50 Epilogue: Bench to bedside and bacic to bench
STEPHAN R. TARGAN, LOREN C. KARP AND FERGUS SHANAHAN
Where we have been and where we are going
The preceding chapters combine to provide an excellent depiction of all facets of the study of IBD, from bench to bedside. The individual chapters are components of a coherent, albeit detailed, story of the local and systemic pathophysiology of intestinal inflammation, presented in association with a well- reasoned series of management strategies based on this knowledge. Research advances and current concepts of etiopathogenesis are described in the context of what is already known of the clinicopatho- logic features of these disorders. The book as a whole illustrates the effectiveness of a multilateral approach by basic scientists and clinician investiga- tors in the field of IBD. The conclusions of the chapters were drawn purposefully to offer a glimpse of where the field is moving to stimulate research back at the bench.
The interplay of state-of-the-art technology and greater understanding of the pathophysiology in both human and animal disease has brought us far along our road, and has led to excellent translational research resulting in the development of better ther- apeutic options. Upon this model our efforts can be maximized, our scope expanded, and our progress expedited to take us the rest of the way. We know IBD comprises numerous discrete entities that can be stratified by a variety of subclinical and clinical markers that may well identify which patients are likely to respond to any particular therapeutic inter- vention. We also have made great progress toward defining the roles played by infectious agents in the etiology or pathogenesis of IBD. Evidence suggests that the elevated immune function, characteristic of these diseases, is in response to normal commensal bacteria rather than any specific pathogen. It is unlikely that the role of infectious agents correlates directly with the disease responses that are charac- teristic of IBD. More likely the process is indirect,
requiring any number of combinations of genes, immune responses, and environmental triggers, to manifest as disease. Specific manipulation of bacter- ial expression and the corresponding immune response may be the basis of very effective therapies in the near future. Research over the past several years has changed the management of IBD. The effectiveness of therapeutic anti-tumor necrosis factor-oc monoclonal antibodies (anti-TNF-oc) in a portion of patients with Crohn's disease illustrates the unique immune mechanisms that underlie the intestinal inflammation in different subpopulations of patients.
The foundation of our Bench to Bedside and Back to Bench concept has been laid, and the planks of the foundation are described in detail by the chapters of this book. The final decades of the last century saw the Bench to Bedside portion of the equation evolve.
New therapies based on laboratory science were developed, tested and some came to be used in clinical practice. At the very end of the last century only a few compounds were tested in humans with associated parallel basic science, or translational, investigations. Although such studies were few, the yield in terms of information was great.
Not long ago only a few animal models were available for the study of dysregulated intestinal inflammation. In the intervening years the number of models has grown to well over 30, mainly in the form of genetically engineered mouse strains. Analy- sis of these models has taught us that chronic colitis can be induced by diverse genetic abnormalities.
However, the inflammatory patterns and their clinical expressions are also distinct, yielding new opportunities to link alterations in specific genetic loci with alterations of T cell responses, mucosal permeability, and in-vivo clinical patterns.
Another striking area of progress in IBD research that has traversed from the Bench to the Bedside and back again is related to the issue of a microbial pathogenesis of IBD. Despite more than 50 years of
Stephan R. Targan, Fergus Shanahan and Loren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 887-891.
© 2003 Kluwer Academic Publishers. Printed in Great Britain
research a specific bacterial or viral organism has not been identified as a direct cause of IBD pathogenesis.
Investigations in animals, however, identified a role for commensal bacteria in induction of dysregulated mucosal inflammation. Verification of bacterial involvement came with the finding that, despite the genetic or cellular manipulation in any given model, its attendant mucosal inflammation is lost in a pathogen-free environment.
We now know that there are two ways in which the mucosa responds to bacterial antigens. The antigen- specific T cell response is well characterized. A more complex relationship exists between bacterial structures such as lipopolysaccharides and peptido- glycans that have repetitive sequences that are recog- nized by cell receptors. The relationship is receiving a great amount of attention at the time of this writing.
The most exciting of these receptors appear to be present on epithelial cells and macrophages and are known as loll-like receptors. First described in Drosophila, toll-like receptors engage bacterial structures and generate mucosal inflammation.
Alterations in such receptor signaling may well play a key role in the inflammation of IBD. Bacterial antigens are likely to trigger certain T cell responses in the mucosa, and the balance of these responses is critical to regulating the intestinal immune response.
T cell responses can be divided broadly into 'disease- inducing' and 'disease-preventing', depending upon the cytokines released. In the majority of animal models, the presence of effector cells secreting T helper-1 (Thl) cytokines (interferon-gamma, IFN- y, TNF-oc) is responsible for disease, and by this criterion these models are reflective of human Crohn's disease. One clear exception is the T cell receptor-alpha (TCR-a) 'knockout' mouse in which a unique subset of TCR-P dim/interleukin (IL)-4 producing cells is responsible for inflammation; this model may therefore be considered to be more similar to human ulcerative colitis. Two populations of disease-preventing or down-regulatory T cells have been described; Tr-l/IL-10 producing T cefls and the T-helper-3 (Th3)/transforming growth factor-beta (TGF-P) producing T cells. Studies have demonstrated that these two T cell populations counterbalance the effector cells and protect against dysregulated inflammation. It is an imbalance between these effector and regulatory populations that presumably results in disease expression.
Renewed interest in commensal bacteria and their relationship to T cell responses in the mucosa has led to investigations demonstrating that the transfer of a
specific subset of bacterial antigen-stimulated Thl cells can induce colitis in animals. Furthermore, down-regulatory or T-regulatory-1 (Tr-1)/IL-10- producing cells, stimulated by the same bacterial antigens as the pathogenic Thl cells, prevent expres- sion of colitis when co-transferred with these anti- gen-specific Thl cells. These bacterial antigen-trig- gered effector and regulatory cell populations clearly influence the interplay between clinical expression and mucosal inflammation. The diversity in expres- sion of colitis, i.e. cecal involvement, rectal ulcera- tion, specific ileal involvement, ulcerative lethal dis- ease, is dependent upon any one or more genetic abnormalities. Furthermore, the same genetically altered animal can manifest a different expression of disease when raised in different housing facilities, suggesting an additional mechanism for the hetero- geneity of these diseases. It is quite possible that genetic and microenvironmental heterogeneity in these animal models is similar to what is seen among patients with IBD. This paradigm is now the founda- tion of our understanding of pathogenic mechanisms in human disease.
The investigations in animal models of mucosal inflammation have proceeded in parallel with those in human IBD. Much progress has been made in identifying the genes associated with these inflam- matory disorders. Genome-wide scanning and can- didate gene approaches have revealed several dis- ease-associated loci. One important implication of these studies is that IBD is an oligogenetic disease:
genome-wide scans thus far have revealed at least four loci confirmed by independent groups, and an additional three others are suggested. Some of these chromosomes and loci are associated with Crohn's disease and/or ulcerative colitis individually, and others are associated with both. Marker antibodies and/or disease expression further refines the associa- tions. Multiple studies in humans, as well as the animal studies summarized above, have demon- strated diff^ering genetic associations with clinical phenotype as well as marker antibody expression.
Allelic diversity at the susceptibility loci thus prob- ably underlies the immunologic and clinical hetero- geneity of IBD. The chromosome 16 locus was the first to be identified by a whole genome approach.
The IBDl gene (NOD2) has been identified by using a positional cloning strategy - linkage followed by linkage disequilibrium mapping. Three independent associations for Crohn's disease were discovered.
Simultaneously, by a candidate gene approach, an
independent group also demonstrated association
between the NOD2 and Crohn's disease. Clearly, NOD2 is a susceptibility gene to Crohn's disease, as it increases an individual's risk for Crohn's disease, but it is neither sufficient nor necessary for the development of the disease.
Significant advances have already been made, based on the identification of NOD2 and its relation- ship to subtypes of Crohn's disease. Hugot et al and Abreu et al. have recently published separate reports demonstrating that fibrosing Crohn's disease is asso- ciated with the dose effect of two mutations, and several other groups have further confirmed these findings, with results not yet published at the time of this writing. These studies indicate that abihty to predict which patients will experience a more complicated course is in the immediate future.
NOD2 will no doubt continue to shed light on the etiologic pathways of IBD and accelerate the discovery of additional susceptibility genes for IBD.
The identification of IBD-associated marker anti- bodies at the bench has made a great impact on patient care at the bedside, and in turn has stimulated further laboratory research. Since the discovery of perinuclear antineutrophil cytoplasmic antibodies (ANCA) more than 10 years ago, much has been learned about the significance of marker antibodies in the serum of patients with IBD. In addition to pANCA, interest has focused on antibodies to Saccharomyces cerevisiae (ASCA), which have been shown to differentiate between ulcerative colitis and Crohn's disease. Progress continues with these anti- bodies, and other bacterial antigens have been detected which identify strongly with disease-asso- ciated antibodies in IBD. These marker antibodies are important for their diagnostic utility and stratifi- cation of patients into distinct clinical phenotypes.
These antibodies are associated not only with the different l o c a t i o n s of disease, but also with aggressiveness of disease course and responses to treatment by specific manipulation of cytokines. It is of particular interest that some of the marker anti- bodies, such as pANCA, appear to be cross-reactive with commensal bacterial antigens. Therefore, the presence of marker antibodies provides a platform as a basis to study bacterial-host interactions as well as to define subsets of patients with Crohn's disease and ulcerative colitis. At the bench we shall now endeavor to define the relevance of these findings in animals to the diversity of human disease seen among patients with IBD. In addition we must determine which immune alterations and bacterial sensitivity in the different animal models are represented among the
varied clinical expressions of disease seen in these disorders. From our findings in animal models, we assume that patterns of immune responses to environmental and bacterial antigens differ widely among groups of patients with Crohn's disease.
Landers et al have recently reported that expression of certain serologic markers clusters in certain patient subgroups. Even in groups of patients that have been classified as 'indeterminant', expression of certain markers remain consistent over time and suggest the clinical course that is ultimately taken, as has been shown by Joossens et al. Thus, what seemed futuristic only recently, clinically useful information derived from studying genes and their relationship to immunophysiologic responses will be obtainable in a much shorter time. It is therefore possible that certain antibiotics might be most effective in those patients who have the most robust responses to bacterial antigens. Likewise, the most robust responses to the broadest number of these antigens may define those patients who can best be treated by manipulation of the bacterial flora. If the antigens to which an individual is sensitive are among the anaerobic bacteria then metronidazole may be more effective. If these antigens are among the aerobic bacteria, then these patients may best be treated with ciprofloxacin. Finally, in patients who are immunoreactive to both aerobic and anaerobic populations, such patients may require a combina- tion of both metronidazole and ciprofloxacin.
Although many inflammatory molecules are elevated in the mucosa of patients with IBD, their precise role in disease pathogenesis has never been determined. In animal models the success of treat- ment by manipulation of Thl cytokines, e.g. anti- bodies toTNF-oc, IFN-y or IL-12, led to trials of anti- TNF-a for the treatment of Crohn's disease. Results of these trials showed for the first time that cytokine expression in the mucosa could alter disease course in the majority of patients with Crohn's disease.
'Back to Bench' research demonstrated down-regula-
tion of Thl cytokine-producing cells in the inflamed
mucosa to the level of those in the uninflamed areas
as one of the mechanisms responsible for the disease
ameliorating effect. Unlike some animal models in
which cecal bacterial antigens drive the dominant
Thl profile, no similar antigen trigger has been
identified in human disease. A proportion of Crohn's
disease patients can be treated by manipulation of
the bacterial flora by either antibiotic therapy or a
diversion of the fecal stream, which suggests diverse
responses to bacterial antigens among the Crohn's
disease population. In ulcerative colitis, patients with high levels of pANCA and serum IgG reactivity to the E. coli antigen, OmpC, are three times more likely to develop chronic pouchitis than patients without high levels of pANCA. These findings, taken together, highlight the potential importance of bac- terial interactions in the induction or maintenance of inflammation. Overall, the diversity of clinical expression seen in IBD is very likely dependent upon a combination of genetic abnormalities and the presence of certain commensal bacteria.
The 21st century - the age of translationalism
With the rapidly growing body of knowledge, it behooves us to effect the most expeditious applica- tion of our findings to patient care and research in return. We must now determine which of the many discoveries made in animals are relevant to human disease. To do so requires a process by which ideas can be exchanged among investigators and work done simultaneously, crossing between animal mod- els and in-vitw testing for human disease. Given what is already known about the heterogeneity of these disorders, this process is likely to be complex, and it is not yet possible to say how many combinations of genes and antigens are likely to result in some sort of chronic intestinal inflammation or extraintestinal disease. Nevertheless, these investigations are crucial to developing phenotypically defined disease groups.
Translational research will lead to some very exciting discoveries, based on the 'omics', genomics, proteinomics and phenomics. Genomic investiga- tions of genes and their function are bringing about a revolution in our understanding of the molecular mechanisms of disease, including the complex inter- play of genetic and environmental factors. Genomics is also stimulating the discovery of breakthrough pharmaceuticals by revealing hundreds of new bio- logical targets for the development of drugs, and by giving scientists innovative ways to design new drugs, vaccines and diagnostics. Genomics-based thera- peutics include 'traditional' small chemical drugs, protein drugs, and potentially gene therapy.
Proteinomics refers to the study of proteins expressed by genes and understanding the role of every human protein. By manipulation of these proteins the course of the disease process can be altered and potentially the expression of disease can be eliminated altogether.
Phenomics is the study of the result or sequelae of a gene's function, i.e. the disease manifestation. Phe- nomics is the ultimate point of tomorrow's drug discovery effort as it allows researchers to rapidly validate genes as potential targets for drug discovery and point them to the right individuals. At present, researchers use genomic 'knockout' animal models to determine the biological function of a gene, and then compare the function to that of humans. As diseases are related to subtle changes in protein activity, rather than to complete loss or gain in any one protein, researchers must produce varying changes in protein activity to determine a gene's relation to the specific disease process.
Through these efforts we can meet our challenge to refine patient subsets, not strictly from clinical points of view. There is a tendency among clinicians to 'lump' patients into groups based on certain clinical aspects of their disease; e.g. disease in the ileum.
Nevertheless, not all patients with ileal disease are the same. Simultaneously we must refrain from 'splitting' so many subtypes so as not to lose sight of 'the forest for the trees.' To determine the pheno- types, the genetic and immunologic origins and clinical significance of the phenotypes requires a very certain type of investigator the translationalist. We need to encourage a cadre of educated and creative translational investigators. Such individuals will have in-depth knowledge of disease pathogenesis combined with medical training and education in clinical investigation, and with these tools, they will bring about our expectations for the 21st century (Table 1).
Too many targets, too few patients
Estimates have placed the number of individuals
affected with IBD at as high as two million. The
actual pool of patients willing and able to participate
in clinical trials is much, much smaller. A primary
goal of the translationalists, then, must be the devel-
opment of 'reagent-grade' patients. Well-defined
populations will permit highly specific eligibility
criteria for study participation. It is our challenge to
accelerate our knowledge of these diseases without
wasting our precious patient resources. In the 1990s
we were successful in our efforts to capture the
interest of the biotechnology and pharmaceutical
industries to encourage them to focus on developing
new and improved treatments for IBD. This success
has strained our resources and there are now not
Table 1. IBD research in the 21st century
Effort Result
1. Intercommunication of rodent and human genetic research findings
2. Definition of IBD relevant bacterial antigens, i.e. commensal associated molecular patterns (CAMPs)
3. Prioritization of clinical trials via cooperation between academic institutions and the pharmaceutical industry
4. Definition of 'reagent grade' patients
5. Crossbreeding of relevant rodent models to better approximate human IBD subtypes
Identification of genetic mutations and interactions with relevance to human diseases
Definition of environmental triggers
Rapid answers, targeted therapeutic efficacy
Customization of diagnosis and prognosis and therefore, therapeutic targets and treatment plans
Acceleration of the progress of 1-4
