Tuesday, June 12, 2012

Animal Models and the Future of Psychiatric Research

"Few medicines, in the history of pharmaceuticals, have been greeted with as much exultation as a green-and-white pill containing 20 milligrams of fluoxetine hydrochloride — the chemical we know as Prozac" wrote Columbia University Assistant Professor of Medicine Siddhartha Mukherjee in a recent New York Times editorial on antidepressant efficacy. As Dr. Mukherjee points out, the rise of antidepressants over the past several decades has been swift and staggering. A recent CDC study found that antidepressants are the most commonly prescribed drug for Americans between 18 and 44 years of age, with 11% of Americans over the age of twelve utilizing such medications.
Original image from the Global Information Exchange Network

In light of such overwhelming usage rates, one would think that pharmaceutical companies would be falling over themselves to invent the next antidepressant superdrug. As it turns out, however, exactly the opposite trend is emerging. Pharmaceutical giants AstraZeneca and GlaxoSmithKline recently cut funding for new psychiatric drugs, leading psychiatrists David Nutt and Guy Goodwin to lament that "The well-reported pull-out of pharmaceutical companies from neuroscience research that has occurred in the past year is a major concern... the withdrawal of research resources is a withdrawal of hope for patients and their families."

One popular explanation for pharmaceutical disinterest in new antidepressant drugs stems from what Dr. Steven Hyman has called a surfeit of "refried serotonin" - that is, the continual redesign of antidepressant drugs that affect levels of monoamine neurotransmitters such as serotonin and norepinephrine in essentially similar ways. While this paradigm has led to the development of useful antidepressant drugs (Lexapro, Paxil, etc.), Hyman argues that the great advances of recent decades have been in safety and tolerability, but not in efficacy, which may have peaked in the 1950s. Given that only one third of patients fully recover from depression on the current generation of antidepressants, it's clear that a need for more effective therapies persists.

Are antidepressants based on altering synaptic levels of serotonin and norepinephrine the best we can do, or are there new drugs, yet to be discovered, that may revolutionize the treatment of depression? Some have interpreted pharmaceutical funding cuts as indicative of an over-reliance on drugs in the treatment of psychiatric disorders and have suggested increased emphasis on non-pharmacological treatments such as psychotherapy; others, however, have called for the development of new methods and paradigms within pharmacology as a means to develop more effective psychiatric drugs. With Dr. Steven Hyman's visit to Emory to accept the Neuroscience and Ethics Award, we had the opportunity talk to one of the most respected individuals in the field about future research directions in psychopathology.

Moving Beyond Animal Models

Dr. Hyman began by criticizing reliance on animal-based assays that were originally developed to produce new antidepressant drugs that acted like older drugs. The classic "forced swim" and "tail suspension" tests were developed based on antidepressant drugs discovered in the 1950s. In the forced swim test, rats are placed in a water-filled cylinder from which they cannot escape. Researchers found that rats pretreated with certain antidepressants will attempt to swim for longer periods of time before giving up. Similarly, rats suspended by the tail and given an antidepressant will struggle for longer in comparison to rats which are not given an antidepressant. Because these assays were worked out using antidepressants already known to work in humans, and because they successfully predicted the efficacy of similar antidepressant compounds, they became standard tools in both academic and industry labs.

Hyman expressed concern on several fronts: 

"First, these are 'black box assays,' not disease models. As a result they may have limited predictive power and may actually screen out potentially useful drugs. Second, they are not even good models of what antidepressants do in humans. In these assays, antidepressants are active following a single dose, whereas in humans therapeutic responses required weeks. At a minimum this should remind us that there is much that we do not know.” 

He added that these screens have invaded basic academic research. Investigators may laboriously investigate the function of a gene by knocking it out in the mouse genome or through injecting it into specific brain regions using a viral vector. They will then use forced swim or tail suspension to argue that their mouse may be a depression model, using the term “behavioral despair” to describe rodents who fail to swim or struggle. This, said Hyman, is “an anthropomorphism that behavioral pharmacologists use to deceive themselves about our continuing difficulty of modeling depression in animals.”

Not the same thing.
Original image from drugdiscoveryopinion.com

Moreover, many, including Dr. Alan Schatzberg, a former Chairman of the Department of Psychiatry at Stanford University's School of Medicine, have argued that animal models for depression simply cannot model the cognitive and emotional aspects of depression that are central to human experience. Insofar as animals do not feel guilt or hopelessness, researchers such as Dr. Schatzberg argue that animal models cannot measure the same disease processes that exist in depressed humans.

Dr. Hyman noted that animal models have yielded clear success in studying highly penetrant monogenic disorders such as Rett Syndrome and Fragile X syndrome, as well as the functions of circuits that are highly conserved in evolution such as "basic emotions of fear and reward." However, he also noted that these models have had limited utility to date in studying common mental disorders, which are generally polygenic, heterogeneous, and involve circuits in the prefrontal cortex that are not well modeled in rodents.

Animal models have proliferated largely as an alternative to testing novel pharmaceuticals on humans, a practice which holds potentially problematic ethical implications. As a result, a difficult and important question emerges: how can testing procedures be refined to produce effective psychiatric medications without putting humans at excessive risk?

New Directions in Human Research

Dr. Hyman suggested a number of novel methods that he believes could reinvigorate research into psychopathology and therapeutics. One possibility, which he noted is "still almost as much science fiction as real," is based on the rapidly progressing technology of reprogramming peripheral cells such as human fibroblasts into neurons. Future work in this area may be capable of creating replicable neurons of specific types and assembling them into circuits, perhaps on a chip. As knowledge of the genetics of mental disorders advances, disease-risk versions of genes could be engineered into such cells and then compared to patient cells. Dr. Hyman suggested that study of biochemical and cellular phenotypes may yield therapeutic benefits that exceed those produced by studies of rodent behavior.

Dr. Hyman also saw promise in innovative forms of "human experimental biology" that combine treatment interventions such as deep brain stimulation with brain imaging to understand underlying circuits in neuropsychiatric disorders. In particular, he praised Emory researcher Dr. Helen Mayberg for continuing her research on deep brain stimulation (DBS) as a treatment for depression, even after DBS had demonstrated therapeutic benefit:

"I think I have great respect for what Helen Mayberg and the group are doing here because they didn't stop at saying, well, this works, meaning it improves the Hamilton Depression rating in otherwise treatment refractory patients 60% of the time. They've also followed it up by -- with very inventive imaging approaches to see what circuits are involved. I think that within the bounds of careful, ethical and safety considerations we must learn from humans what we cannot readily learn from animals.”
Dr. Helen Mayberg, neurologist at Emory University.
Original image from emory.edu
More broadly, Dr. Hyman emphasized the need to think creatively about novel ways to approach long-standing research problems. He advocated convergent interdisciplinary approaches to problems and avoidance of “cottage industries” that lead nowhere. In particular, Dr. Hyman noted the hundreds of genetic associations studies that have been done on genetic variants such as the length polymorphisms in the serotonin transporter gene, which have already been identified as the target of a host of moderately efficacious drugs.

Neuroethical Implications

Research with human subjects raises a number of unique problems. The first is a problem of external validity: current trials with humans have been criticized on a number of grounds, including excessively stringent exclusion criteria and selective publication of studies demonstrating positive results. The second is an ethical problem: research with human subjects involves issues of safety, informed consent, and possible exploitation of vulnerable populations such as prisoners or children. These and similar concerns have led some, such as the Alliance for Human Research Protection, to oppose past efforts to expand drug experimentation on humans.

More broadly, the insufficiency of current research paradigms in psychiatric drug development raises questions about the fundamental nature of psychiatric practice. Are mental disorders best understood as physical illnesses like any other, discrete neurological deficits that can be cured through the use of appropriate medications, or are they what some psychiatrists have termed "problems in living," more loosely defined constructs that don't necessarily imply biological solutions?

Robert Whitaker, a vocal critic of the biomedical paradigm in psychiatry, suggests that funding cuts for new pharmaceuticals may have a "silver lining": "it is possible that the drug tsunami that has swept over our society will begin to ebb, and this will provide our society an opportunity to rethink its psychiatric care." For Whitaker, there is little hope that new research methods will substantially contribute to our capacity to treat mental illness, and we would be wise to reconfigure our understanding of mental illness to emphasize aspects of "wellness" rather than "disease." Such claims imply that work with animals is likely to yield little benefit in the treatment of depression, as animals cannot model uniquely human aspects of depression such as guilt and hopelessness. Whitaker's arguments are also likely to be met with skepticism in research circles, where the disease model of mental illness still prevails. Whatever new paradigms are adopted within psychiatric research, then, are likely to hold significant implications not only for what sorts of treatments are developed, but also for our understanding of what it means to be "normal" and "disordered" more generally.

Want to cite this post?
Gordon, R. (2012). Animal Models and the Future of Psychiatric Research. The Neuroethics Blog. Retrieved on , from http://www.theneuroethicsblog.com/2012/06/animal-models-and-future-of-psychiatric.html


Kristina Gupta said...

Hi Ross,

Thanks for this thoughtful blog. You cover a lot - from the usefulness of animals models, to "human experimental biology," to whether mental illness should be conceptualized as physical illnesses or "problems of living."

In regards to the last point, my position has been that we need to approach mental illnesses as BOTH physical illnesses and social problems or problems of living. Do you think the two approaches can be pursued at the same time or are they inherently opposed to each other?

Ross Gordon said...

Hey Kristina, thanks for the comment.

I agree that both approaches can absolutely be pursued at the same time. I think what's important isn't whether depression is "really" a brain disease (or "really" a social/existential problem) but rather which attribution is most likely to produce positive results in a particular case.

If, hypothetically, someone invented a new drug that was 95% effective at ameliorating depression without incurring significant side effects, I think that depression would be, for all intents and purposes, a "brain disorder." If you take the 33% efficacy of current antidepressants at face value, though, I think you have to (at least provisionally) be a little more flexible in what you take "depression" to mean.

That's why I think new psychiatric drug development is so important in thinking about what it means to be "mentally disordered": in some ways, I think the "brain disease" model as much hypothesis than fact, and the development of a new, supereffective antidepressant would go a long way towards supporting the latter characterization.