Critical Reviews in Clinical Laboratory Sciences, 38(1):33–108 (2001)
`
`Recent Advances in Inflammatory
`Bowel Disease
`Richard J. Farrell,* Subhas Banerjee, and Mark A. Peppercorn
`Divison of Gastroenterology, Beth Israel Deaconess Medical Center, Harvard
`Medical School, Boston, Massachusetts
`
`Referee: Fabio Cominelli, University of Virginia, Charlottesville
`
`*
`
`Correspondence to be addressed to: Richard J Farrell, MD, Department of Gastroenterology,
`Beth Israel Deaconess Medical Center, Harvard Medical School, Dana 501, 330 Brookline
`Avenue, Boston, MA 02115
`
`ABSTRACT: The last decade has seen tremendous advances in our knowledge, which has led
`to genuine improvements in our understanding of the pathogenesis and management of
`inflammatory bowel disease (IBD). The combined power of cellular and molecular biology has
`begun to unveil the enigmas of IBD, and, consequently, substantial gains have been made in
`the treatment of IBD. Refinements in drug formulation have provided the ability to target
`distinct sites of delivery, while enhancing the safety and efficacy of older agents. Simultaneous
`progress in biotechnology has fostered the development of new agents that strategically target
`pivotal processes in disease pathogenesis. This article addresses our current understanding of
`the pathogenesis of IBD, including the latest developments in animal models and covers agents
`currently used in the treatment of IBD as well as emerging therapies.
`
`KEY WORDS: inflammatory bowel disease, Crohn’s disease, ulcerative colitis, pathogenesis,
`animal models, medical therapy.
`
`I. INTRODUCTION
`
`Over the last decade, substantial gains have been made in the treatment of
`inflammatory bowel disease (IBD). Despite expanding knowledge of cellular
`and molecular mechanisms of intestinal inflammation, the etiology and patho-
`genesis of IBD remain obscure. Consequently, approaches to the diagnosis and
`management of both Crohn’s disease (CD) and ulcerative colitis (UC) remains
`largely empiric. However, progress in biotechnology has fostered the develop-
`ment of new agents that strategically target pivotal processes in disease patho-
`genesis. This review addresses our current understanding of the pathogenesis
`of IBD, including the latest developments in animal models and covers agents
`currently used in the treatment of IBD as well as emerging therapies.
`
`1040-8363/01/$.50
`© 2001 by CRC Press LLC
`
`33
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`SECTION 1: PATHOGENESIS OF INFLAMMATORY
`BOWEL DISEASE
`
`No single agent or distinct mechanism explains all aspects of IBD, and
`several distinguishing factors are likely to result in either CD or UC (Table 1).
`The pathogenesis of CD and UC involves an interplay of environmental,
`genetic, microbial, immune and nonimmune factors, which, in combination,
`result in chronic intestinal inflammation (Figure 1). These factors are dis-
`cussed below.
`
`II. ENVIRONMENTAL FACTORS
`
`Although the continuous rise in the incidence and prevalence of IBD
`throughout most of the last 50 years highlights the important role played by the
`environment in its pathogenesis, the link between the environment and IBD is
`still circumstantial. The relative importance of environmental compared with
`genetic factors in the initiation or progression of UC and CD remains unclear
`and controversial. In a long list of potentially relevant environmental factors,
`smoking, childhood infections, oral contraceptives, diet, prenatal events,
`breastfeeding, hygiene, microbial agents, occupation, pollution, and stress are
`the most commonly cited.
`
`A. Smoking
`
`Of all environmental factors, the opposite effect of cigarette smoking in
`IBD, beneficial in UC and detrimental in CD, remains the most puzzling, the
`most extensively studied, but also the most reproducible.1-3 The studies fol-
`lowed the unusual finding by Harries et al.4 that smoking seemed to protect
`against UC, with the risk of developing UC among current smokers only 40%
`that of nonsmokers. In general, the protective effect was similar in men and
`women.4 In contrast, several studies have demonstrated that smoking increases
`the risk for CD by over twofold.5,6 Sutherland et al.7 found that women
`smokers were greater than four times more likely to develop CD recurrence
`than nonsmokers, while male smokers had a slightly increased risk as well. A
`recent investigation of a large cohort of patients with CD has confirmed that
`smoking is an independent risk factor for recurrence, particularly following
`medically induced remission.8 The mechanism of the different impact of
`smoking on CD and UC still remains a mystery. Changes in colonic mucus
`have been proposed as one possibility. Following the initial observation that
`
`34
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`35
`
`Distingushing Features of UC and CD Pathogenesis
`TABLE 1
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`FIGURE 1. Diagram showing the components and events involved in IBD pathogenesis.
`
`36
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`colonic mucus in UC was both quantitatively and qualitatiively abnormal,9
`Cope et al.10 demonstrated that mucus production in UC patients who smoked
`was similar to that of healthy controls, suggesting that restoration of a normal
`pattern of mucus production, as a consequence of smoking may protect against
`disease. However, subsequent studies found that the composition of colonic
`mucus in UC and controls was similar.11 While it appears not to be related to
`an effect on intestinal permeability,12 one possible mechanism is that the
`thrombogenic properties of cigarette smoking may lead to vascular changes
`that would predispose susceptible persons to CD. Cytokines have also been
`proposed as part of the explanation of the peculiar effects of smoking in IBD.
`Colonic mucosal concentrations of proinflammatory cytokines interleukin
`(IL)-1β and IL-8 have been found to be significantly reduced in smokers with
`UC compared with nonsmokers with UC in contrast to higher levels of IL-8
`in healthy controls who smoke compared with nonsmokers. Regarding smok-
`ers with CD, only IL-8 levels were reduced compared to nonsmokers with CD,
`with no significant reduction in IL-1β levels.13 As discussed later in this
`review, the cytokine profile in CD is of a dominant Th1 pattern whereas in UC
`a Th2 cytokine pattern, predominates. Madretsma et al.14 have demonstrated
`that nicotine has an inhibitory effect in vivo on Th2 mononuclear cell function
`as measured by the inhibition of IL-10 production, but does not appear to have
`any effect on Th1 cell function as measured by IL-2, and tumor necrosis factor-
`alpha (TNFα) production.
`
`B. Diet
`
`A possible relation between diet and IBD is a logical link because inflam-
`mation affects the very site of nutrient absorption. To date, however, no
`specific food has been clearly linked to IBD. Although the consumption of
`cola drinks, chewing gum, and chocolate recently has been associated with
`CD, and cola drinks and chocolate with UC, all of these substances may simply
`be the expression of a “modern lifestyle” that involves more relevant risk
`factors that are presently unknown. Nutritional deficiencies such as zinc and
`selenium are common in CD, and some authors suggest that these deficiencies
`contribute to immunological dysfunction.15,16 While there have been a number
`of uncontrolled studies assessing the efficacy of elemental diets in CD that
`suggested some degree of clinical improvement, most controlled studies have
`found that enteral nutrition is less effective than steroids and
`aminosalicylates.17,18 Meta-analyses found that the overall remission rates for
`diet- and steroid-treated patients were 58 and 80%, respectively.19,20 While
`some data suggest that elemental diet may improve CD by reducing intestinal
`
`37
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`permeability,21 it remains unclear why nutritional therapies improve CD but
`not UC.
`
`C. Drugs
`
`Research on IBD risk among oral contraceptive users has reached conflict-
`ing conclusions. A population-based case-control study found that women
`who reported oral contraceptive use within 6 months before disease onset were
`at increased risk for both diseases compared with those who never used oral
`contraceptives (relative risk (RR) of UC = 2.0; RR of CD = 2.6). While women
`who had used oral contraceptives for more than 6 years had the highest risk of
`CD (RR = 5.1), increasing duration of use was not associated with increased
`risk of UC. Adjustment for race, smoking, income, or pregnancy history did
`not substantially alter these results.22
`Anecdotal, rechallenge, and limited epidemiological data indicate that
`NSAIDs can cause exacerbations of preexisting IBD: indeed these drugs can
`occasionally induce a colitis de novo.23-25 Recent studies indicate that fecal pH
`is more acidic in patients with IBD than in healthy controls. Fecal pH is
`reduced in the indomethacin model of chronic ileitis in rats, which is similar
`in some respects to human IBD.26 In this model, indomethacin significantly
`changed microcirculation parameters causing a twofold increase in leukocyte
`adherence and a threefold increase in leukocyte emigration, thus providing a
`possible pathophysiological mechanism for the observations mentioned above.
`Demling observed that the increase in CD that has taken place over the last
`50 years paralleled the increasing use of antibiotics in human and veterinary
`medicine. He has postulated that insufficient doses of antibiotics can induce a
`capacity for toxin production in bacteria, or can make them invasive. It is
`conceivable that CD is caused in genetically susceptible persons by intestinal
`bacteria whose biological (but not morphological) properties have been altered
`by antibiotics.27
`
`D. Childhood Factors
`
`Epidemiological studies have found an increased frequency of childhood
`infections and tonsillectomies in patients with IBD. The ileum is the most
`prevalent location of disease in CD patients with previous tonsillectomy.28 In
`a Belgium study, it was found that absence of appendectomy was closely
`related to the development of UC. Before the onset of UC, only one of the 174
`patients (0.6%) had undergone an appendectomy compared with 41 of 161
`
`38
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`(25.4%) of age- and sex-matched controls. The difference between the two
`groups was highly significant with an odds ratio of 59.1 compared with a
`relative risk of getting the disease in non-smokers of 2.95.29 While this overall
`protective role of appendectomy for UC has been confirmed in several studies,
`observations in CD suggest that appendectomy in some cases was a result of
`still undiagnosed CD.30 One possible interpretation of this finding may be that
`an infectious agent or a related antigen is present in the appendix and may be
`involved in the pathogenesis of UC in the predisposed individual.
`One American study found that compared with neighbor controls, patients
`with CD were more likely to report an increased frequency of childhood
`infections in general (odds ratio 4.67) and pharyngitis specifically (odds ratio
`2.14), while patients with UC reported an excess of infections generally (odds
`ratio 2.37).31 A large international cooperative study found very few statisti-
`cally significant differences in childhood factors between index and control
`families.32 Except for increased incidence of childhood eczema, no other
`differences were found with respect to a multitude of factors traditionally
`linked to IBD, including breastfeeding, childhood gastroenteritis, sugar added
`to milk in infancy, cereal consumption, and major stressful life events.
`
`E. Stress
`
`As recently as 1994, a significant proportion of 247 women with IBD
`thought that they may have acquired the disease due to stress and emotional
`issues (UC 40% and CD 29%) and internalizing issues (2.3%).33 Substantial
`evidence has been found for an association between stress and increased
`illness behavior and possibly between stress and infectious pathology. Intro-
`verts, isolates, and persons lacking social skills may be at an increased risk for
`both illness behaviors and pathology.34 There is substantial evidence for a
`relation between stress and decreases in functional immune measures (pro-
`liferative response to mitogens and natural killer activity), percent of circulat-
`ing white blood cells, immunoglobulin levels, and antibody titers to herpesvi-
`ruses.35 Therefore, there are ample stress-related factors that may contribute to
`both the pathogenesis and expression of IBD, but future controlled studies are
`required.
`
`F. Vascular Factors
`
`In a prospective study by Wakefield et al.,36 specimens of resected small
`and large intestine from patients with IBD, and CD in particular, were exam-
`
`39
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`ined. A number of pathogenetic sequences of events were seen, such as
`vascular injury, focal arteritis, fibrin deposition, and arterial occlusion, mainly
`at the level of the muscularis propria, followed by tissue infarction or
`neovascularisation. It has been suggested that this pathogenetic process may
`be compatible with many of the clinical features seen in IBD patients. Further-
`more, Wakefield et al.37 have shown that 85% of identified granulomas in CD
`have a vascular localization. Vascular localization of granulomatous inflam-
`mation suggests that the intestinal microvasculature contains an early element
`in the pathogenesis of CD. These findings may have important implications for
`the identificaton of the primary cause of IBD and advances in clinical manage-
`ment. In an attempt at a unifying ‘vascular theory’, Wakefield and colleagues38
`have proposed that smoking, the oral contraceptive pill, and measles virus may
`potentiate a tendency for focal thrombosis and hence contribute to the patho-
`genesis of CD.
`
`III. FAMILIAL FACTORS
`
`A. Family Studies
`
`Multiple epidemiologic studies have documented significant ethnic and
`familial predisposition to IBD. One of the most consistent observations in
`studies of IBD populations is the high incidence of CD and UC among family
`members of the index case.39 The frequency of IBD in first-degree family
`members may be as high as 40%.40 The basis of this genetic predisposition
`remains obscure — families tend to live in the same place so that environmental
`factors can still play a role in familial aggregation. Concordance rates in twins
`provide probably the strongest evidence that both genetic and environmental
`factors are important in IBD pathogenesis. Tysk and colleagues41 performed a
`detailed analysis of the Swedish registry of 80 twin pairs in whom at least one
`twin was known to suffer from IBD. A higher rate of concordance was found for
`monozygotic twins with CD or UC compared with dizygotic twins, with the
`genetic contribution for CD considerably greater than that for UC. However,
`distinguishing between the genetic and environmental contributions to IBD
`pathogenesis has been difficult. In large studies performed in Scandinavian
`countries, a 10-fold increase in familial risk was interpreted to be strongly
`suggestive of a genetic cause,42 while seasonality of the cohort effect, with
`relapse rates in UC peaking from August to January, combined with urban
`preponderance was interpreted as pointing to environmental factors.43 However,
`the relatively high concordance of clinical characteristics between the index case
`and family members with respect to age, site of inflammation, and type of
`
`40
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`clinical manifestations,44 coupled with the concept of ‘genetic anticipation’,
`defined by a progressively earlier disease onset in successive generations further
`supports the relevance of genes in IBD pathogenesis.45 Other potential familial
`factors explored include an abnormal immune response, exemplified by abnor-
`mal activation of complement via the alternate pathway in both patients with
`IBD and their relatives,46 and a high prevalence of antibodies to intestinal
`epithelial antigens in healthy first-degree relatives of IBD patients.47
`
`B. Intestinal Permeability
`
`Increased intestinal permeability has also been invoked as a pathogenic
`factor for CD with a postulated increased uptake of antigenic material by the
`mucosal immune system that initiates or perpetuates inflammation. Hollander
`and colleagues48 were the first to demonstrate increased intestinal permeability
`in healthy first-degree relatives of CD patiemts. More recently, it has been
`shown that in patients with CD, increased intestinal permeability apparently
`predates clinical relapse and can normalize after surgery, implying that it may
`simply be a phenomenon secondary to inflammation.49 Nevertheless, it re-
`mains to be established whether the defect in relatives is genetically predeter-
`mined and independent of intestinal inflammation or represents subclinical CD
`with asymptomatic intestinal inflammation.
`
`IV. GENETIC FACTORS
`
`It is generally accepted that genetic factors play a significant role in IBD
`pathogenesis.50 The epidemiological data presented above are not consistent
`with a simple Mendelian pattern of inheritance for either CD or UC and a
`polygenic model of disease inheritance has been proposed. McConnell and
`Vadheim51 initially suggested that 10 to 15 genes may account for susceptibil-
`ity to IBD as well as account for the relationship between CD and UC. He
`suggested that an individual with a complete genotype would be more likely
`to develop CD than an individual with fewer genes, who would be predisposed
`to UC. Current thinking, based on the epidemiological data, the clinical
`variability of presentation of IBD, and the results of genetic marker studies has
`refined this model further to suggest that IBD is a number of complex poly-
`genic diseases, each sharing some but not all susceptibility genes. This com-
`plex model incorporates both polygenic inheritance and genetic heterogeneity
`and would explain not only the variability of clinical presentation but also
`ethnic differences in disease prevalence.52
`
`41
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`Two complementary strategies have been used to identify genes involved
`in determining susceptibility to disease and disease behavior. The candidate
`gene directed strategy presumes an understanding of disease pathophysiology
`and uses either population-based association studies, family-based studies
`such as linkage analysis, or the transmission disequilibrium test. By contrast,
`the relatively new technique of genome wide scanning involves a systematic
`analysis of the entire human genome, and is directed at identifying the chro-
`mosomal location of susceptibility loci. Genome-wide scanning currently uses
`family-based linkage analysis for ‘coarse’ mapping and both population-based
`studies and larger family studies for ‘finer’ mapping.
`
`A. Candidate Gene Studies
`
`Candidate gene studies have investigated the contribution of genes in-
`volved in the regulation of the immune response, particularly the major histo-
`compatibility complex (MHC), which is situated in the human leucocyte
`antigen (HLA) region in man. Data available so far provide evidence for
`considerable heterogeneity, not only between CD and UC, but also within both
`CD and UC, with heterogeneity between different populations a further con-
`founding factor. The evidence for a contribution of the HLA region to genetic
`susceptibility in IBD is most strong in UC. Significant linkage between the
`HLA Class II region and UC has been demonstrated in European Caucasian
`families.53 Allelic associations have been reported in Japanese, North Ameri-
`can, and European studies. Allelic association studies are most consistent in
`the Japanese population, where the HLA-DRB1*1502 allele (DR2) is associ-
`ated with susceptibility to UC, and alleles representing DR4 are associated
`with decreased susceptibility to UC.54,55 While the negative association with
`DR4 has been reported widely in both Jewish and non-Jewish Caucasian
`populations, the positive association with DR2 has been demonstrated in
`Californian patients (a mixed Jewish/non-Jewish population),56 but not in the
`majority of European studies (mostly non-Jewish). Population-based associa-
`tion studies suggest that the relatively rare HLA class II gene DRBl*0103 may
`be predictive of severe UC disease behavior, particularly with the need for
`colectomy.53,57 Further data from Caucasian populations also suggest that the
`HLA Class II genotype may influence disease extent and the presence of
`extraintestinal manifestations, particularly arthropathy.58
`The contribution of genes encoding cytokines or cytokine antagonists have
`also been studied by both association studies and linkage analysis. While
`strong associations have been reported in Jewish populations between poly-
`morphisms of the gene encoding the interleukin 1 receptor antagonist (IL-
`
`42
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`1RA) and susceptibility to UC,59 data from non-Jewish Caucasian populations
`are contradictory.60 Similarly, although association between susceptibility to
`CD and TNFα polymorphisms are reported in Californian populations,61
`northern European studies found little evidence for this association. Genes
`encoding IL-2, IL-6, and IL-10 are also currently being studied. There is also
`particular interest in the contribution of genes encoding intestinal mucins. In
`total, nine mucin genes have been characterized, designated MUC1 to MUC9.
`Recent molecular genotyping data have provided preliminary evidence that
`inherited polymorphisms of MUC2 and MUC3 may be involved in disease
`pathogenesis.62 Adhesion molecules,63 trefoil peptides, and proteins involved
`in intracellular signaling are other candidate genes that have also been studied.
`
`B. Genome-Wide Scanning Studies
`
`The application of genome-wide scanning in IBD has arguably been
`more successful to date than many other complex disorders. Hugot et al.64
`reported the first genome screening in CD patients. The genome was system-
`atically screened using polymorphic microsatellite markers distributed on all
`autosomes. Using nonparametric linkage analysis, they identified a suscep-
`tibility locus in the pericentromeric region of chromosome 16, designated
`IBD1 by the authors. Putative candidate genes within this region include the
`gene encoding the IL-4 receptor and CD11 integrin. This locus, which
`appears to be important in CD but not in UC, has been corroborated by many
`investigators throughout the world.65,66 Subsequently, Satsangi et al.67 re-
`ported the results of a genome-wide scan involving 186 affected sibling pairs
`with IBD. The data provide the strongest evidence to date that CD and UC
`are related polygenic disorders. They demonstrated linkage between IBD
`overall and regions on chromosome 12, 7, and 3. Linkage with CD was noted
`with the pericentromeric region on chromosome 16, while linkage with UC
`was noted with individual markers on chromosome 2 and 6 (HLA region).
`While the linkage between susceptibility to IBD and the region on chromo-
`some 12 has been replicated in North America and Europe, recently Rioux
`et al.68 found no evidence for linkage between IBD and selected susceptibil-
`ity loci on chromosome 3, 7, 12, or 16.
`The genetics of IBD has become an area of considerable active research,
`with novel linkages continuing to be reported.69 In view of the rapid pace of
`the Human Genome Project, we should soon learn more about specific genes
`at these putative susceptibility loci and their role in IBD pathogenesis. How-
`ever, the most critical factor remains the requirement for large numbers of
`multiply affected families and affected individuals for study. It is estimated
`
`43
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`that as many as 700 sibling pairs may be required to allow fine mapping of a
`putative susceptibility locus that contributes a sibling relative risk (λs) of only
`2. Collaboration between investigators and support from biotechnology com-
`panies are helping to overcome many of these obstacles.
`
`V. MICROBIAL FACTORS
`
`The dilemma remains: germs or genes, or a combination of the two
`with excessive immune responses to gut-associated bacteria. Because
`both CD and UC closely mimic defined intestinal infections and occur in
`areas of highest luminal bacterial concentrations, investigators have long
`searched for a transmissable agent responsible for IBD.70 While several
`pathogens, including the measles virus, Mycobacterium paratuberculo-
`sis, and Listeria monocytogenes, have been suggested as etiologic agents
`in CD, the notion that normal enteric flora play a role in initiating or
`maintaning IBD is gaining momentum. Further dissection of the para-
`digm of microbial-mucosal interaction suggests that while some compo-
`nents of the flora may act as noxious agents, others (probiotics) seem to
`have a protective effect.
`
`A. Bacteria
`
`Experimental and clinical data support an ever-expanding role for the
`intestinal flora in the initiation or perpetuation of gut inflammation. New
`studies have shown that in colitic rats common strains of anaerobes induce
`inflammatory mediators, including transforming growth factor (TGF)-β1,
`which leads to intestinal fibrosis.71 Recent work demonstrates that Escheri-
`chia coli isolated from CD ileal mucosa adhere to intestinal cells and synthe-
`size cytotoxins, which may cause disruption of the intestinal barrier.72 Of
`direct clinical relevance to the relation between bacteria and IBD, antibiotics
`continue to be evaluated in the treatment of UC and CD, but results remain
`equivocal. A recent 6-month study of ciprofloxacin in addition to
`5-aminoslicylates and corticosteroids suggested clinical benefit in patients
`with UC, particularly those failing under corticosteroid treatment,73 but more
`studies of this and other antibiotics are needed to assess these promising
`results. Recently, fecal hydrogen sulfide (H2S), a volatile, highly toxic but
`relatively uncommon sulfide product of sulfate-reducing bacteria, has been
`proposed to play an etiologic role in UC. H2S release by UC feces has been
`shown to be three to four times higher compared with normal feces.74 This
`
`44
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`increased H2S may reflect abnormalities of the fecal bacteria and/or sub-
`strate availability.
`
`B. Probiotics
`
`Although most microbiological research in IBD has been focused on
`the harmful effects of bacteria, there is growing interest in the potentially
`beneficial effects of manipulating the normal intestinal flora using probiotics.
`Although there are only very limited data to suggest differences between
`the microbial ecology of UC and CD patients and normal individuals, there
`is increasing support for the concept that some components of the flora
`downregulate inflammation. This regulatory function may be lost or abnor-
`mal in IBD, and probiotic therapy may help restore microbial homeostasis
`in the gut. A probiotic formula has been shown to increase fecal IgA levels
`in infants and perhaps enhance mucosal resistance.75 Venturi et al.76 re-
`cently have reported that a probiotic mixture of lactobacilli, bifidobacteria,
`and streptococci is quite effective in maintaing clinical remission in UC
`patients.
`
`C. Fecal Stream
`
`Initial studies relegated the importance of fecal stream, suggesting that
`if it may play a role in contributing to secondary tissue damage once
`inflammation has commenced.77 However, more recent work has estab-
`lished the increasingly important role of intestinal contents in the patho-
`genesis of IBD. An important clinical observation has been that patients
`with CD with a diverting ileostomy fail to develop recurrent disease until
`reanastomosed.78 Furthermore, infusion of a CD patients’s ileostomy efflu-
`ent into the excluded normal ileal loop of the same patient rapidly induces
`local immune activation and de novo inflammation.79 Some components of
`the fecal stream promote inflammation, but whether they are intrinsically
`noxious or whether CD patients are particularly susceptible to them is
`unknown.
`The most compelling support for a crucial role of normal enteric bacteria in
`IBD comes from studies in experimental animals. Intestinal inflammation is
`absent in animals raised in germ-free conditions, but appears when the gut is
`repopulated with normal flora, particularly anaerobes.80 In one of these animal
`models, the C3H/HeJBir mouse, mucosal T cells proliferate in vitro when
`exposed to cecal bacteria, but not to food or intestinal epithelial cell antigens.81
`
`45
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`Interestingly, these T cells can transfer experimental IBD to other mice, estab-
`lishing a strong link among flora, immune reactivity, and symptoms.
`
`D. Measles Virus
`
`Wakefield et al.37 provided evidence supporting a role for the measles
`virus in the pathogenesis of CD, contending that the underlying pathogenic
`abnormality in CD is granulomatous vasculitis induced by persistent measles
`infection of mesenteric endothelial cells. The same researchers also presented
`epidemiologic data showing an increased risk of CD with perinatal exposure to
`measles virus and of IBD with measles vaccination.82 However, other investiga-
`tors have failed to reproduce their observations. Iizuka et al.83 failed to detect
`measles DNA in the intestines of patients with CD, UC, and other bowel diseases
`using very sensitive polymerase chain reaction primers. Liu et al.84 found that
`Wakefield’s nonaffinity purified polyclonal serum exhibited nonspecific nuclear
`staining, and they were unable to demonstrate any staining of Crohn’s tissues
`with a specific monoclonal antibody. More recent studies found neither molecu-
`lar nor epidemiologic evidence supporting an etiologic or pathogenic link be-
`tween measles virus and CD.85,86
`
`E. Mycobacteria paratuberculosis
`
`Since the original description of regional enteritis, an association between
`CD and mycobacteria has been postulated. After a wave of intense interest raised
`by the finding of Mycobacterium paratuberculosis in some CD tissues but not
`tissues from patients with other conditions,87,88 long-term bacterial cultures and
`sensitive polymerase chain reaction tests have detected M. paratuberculosis in
`a variable and small proportion of patients with CD.70,89 However, other studies
`either failed to detect IS900, a repeating genomic DNA insertion element
`specific for M. paratuberculosis, in CD tissues,90 or showed a complete lack of
`specificity.91 Based on these inconsistent results, only a marginal case can be
`made for a role of M. paratuberculosis in CD. Furthermore, the results of trials
`with antituberculous therapy has been disappointing raising the question of
`whether further investigation of M. paratuberculosis in CD is warranted.92-94
`
`F. Listeria monocytogenes
`
`The most recent putative agent in CD is Listeria monocytogenes.84 Antigen for
`this organism was found in macrophages and giant cells beneath mucosal ulcers,
`
`46
`
`Downloaded by [Reprints Desk Inc] at 06:42 29 August 2017
`
`

`

`adjacent to fistulae, near abscesses, and within the lamina propria, granulomas, and
`mesenteric lymph nodes of 75% of CD patients, compared with 13% of UC
`patients and 0% of control subjects. The significance of these observations remains
`to be determined, because Listeria is a relatively frequent environmental contami-
`nant, and 83% of the patients with positive results had immunohistochemical
`evidence of coexistent Escherichia coli or steptococcal antigen, suggesting non-
`specific secondary invasion of ulcerated tissues by luminal organisms.
`
`VI. IMMUNE FACTORS
`
`A. Immunology of Inflammatory Bowel Disease
`
`UC and CD form the ends of a spectrum of chronic inflammatory bowel
`disorders of the intestine for which no specific pathogen has been found. In the
`absence of an identifiable cause, attention has focused on the host immune
`response, both at mucosal level and systemic levels, in an attempt to unravel
`the immune reactions that engender inflammation and tissue damage.
`Three major hypotheses as to the antigenic triggers in IBD have been
`postulated. One hypothesis is that the antigenic triggers are microbial patho-
`gens that have not yet been identified because of fastidious culture require-
`ments. According to this hypothesis, the immune response in IBD is an
`appropriate but ineffective response to these pathogens. Such a ‘hypoactive’
`immune response could lead to proliferation and persistence of microbial
`agents. The second hypothesis suggests that the antigenic trigger in IBD is
`some common dietary antigen