Tuesday, November 29, 2011

Understanding pathways leads to drug discovery. Really?

One of the most often heard rationales for studying the genetics of disease is that increased understanding of gene pathways, or the products of gene pathways, will lead to new drug discoveries.  The idea is that even if mapping studies, like GWAS (see many of our earlier posts) don't discover 'the'  or 'blockbuster' genes 'for' a trait, they reveal genetic pathways in the biology of the trait that drugs can target.

However, while there have been some successes or hopeful trials, overall there is only limited evidence yet that this is as yet happening -- in fact, after a burst of initial enthusiasm for the new doors opened by the sequencing of the human genome, pharmaceutical companies have been cutting back on research and development, and few new drugs are currently in the pipeline -- drug discovery is risky, and payoffs are few.  This constriction of the new drug pipeline is why Francis Collins, director of the National Institutes of Health, has pushed "translational medicine" and NIH's involvement in drug development.

Well, as the special report on allergies in the Nov 24 issue of Nature notes, immunologists and allergists have long thought they understood what causes asthma, but the disease has increased dramatically in the last 3 decades or so, and new treatments are few and far between.  So much for the benefits of understanding pathway components. 
Since the discovery of immunoglobulin E (IgE) almost half a century ago, there has been a massive expansion in knowledge about how IgE antibodies work. Research has unravelled IgE's role in a myriad of cellular and molecular targets driving inflammatory responses and underlying complex allergic disorders. This knowledge might have been expected to lead to novel preventative and therapeutic pathways — unfortunately, this has not been the case.
The dramatic rise in allergy and asthma worldwide has increased the clinical need for treatment, but research focusing heavily on IgE as the main malefactor in allergies has not been translated into widespread patient benefit.
One problem, according to a piece by Stephen Holgate on why this is so, has been the pharmaceutical industry's reliance on animal models to both better understand the disease as well as to test new treatments.  But, humans are not mice.  In addition, allergic disease is complex, and involves not only biological pathways, but their interactions with environmental triggers as well as, presumably, an underlying genetic susceptibility.
Traditional therapy of allergic disease has in large part relied on the abatement of symptoms with H1-antihistamines (rhinoconjunctivitis, food allergy, urticaria), adrenaline (anaphylaxis) or ╬▓2–adrenoceptor agonists (asthma), and the suppression of inflammation with corticosteroids. Besides improving the pharmacology of known drugs, the only novel asthma therapies to emerge are leukotriene inhibitors (for example, montelukast) and the non-anaphylactogenic anti-IgE, omalizumab, both of which are directed at targets identified well over 40 years ago.
There have been disappointments with a wide range of biologics targeting activating receptors on T cells, cytokines, chemokines, adhesion molecules and inflammatory mediators. Having shown convincing efficacy in in-vitro cell systems and animal models, and possibly some level of efficacy in acute allergen challenge in mild asthma, all of these have fallen short of expectations when trialled in human asthma. In moderate–severe asthma, where the unmet therapeutic need is greatest, trials of novel biologics have revealed only small subgroups in which efficacy has been shown or is suggestive.
And, it has long been assumed that asthma is primarily an allergic response, but this is no longer thought to be so.  The idea now is that perhaps impaired innate immunity comes first, and leads to allergy.  So, much is known about allergies and asthma, but nowhere near enough.

The asthma question is one that highlights many of the current problems in epidemiology, genetics, and the understanding of causation.  Asthma prevalence has been climbing in the US and other industrialized countries since the 1980s.  Given its precipitous rise, it would seem that the problem is quintessentially one for environmental epidemiology, but even when looked at from that perspective, no convincing environmental cause has yet been identified, and in fact studies have produced a frustrating litany of contradictions -- it's cleanliness or dirt, breast feeding or bottle, this gene or that.  Yes, epidemiologists have turned to genetics to try to understand the disease, but clearly there's no genetic explanation for such a recent epidemic. 

Like most other common chronic conditions, asthma is a complex disease, with multiple causes and multiple pathways.  As Holgate concludes:
In the future, it is essential that asthma is not treated as a single disorder, but rather defined by causative pathways. We need new diagnostic biomarkers to identify patients most likely to respond to highly selective biologics, such as anti-IL-5 biologic (mepolizumab) and anti-IL-13 (lebrikizumab). These therapies are only active in particular subtypes of asthma, when the molecules they target lie on a causative disease pathway. 
Studies of large numbers of people with a common chronic disease like asthma, or heart disease or hypertension, are necessarily pooling cases with different causes, pathways, genetic backgrounds, and outcomes.  This limits the potential for successful findings.

Biological traits are the result of interactions among many different factors, genetic and otherwise.  Such interactions, and the way that evolution works, leads to redundancy, alternative pathways, overlapping pathways, and the like.  This was a major theme of our book The Mermaid's Tale.  Complexity is an easy word to say, and perhaps it's easy to use it to excuse failure to find blockbuster findings.  But the last couple of decades have systematically shown causal complexity to be real.

Besides complexity itself, a major problem is not simply that humans aren't mice.  It's also that we are all unique in our environmental exposures, genomics, and how our bodies respond.  Identifying single genes that may be involved in complex diseases, or even biophysiological pathways, may be a rationale for sticking with the genetic approach to understanding disease, but it's a far cry from prevention, treatment or cure.

Rather than promising simplistic causation and consequent intervention miracles, we feel that students and young investigators, and the funding mechanisms, should be geared towards coming to grips with complexity, rather than just spinning out ever more details.  Major practical, and we also believe conceptual challenges lie ahead.  Asthma is a good case in point.

No comments: