Base hits
After two days of talks replete with exciting molecules, pathways and animal models, on Thursday afternoon the organizers of the Cell Symposia: Autism Spectrum Disorders: From Mechanisms to Therapies conference in Washington, D.C. got down to business.
They convened a panel of heavyweights from pharmaceutical companies and private funding organizations and asked: How can bench discoveries be translated to the clinic?
The responses from two pharmaceutical representatives were surprisingly pessimistic — especially regarding mouse models — and elicited palpable disappointment among basic scientists in the crowd.
The bottom line, according to these two experts: Drugs for autism aren't going to appear anytime soon. If and when they do materialize, they will probably treat a small and clearly defined subgroup of the autism spectrum, and provide a relatively small improvement in symptoms.
In basic autism research, "progress has been remarkable, especially the genetics and the creation of animal models," said panelist Morgan Sheng, vice president of neuroscience at Genentech. "But frankly, I don't think it's really ready for drug discovery."
Despite a flurry of interest in autism a couple of years ago, pharmaceutical companies are gradually pulling away from their neuroscience programs. In fact, according to panelist Michael Ehlers, chief scientific officer for neuroscience research at Pfizer, the crowd in front of him, which included a couple of hundred researchers, was "eight or nine times larger than all of the people working on autism in industry."
Developing drugs to treat autism is difficult for many reasons. First, the disorder is notoriously variable. A treatment might be effective in a subset of those with the disorder, but that result might be obscured by the group data. Second, there aren't yet any reliable and objective biomarkers to prove that a drug works.
"The core symptoms of autism are subjective and behavioral," Sheng said. "You really want to have some quantifiable measure that can be followed and hopefully correlated with outcome. Unfortunately, that kind of research is not extremely intellectually exciting."
Third, pharma companies aren’t interested in drugs that work on mouse models of autism. As audience member William Snider of the University of North Carolina pointed out, some drugs cure stroke and Alzheimer's disease in mice. But when companies tested those drugs in hugely expensive human clinical trials, they flopped.
"These models, it's time for them to go," Ehlers said.
Successful psychiatric drugs have rarely resulted from studies on genetics or animal models, Ehlers noted. For example, for six decades the only drugs that have consistently treated schizophrenia are D2 agonists, which have no genetic link to the disease.
Instead of aiming for a blockbuster autism drug based on biology, he says, researchers should try to tweak a symptom or two. For example, a drug might make children more responsive to behavioral therapy. "We don't need to hit home runs. We just need singles," Ehlers said.
The trouble is, clinical trials, even small, preliminary ones, cost hundreds of millions of dollars. Seaside Therapeutics, a small biotech in Cambridge, Massachusetts, has had considerable success testing drugs for fragile X syndrome. But the company's founder, panelist Mark Bear, said that it never would have happened without the help of an angel investor who has a personal connection to the disease. "A lot does come down to money," Bear said.
The two panelists from private organizations, Rob Ring of Autism Speaks and Gerry Fischbach, SFARI’s scientific director, both said that nonprofits and the government can help with the expensive and less exciting parts of preclinical testing. But no public-private partnerships are in the works.
After 90 minutes of sobering discussion, Sheng ended the session on a surprisingly cheery note. Basic scientists, he said, should continue their work at the bench without fretting about base hits or homeruns. "I wouldn't worry too much about trying to translate.”
For more reports from the 2011 Society for Neuroscience annual meeting, please click here.
Comments
I understand the caution expressed by Drs. Sheng and Ehlers during the animated Panel discussion. They pointed out how difficult it is to bring a drug to market citing many recent failures in the area of neuro-psychiatric disorders. The endgame is expensive so companies must be selective.
But I am a congenital optimist, and that leads me to believe (hope) that new candidate therapies for ASDs are on the horizon. “Chance favors the prepared mind” so we must sustain our efforts in genomics and expand our efforts in mouse models.
During the Cell Press Symposium, exciting results were presented on several mouse models including CNTNAP2, NLG3, Shank 3, Shank 2, DISC1 and MET. Computational approaches to metabolic pathways may reveal “druggable” targets associated with these genes. The mTOR and WNT pathways are prime candidates. Chance does favor the prepared mind after all.
The alternative to mechanism based screening is mindless assays of thousands, perhaps millions of molecules – or waiting for them to arise by chance while treating other disorders. I am not opposed to that, but we should not give up on mechanism based biomarkers in mice and other organisms and in cells derived from such models and from people diagnosed with ASD. Behavioral measures in people are precise and should not be pushed aside in the as we search for molecular biomarkers.
The heterogeneity of ASD is often cited as a challenge to drug discovery, but I think quite the opposite is the case. A multiplicity of risk factors, if they are indeed convincing as are many of the newly discovered de novo variants will open up a multiplicity of therapeutic approaches. None of this would have been discussed at the Society for Neuroscience five years ago!
The techniques of genomics and the mathematical analyses of the results advance every month it seems, imaging has reached a new level and our understanding of synaptic biology is more sophisticated.
We should not expect a drug or device that will reverse all aspects of the ASD complex. A modest effect on social cognition that improves the quality of life of those on the spectrum would be considered a “blockbuster”. That is an optimistic, not a pessimistic vision.
Beyond mice we should invest in flies and fish and worms and non-human primates. Human subject research is, of course the endgame. This is what will bring Pharma back to the table. Public/private/industrial partnerships are needed in the “precompetitive space” and also in the endgame. How can we overcome or modulate the drive for profits – by devising a model in which risks and profits are shared. The model now in development by the European Union may be a beginning.
While the era of rational drug development has proved more difficult than most people originally anticipated, I also think that autism drug development is now systematically feasible, even within the tight constraints faced by industry programs. All agree that there are a growing number of handholds on autism pathogenesis, driven primarily by the ongoing rapid elucidation of risk genes. Some intriguing pathway and circuit convergences, as well as animal models, are emerging; in addition, refinements in delineating cognitive and behavioral changes in individuals with ASDs is combining with advances in neuroscience to identify neural system targets, for example oxytocin receptor-mediated social affiliation, that may not be far removed from desired therapeutic endpoints. Furthermore, as Gerry Fischbach points out above, a therapy achieving even modest benefits on core symptoms would constitute a very meaningful breakthrough, and ASDs are sufficiently common that even a sliver of the affected population could constitute a market size sufficient to support commercialization. Several therapeutic approaches have shown rescue benefits when administered to older animal models, and a few pioneering clinical trials are underway.
The key to sustained success I believe will lie in acknowledging upfront that understandings of relevant biology will be incomplete for the foreseeable future, and planning to take an expanded number of "shots on goal" in the form of early and relatively inexpensive exploratory clinical trials (1). To be able to mount an increased number of trials in a world of reduced resources, we will need to develop several significant platform capabilities including trial networks, efficient methodologies, and suitable early efficacy biomarkers. Creation of these enabling platforms will mostly be beyond the reach of individual organizations, but is achievable by broad public-private partnerships engaging academia, companies, government, regulatory agencies, and private foundations in pre-competitive space. Fortunately, such partnerships are indeed beginning to form in the autism field, with one example being the autism-targeted component of the European Innovative Medicines Initiative (EU-AIMS). SFARI seeks to contribute to the effort, and among several ongoing projects has joined with the Foundation for NIH and Autism Speaks to try to nucleate a public-private partnership focusing on the development of early efficacy biomarkers.
--------------------------------
(1 - Perhaps in this regard the autism field is not really so different from other fields more widely considered today to be ready for drug development? The long litany of recent drug program failures across therapeutic areas, derailed by factors yet unknown or only coming into focus well after program launch, supports this perspective.)
As a neuroscientist who now works in nonprofit research administration and who has a child with autism, it seems to me that efforts to better characterize the variability of the population of people with autism, clincally, genetically and from a pathophysiological standpoint, would go a long way toward "clinical trial readiness." there may never be a one size fits all therapeutic and there probably shouldn't be one size fits all trials. These are things that can be done in parallel to the mechanistic work that may lead to therapeutic targets. I hope this on the radar for both SFARI and Autism Speaks.
Sharon, I agree that must do everything possible to "stratify" the population of individuals considered to be on the autism spectrum. Different systems may be involved to different degrees. We certaily would not treat all infections with the same antibiotic. Reliable, sensitive biomarkers are needed for diagnosis and for monitoring candidate therapies.
Parents are contenders in this battle and IMO the two pharmaceutical "heavy weights" couldn't be more wrong. In the last 10 years, autism numbers have skyrocketed with 1 in 88 American children now effected. NOW is the time for autism drug research. Autism families have waited long enough. Thankfully biotech companies like Seaside Therpeutics has teamed up with Roche to bring one of the first autism drugs to market to treat Retts syndrome. The drug called STX209 a GABA-B Agonist is currently in Phase 3. Results so far have been very positive. Hopefully the success of this drug will open doors for futher autism drug research.