ARTICLE IN BRIEF
Two disappointments in phase 3 trials of ceftriaxone and dexpramipexole for amyotrophic lateral sclerosis (ALS) underscore the need for better pharmacodynamic markers to indicate whether experimental drugs are hitting their target and the recognition of heterogeneity among ALS patients, investigators say.
SAN DIEGO—Two recent large clinical trials in amyotrophic lateral sclerosis (ALS) — testing two different drugs and two mechanisms of disease — both failed to delay disease progression. The final results from these trials, announced at the AAN annual meeting here in March, have led to much soul-searching among ALS researchers about the best way to make progress in discovery of new therapies. “This was a disappointment, but we can learn from it,” said Merit Cudkowicz, MD, a principal investigator on both trials, and chief of neurology at Massachusetts General Hospital in Boston.
The two drugs are ceftriaxone and dexpramipexole. Ceftriaxone is an antibiotic whose promise as an ALS drug arose from a high-throughput screen of FDA-approved drugs that could upregulate the astrocytic glutamate transporter, EAAT2. Loss of this transporter is seen in patients, and is thought to contribute to motor neuron excitotoxicity. In subsequent preclinical testing, it prevented cell death in culture and prolonged survival, albeit modestly, in a mouse model of the disease.
Ceftriaxone was used in a three-part trial in ALS patients beginning in 2006, testing first its ability to enter the cerebrospinal fluid, then its safety, and then its ability to slow clinical decline. Though the second phase was not designed to test efficacy, patients taking ceftriaxone had a 38 percent slowing of their disease progression over 20 weeks compared with placebo (p=0.04), as measured by the ALS Functional Rating Scale (ALSFRS). But the phase 3 portion of the trial, which enrolled over 500 patients in the United States and Canada, was stopped early when interim data indicated it provided no benefit on survival or rate of functional decline.
The second drug, dexpramipexole, is an isomer (a mirror-image molecule) of the anti-parkinson drug. It improves mitochondrial function, protects neurons in culture, and prolongs survival in mice. A phase 2 trial of dexpramipexole in a small number of ALS patients suggested dose-dependent effects on motor function and functional decline, whose sizes were modest but encouraging. That led Biogen-Idec to fund a phase 3 trial, enrolling almost 1,000 patients with early disease — the largest clinical trial so far in ALS. Patients were randomized 1:1 to placebo or 150 mg dexpramipexole twice daily. Randomization was stratified by usage of riluzole, the only approved drug for ALS, so that each arm contained equivalent numbers of patients on and off the drug.
The primary endpoint was the Combined Assessment of Function and Survival (CAFS), a new measure for the ALS field, in which each patient's outcome was ranked against all others, with higher scores for better function and longer time of survival. Treatment arms were compared for their total ranking.
Analysis of CAFS scores at the end of the trial indicated there was no difference between the two groups overall. Similarly, there was no efficacy signal from any of the secondary endpoints, including change in the ALSFRS, time to death, or vital capacity. Dexpramipexole did lower eosinophil counts in some patients, indicating a biological effect of the drug, but there was no correlation between this effect and an effect on CAFS score.
While there are still some post-hoc comparisons left to do, Douglas Kerr, MD, PhD, medical director of neurodegeneration for Biogen Idec, and a co-author on the study, indicated in an interview with Neurology Today that it was unlikely there would be any surprises. “This was a very robust study, with high-quality data, and patients followed to the very end, so we really felt we got a good look at the efficacy of this drug,” he said.
There was one reassuring outcome from the data analysis: patients who were taking riluzole survived longer than those who weren't. The difference on average was about two months, similar to the effect seen in the original trials of the drug in the early 1990s, strengthening the sometimes-questioned conclusion that the drug has a real effect on survival for ALS patients.
Dr. Cudkowicz said there were several important lessons coming out of the two trials. Perhaps the most important was the need for pharmacodynamic markers to indicate whether the drug is hitting the target. “We don't have any good phase 2 readouts,” she said. “We are basically using the same outcome measures in phases 2 and 3,” namely functional decline, “so a phase 2 trial really ends up being an underpowered efficacy study.”
In the ceftriaxone trial, the ideal marker would have been a measure of the density of glutamate transporters in the spinal cord, she said. A PET imaging ligand that can tag these transporters is in development, but was not ready in time for the trial. The ceftriaxone investigators tried to obtain permission to measure transporters in the nasal epithelia as a surrogate, but without a long data trail connecting change in the surrogate with change in the CNS, they were turned down by the FDA. The consequence is that, despite the failure of the trial, it is still unknown whether glutamate transporters are legitimate targets for ALS. “I don't think we can rule out that the pathway is the right one,” Dr. Cudkowicz said.
In the dexpramipexole trial, there was no clear choice for a validated marker that was linked to the presumed mechanism of the drug, and so, again, mitochondria are not ruled out as a target for future trials. Dr. Kerr recognizes that the outcome of the trial underscored the point. “We need pharmacodynamic biomarkers, so we can fail early, not later,” he said. “If it's not working on the target, move on. Those types of decisions can be made when you have good pharmacodynamic markers.”
A second problem is the growing recognition of the heterogeneity among ALS patients, not only clinically, but also mechanistically. “This is a huge challenge in clinical trials,” said John Ravits, MD, professor of clinical neuroscience at the University of California, San Diego School of Medicine. It is probably part of the reason for the failure of recent trials, he said, since it may be that only a subset of patients have disease due to mitochondrial dysfunction, while glutamate transporters are responsible in another subset, and other pathways accounts for other groups.
Dr. Cudkowicz agreed: “We have to get smarter at figuring out the heterogeneity of the [disease] biology.” Future trials might be more successful, she said, if investigators can target a subset of patients with the same disease process. That likely means smaller, not bigger, trials in the future, with a process-specific drug and related pharmacodynamic marker.
The fundamental problem, Dr. Ravits said, “is that we really don't understand disease pathogenesis. We are taking ideas not really based on strong data of disease mechanisms,” lowering the odds of success without the ability to learn clear lessons about mechanisms from failures.
However, he said, modern genomic tools “are bringing the field to the beginning of a renaissance in understanding mechanisms.” Those tools are beginning to allow researchers to conduct “hypothesis-free” searches for disease pathways, evaluating enormous amounts of data to see where it points. And patient-derived stem cells may provide sub-type specific preclinical models in which to test new drugs for their ability to engage putative targets. Combining these approaches may rapidly lead to better drug candidates. Despite the recent disappointments, he said, “I'm not discouraged. The good stuff is yet to come.”
TUNE IN: What can be learned from the failure of phase 3 trials of two drugs for amyotrophic lateral sclerosis? In a video interview, Neurology Today's Editor-in-Chief Steven P. Ringel, MD, and Associate Editor Robert Holloway, MD, discuss the key clinical takeaways from the latest analysis of data on the trials involving dexpramipexole and ceftriaxone for ALS: http://bit.ly/aNQ4KB
A ‘BEST PAPER’ PICK: Neurology Today Editor-in-Chief Steven P. Ringel, MD, and Associate Editor Robert Holloway, MD, selected this as one of the noteworthy papers highlighted by the AAN Science Committee for presentation at the AAN annual meeting.