In 1999, a paper in Nature Medicine reported that mouse models of the fatal neurodegenerative disorder amyotrophic lateral sclerosis fared better with a simple treatment: a diet supplemented with creatine, a compound that helps regulate energy levels in the brain and muscles (5:347–50). That promising, albeit preliminary, result soon launched not one but three clinical trials, with a total of 386 patients in the US and Europe. Disappointingly, the trials revealed that creatine had no effect in people. It was a familiar outcome: more than 50 other clinical trials of potential amyotrophic lateral sclerosis (ALS) drugs, ranging from lithium to celecoxib (Celebrex), have failed.
Also known as Lou Gehrig’s disease, ALS results from the degeneration and death of motor neurons, and affects approximately two to five of every 100,000 people worldwide. ALS’s devastating symptoms—including progressively worsening muscle weakness and spasming, and difficulties with speech, swallowing, and breathing, leading ultimately to paralysis and death—have led to an intense hunt for treatments to halt its progression.
Unfortunately, the desire to give patients hope has often outstripped good scientific sense. “Many drugs that have gone into ALS clinical trials shouldn’t have, because the preclinical data package didn’t support it,” says Steve Perrin, CEO and CSO of the nonprofit ALS Therapy Development Institute (TDI) based in Cambridge, Massachusetts. Only five of the 420 ALS therapy candidates that his center has retested in mouse and cellular models have shown a therapeutic effect.
Progress has been hindered by three main challenges. First, the disease’s causal mechanisms are poorly understood. Even the two ALS treatments currently approved by the US Food and Drug Administration (FDA), small-molecule drugs riluzole and edaravone, have largely mysterious mechanisms of action and targets, and only modestly improve survival and quality of life.
The desire to give patients hope has often outstripped good scientific sense.
Second, ALS is a highly heterogeneous disease in terms of origin (90 percent to 95 percent of cases are sporadic rather than inherited), initial symptoms (patients may report limb weakness or difficulty in speaking or swallowing), and speed of progression (some patients live months, others decades, after diagnosis). This has made it tricky to model the disease. For many years, the only available mouse models were those carrying mutations in SOD1—the first gene linked to the disorder—which affect only 2 percent of ALS patients.
Finally, there are no quantitative biomarkers to track disease progression or serve as clinical endpoints for trials. Currently, physicians use the subjective ALS Functional Rating Scale–Revised, which measures 12 motor skills on a five-point scale.
In the past few years, the ALS research community has gained traction in meeting these challenges. Genetic discoveries have revealed new pathways involved in the disease. Enhanced technologies generate more-realistic models and hypotheses. And new therapeutic approaches, such as oligonucleotide therapeutics, have begun to enter the clinic. Together, these advances have made many researchers optimistic that effective treatments are finally on the way.
Shedding light on ALS biology
In addition to riluzole, on the market since 1995, and edaravone, which the FDA approved in May 2017, treatment currently focuses on managing ALS symptoms through multidisciplinary care to increase weight and aid breathing, communication, and mobility, says Merit Cudkowicz, director of the ALS Clinic and chief of neurology at Massachusetts General Hospital (MGH).
Although it’s still not clear what triggers onset of the disease, research over the past decade—in particular, the cumulative identification of roughly 50 potential disease-linked genes by multiple research groups—has unveiled numerous relevant mechanisms, most of which act in the motor neurons affected by ALS. These new insights are allowing drug developers to shift from downstream pathways, such as inflammation or muscle weakness, to upstream mechanisms more likely to drive the disease.
Many identified mutations occur in genes coding for RNA-binding proteins, such as TDP-43 and FUS, involved in splicing, transcriptional regulation, or other aspects of RNA metabolism. These genetic changes can lead to cytoplasmic aggregation of these proteins and dysfunctional mRNA metabolism. Intriguingly, most ALS patients, even those without these mutations, have abnormally high TDP-43 levels, which can lead to RNA misprocessing, says neuroscientist Clotilde Lagier-Tourenne of Harvard Medical School. This suggests that targeting these pathways could one day result in broadly applicable ALS drugs.
An emerging player in both dysfunctional RNA biology and other mechanisms of ALS is the gene C9orf72, thought to encode a protein involved in cell signaling. Mutated forms of this gene account for 25 percent to 30 percent of familial ALS cases and up to 5 percent of sporadic cases, making it the largest known genetic driver of the disease. Its 2011 discovery was “a key breakthrough in terms of therapeutic development,” says Laura Ferraiuolo, a translational neurobiologist at the University of Sheffield in the UK. Multiple companies have begun to target the gene with therapeutics.
Smaller RNAs might also have a role to play in ALS. The microRNA miR-155 is upregulated in the spinal cord of ALS patients, and reducing its levels in SOD1 mutant mice improved survival and motor function, says Bill Marshall, CEO of miRagen, a Colorado-based company developing the anti-miR-155 compound MRG-107. He thinks that miR-155 may additionally be acting in neural support cells, such as microglia or astrocytes, which also seem to play a role in the disease.
Proteins misbehave in ALS, too, misfolding and aggregating in the cytoplasm and thereby triggering cellular stress pathways. The actin regulator profilin 1 clumps in this manner, and the mutated profilin found in some patients exacerbates TDP-43 aggregation, too. That’s led researchers such as pharmacologist Mahmoud Kiaei at the University of Arkansas for Medical Sciences to search for compounds to prevent profilin clumping. Other drug strategies focus on broader causes of aggregation, such as nuclear pore defects that result in cytoplasmic protein accumulation. Massachusetts-based Karyopharm’s preclinical compound KPT-350 aims to inhibit the nuclear pore protein exportin-1 (XPO1) in the hopes of alleviating this accumulation. “If it works, it will have a huge impact in showing how basic biology can be translated to the clinic,” says Chris Henderson, vice president and head of neuromuscular and movement disorders research at pharmaceutical giant Biogen, which has acquired rights to further developing the compound.
Such discoveries have aided preclinical research, via the development of new rodent models such as TDP-43 and profilin-1 mutant mice. And the recent rise of cellular reprogramming technology has made it possible to work directly with patients’ own cells. Researchers can convert individuals’ skin cells into neurons that recapitulate their specific spinal cord features, such as protein aggregation and RNA metabolism defects, says Ferraiuolo. The cells have also been a boon for therapeutic testing, allowing her and other researchers to screen 300 drugs per patient per day.
Beyond small molecules
Rather than develop new versions of traditional compounds, some companies are exploring entirely novel kinds of drugs. Antisense oligonucleotides (ASOs), for example, are designed to bind to the RNA transcript of a specific gene and trigger its destruction before translation (see “Waiting for Oligonucleotide Therapeutics,” The Scientist, December 2016). For years, ASOs disappointed in clinical trials, often failing to produce convincing efficacy data. That changed with Biogen’s recently approved ASO nusinersen (Sprinraza), which treats the childhood motor neuron disease spinal muscular atrophy. Although the disease and population are different from ALS, that success has boosted optimism for Biogen’s continued partnership with Ionis, Spinraza’s original developer, to test an ASO targeting SOD1 in ALS.
The SOD1 ASO avoids many of the pitfalls of previously proposed ALS treatments: “You know your target, you know your mutation, you have a drug that works at your mutation, and you actually have readouts,” explains Jeffrey Rothstein, director of the Robert Packard Center for ALS Research at Johns Hopkins University. In mutant SOD1 mice, the ASO knocks down overall protein levels and improves survival and muscle strength; currently, a Phase 1 multiple dosing trial seeks to discern a similar effect in people. “It could be the first domino to fall,” says ALS TDI’s Perrin, potentially making way for the development of ASOs for other ALS targets. Biogen and WAVE Life Sciences are developing C9orf72-targeting antisense oligonucleotides that are expected to enter clinical trials soon.
Another promising approach, gene therapy, also received a boost in the therapeutics industry, following the 2009 discovery that the viral vector adeno-associated virus 9 (AAV9), unlike most vehicles, could cross the blood-brain barrier (Nat Biotech, 27:59–65). Gene-therapy company AveXis, recently acquired by Novartis, pairs AAV9 with short hairpin RNAs targeting SOD1 in its ALS therapy AVXS-301. Like the SOD1 ASO, AVXS-301 extends life and improves motor function in SOD1 mice. It’s currently in preclinical safety testing, and AveXis plans to file an investigational new drug application (IND) by early 2019, says CSO Brian Kaspar. He says he hopes that it might even be useful in patients without SOD1 mutations who nevertheless have abnormal levels of the protein, and that, if successful, AAV9 could become the basis for ALS therapies targeting other genes.
I think it will be curable with the right strategy.—Mahmoud Kiaei, University of Arkansas
Rather than altering genes, the cell therapy NurOwn, developed by New York City–based BrainStorm Cell Therapeutics, consists of cells differentiated from patients’ own bone marrow-derived mesenchymal stem cells (MSCs). These MSCs are thought to produce protective neurotrophic factors and immunomodulatory molecules that both help maintain motor neuron function and reduce inflammation, explains Chief Medical Officer and COO Ralph Kern. A Phase 3 trial aims to enroll 200 patients, half of whom will receive three spinal injections of these cells at two-month intervals; results are expected in early 2020. However, despite positive reported results to date, some remain skeptical. “The hype about stem cells is masking the unbelievable ignorance about the biology of these cells,” says Rothstein, who argues that it would be better to administer specific growth factors. Kern counters that the mix of factors provided by the cells is critical for targeting different types of neurons and different patient groups.
Designing better trials
Both Kern and Jean Hubble, VP of medical affairs at Mitsubishi Tanabe Pharma America, the maker of the recently approved edaravone, point to a new strategy to develop promising ALS therapies: restricting clinical trials to patients with rapidly progressing disease, which allows scientists to more easily determine the therapy’s impact. That approach might lead to disappointment for some patients in the short term, but MGH’s Cudkowicz thinks that more-carefully targeted trials are ultimately more likely to identify effective drugs for a broader swath of patients (see “Clinical Matchmaker,” The Scientist, June 2015).
To get there, some ALS researchers are pursuing large-scale studies to develop better ways to stratify their patients and quantitatively track disease progress. For example, through its precision medicine program, which has enrolled 700 patients, ALS TDI uses data from accelerometers strapped to patients’ limbs to monitor disease progression, says Perrin. The initiative also collects standard Functional Rating Scale–Revised results, voice recordings, full genome sequences, and RNA and protein biomarkers, which are made available to each patient through a secure online portal. A separate, multi-institution effort, Answer ALS, is beginning to spot key differences between patient groups based on the 8 billion data points (on motor function, breathing, speech, speed of thought, and molecular profiles of individualized iPS cells) collected from each of the 780 patients enrolled so far, says Rothstein, who is involved in the program.
Both efforts rely on partnerships with big-data experts: ALS TDI is working with Google and the Broad Institute, while Answer ALS links multiple medical centers, IBM Watson, and researchers at MIT. The breadth and scale of these alliances highlight another new trend: the rise of large collaborative teams in ALS spanning basic research, the clinic, and nonprofits. “There’s much more cohesion, and the academic-industry partnerships are extremely vibrant,” says Lucie Bruijn, chief scientist of the ALS Association, a nonprofit aimed at eradicating the disease.
That’s not to say there won’t be failures. In the past year, those include Cytokinetics’s tirasemtiv, a small-molecule no longer in development after a Phase 3 trial failed due to tolerability issues and lack of efficacy; Neuraltus’s NP001, a proprietary formulation of sodium chlorite, which missed clinical endpoints for efficacy in Phase 2 trials; and AB Science’s masitinib, an anti-inflammatory, small-molecule tyrosine kinase inhibitor, which was rejected by European regulators on the basis of unconvincing trial data.
Overall, however, the explosion of biological and technical advances in the ALS therapy field—as well as growing connections between the players—have led to general optimism that future drug development might finally be able to avoid past pitfalls.
“ALS has been labeled incurable, but I think it will be curable with the right strategy,” says the University of Arkansas’s Kiaei. Steve Perrin agrees: “Many of the things that are in clinical development today are better shots on goal than they were a decade ago.”
Jenny Rood is a freelance writer in Cambridge, Massachusetts, and the senior development writer at the Broad Institute of MIT and Harvard.