Essey Award for significant research contributions to fighting ALS.
- Brain-Derived Neurotrophic Factor (BDNF), one of the neurotrophic factors listed in Dr. Dickie’s post, doesn’t cross the Blood Brain Barrier (BBB). It’s almost impossible to get a therapeutic dose into the patient without an intrathecal infusion (directly into spine) using a pump over time. This has numerous obvious drawbacks. It’s also very unclear whether a single neurotrophic factor is useful in ALS, which encompasses a host of deficiencies.
- Cilliary Neurotrophic Factor (CNTF) does cross the BBB and early animal model tests indicated efficacy. However, two human trials in 1996 using subcutaneous delivery and intrathecal delivery as well as a review in 2004 revealed no efficacy in lower doses and serious side-effects at high doses. It’s also important to note that the animal data came well before the excellent work ALSTDI did characterizing the extreme difficulties in using that model. Any ALS mouse data released prior to 2009 (and any subsequent found to not strictly follow those guidelines) should be considered suspect. I have personal knowledge of the difficulty in using this model and the false-positive data which can result from improper use.
- Next, Insulin-like Growth Factor 1 (IGF-1) also crosses the BBB but has a very very short biological half-life, meaning it is broken down and excreted in a matter of hours. That makes therapeutic levels almost impossible to maintain. A form was created with a buffering agent attached to IGF-1 which roughly doubled the half-life but even that was woefully inadequate. Anyone who remembers the IPLEX debacle of a few years ago knows the story.
The comparisons to such single-target neurotrophic factors as BDNF, CNTF, and IGF-1 are therefore flawed in logic and fact. It is very disingenuous for Dr. Dickie to compare GM6 to them as GM6 is a master regulator and acts in 12 relevant pathways simultaneously. This information is already in the public domain freely available for anyone to look up.
Next, Dr. Dickie compares the GM6 results with those of the initial results of NP001 (actually he links to his own blog post where he addresses the anecdotal reports which came before the official results were published). What he failed to mention was the updated post-hoc analysis which showed a halt of disease progression in 27% of patients in the trial. Further, the analysis showed statistically significant evidence of two biomarkers which identified responding sub-groups. This is a tremendous achievement in ALS clinical trial history. Unfortunately the biomarkers aren’t the same in the GM6 trial so the comparison of the two is incomplete at best. The only real similarity is that both trials used biomarkers as secondary end-points. However, the GM6 trial used them in a way that didn’t require post-hoc analysis.
The comparison with lithium is especially troubling. First, it was far from “recent”, with patient excitement starting in 2007. You can find the data collected in the first PALS-led and created clinical trial which coincidentally also involved lithium. What both ALSA and MNDA failed to report about the study which started the excitement (“Fornai, et al., 2008”) was that the study essentially “cooked the books” by assigning PALS with slow progression of disease to the treatment group while putting the more standard PALS in the placebo group. This was revealed only after the paper was published. In the GM6 trial, run by two leading and internationally well-respected ALS researchers and clinicians, all participants were randomly assigned to receive either drug or placebo. The only way the comparison to the lithium study would be accurate is if the researchers deliberately placed certain patients in each cohort. Genervon merely supplied GM6. The trial was run and data collected by the two principle investigators (fancy name for doctors who run clinical trials). The analysis was then also done by a contract research facility. So any implication that Genervon somehow fabricated the data is false and besmirches the reputations of two prominent ALS doctors.
It must be noted and repeated that the standard FDA clinical trial practice is indeed extremely important in terms of protecting the public from the unscrupulous. The collection of objective scientific data is the foundation of good medical care. Nobody calling for the Accelerated Approval of GM6 disputes this. However, because ALS is so rapidly and uniformly fatal, we are calling for FDA to utilize the discretion it was granted in the face of an earlier similar crisis (the AIDS epidemic). Further, we call on everyone to realize that the GM6 trial used much more than the ALSFRS as a metric and thus smaller trial populations are much more statistically significant than before. The other end-points used in the Phase 2 are objective and not subject to placebo effect. Therefore, the indication of efficacy observed in the trial should be considered stronger than in previous trials which relied almost solely on the ALSFRS.
The Accelerated Approval Program was created to bridge the gap between the need for data and the urgent unmet needs of patients with rapidly-fatal diseases. The GM6 trial was unique in the strength of the preliminary efficacy signal, largely due to the objective biomarker end-points used. Just like the early days AIDS crisis, PALS have no meaningful treatment options and thus no hope. We want to use the very FDA program created to deal with that situation. We admit the need for more scientific data. But we don’t want to die while it’s collected.