List of Abbreviations

Introduction

Stricture formation is a significant complication of inflammatory bowel disease (IBD) often requiring endoscopic or surgical intervention. Up to 30% of Crohn’s disease (CD) patients and 5% of ulcerative colitis (UC) patients develop stricturing disease. Strictures are thought to be formed through a combination of inflammatory and fibrotic processes. Although strictures can arise anywhere in the gastrointestinal tract, the most commonly affected locations are the terminal ileum and ileocecal valve because of the fact that the majority of strictures are found in CD patients. Although less common, colonic strictures in UC are typically asymptomatic, benign, and related to degree of inflammation and fibrosis; however, given their location, malignancy must be evaluated for, and factors related to increased likelihood of malignancy include longstanding disease (over 20 years), location proximal to the splenic flexure, and symptomatic bowel obstruction. Because the wealth of information on stricture formation in IBD derives from CD models and studies, the remainder of this chapter will mainly focus on strictures in the context of CD pathophysiology.

Risk Factors

The natural history of CD often begins with a diagnosis of inflammatory disease, with development of fibrostenotic and/or penetrating pathologies over time. Approximately 20%–30% of CD patients are classified into the fibrostenotic phenotype. There exists a variety of clinical, genetic, epigenetic, and serologic risk factors associated with fibrostenotic CD ( Table 3.1 ).

Overall, these risk factors are associated with a complicated disease course rather than stricture formation specifically; nonetheless, discussion of these risk factors is warranted due to their potential importance in identifying patients with, or who could potentially develop, fibrostenotic disease.

The clinical risk factors associated with fibrostenotic stricture development are generally related to disease duration and severity. Diagnosis under the age of 40 years, perianal disease, and need for steroids during the first flare are three of the strongest risk factors that reliably predict severe disease course, and if two of these three criteria are fulfilled, the positive predictive value for disabling disease in the future exceeds 90%. Deep mucosal ulceration reliably puts patients at higher risk of subsequent surgery. Small bowel inflammation also predicts development of clinically significant stricturing disease that may require surgery, primarily owing to the decreased luminal diameter of the small intestine. Smoking tobacco is also a well-known risk factor for worsening CD and has been shown to accelerate time from diagnosis to first stricture. Finally the presence of fistulae has a high-positive predictive value for the presence of strictures.

Several genetic risk factors also exist that may predict stricturing disease. The diversity of function found in these genetic factors is revelatory of the multifaceted nature of stricture pathogenesis. The intracellular bacterial sensor nucleotide-binding oligomerization domain-containing protein 2/caspase recruitment domain-containing protein 15 (NOD2/CARD15) is by far the most widely studied gene in regards to IBD, and certain genetic variations have been shown to be associated with ileal involvement and stricturing disease requiring surgery. Several gene polymorphisms regulating fibrosis have been linked independently with stricturing disease, chiefly among them the 5T5T polymorphism of matrix metalloproteinase–3 (MMP-3) and homozygosity for the rs1363670 G-allele near interleukin-12B (IL-12B). Polymorphisms related to the pro-fibrotic transforming growth factor-β pathway (TGFβ) are gaining increasing attention, although the exact roles of these polymorphisms remain unclear. The molecular pathway of TGFβ signaling will be explored further in the chapter.

Epigenetic risk factors that regulate gene expression are also proving to be vital in understanding fibrostenotic disease. DNA methylation is a process that typically represses gene expression by interfering with the binding of transcription factors to their DNA binding sites or recruiting methyl-CpG-binding proteins that attract histone and chromatin-modifying enzymes. Significant DNA methylation has been found at sites involving IL-27, IL-19, tumor necrosis factor (TNF), soluble latent membrane-type 1 (SMT1), and NOD2, all of which can predict disease activity. DNA methylation has also been found in the vacuole membrane protein-1 (VMP-1) locus, a site that also contains the microRNA-21 (miRNA-21). Differential transcription favoring expression of miRNA-21 has been shown to favor sustained TGFβ signaling, leading to production of collagen, extracellular matrix (ECM), and ultimately fibrosis. Other miRNAs that regulate gene transcription of fibrotic factors have been described: the antifibrotic miRNA-29 and miR17∼92 cluster among them.

The presence of an intestinal microbiota also plays a crucial role in stricture formation, as a dysregulated response to gut microbes is key to the pathogenesis of IBD. Antibodies directed against certain microbes have been increasingly used as indicators of potentially severe disease course. The first discovered and most well-known antibodies are the anti- Saccharomyces cerevisiae antibodies, which have been used to distinguish CD patients from UC patients, and have been associated with complicated disease course (fibrostenotic and penetrating), ileal and ileocolonic involvement, early disease onset, perianal disease, and surgical requirements. Other antibodies have been described which also predict complicated disease course; however, because they do not reliably predict fibrostenotic over penetrating phenotypes, they currently have limited clinical utility ( Table 3.1 ).

Pathogenesis

As stated earlier, stricture formation in IBD is a complex process involving a wide variety of cellular and molecular factors. It is widely accepted that IBD pathogenesis overall involves a dysregulated immune response to gut flora, triggering a variety of leukocytes to release inflammatory cytokines and factors. Inflammation in turn causes disordered tissue repair mechanisms, leading to a final common pathway involving mesenchymal cells: their proliferation, activation, deposition of ECM, decreased secretion of ECM-degrading MMPs, increased secretion of MMP-inhibiting tissue inhibitors of metalloproteinases (TIMPs), and ultimately pathologic fibrosis ( Fig. 3.1 ).

Mechanisms of fibrosis in inflammatory bowel disease.

TGFβ is perhaps the most well-known growth factor associated with wound healing and fibrosis. Elevated levels of TGFβ have been associated with fibrotic disease in a wide variety of organ systems, including heart, lungs, liver, kidneys, skin, and digestive tract. The primary functions of TGFβ in the context of fibrosis include activation and differentiation of fibroblasts to myofibroblasts, upregulating production of ECM proteins collagen and fibronectin, expression of adhesive receptors and contractile elements, inhibition of MMPs, and stimulation of TIMPs. Moreover, TGFβ has been found to stimulate epithelial-to-mesenchymal transition (EMT).

At the molecular level, there are three isoforms of TGFβ that can bind to receptors TGFβR1 or TGFβR2. These receptors form heterodimeric or homodimeric transmembrane serine/threonine kinase complexes that phosphorylate Smad2 and Smad3. The phosphorylated Smad2 and Smad3 complex with Smad4, and this larger complex translocates to the nucleus to activate gene transcription, ultimately with pro-fibrotic effects ( Fig. 3.2 ).

TGFβ signaling.

Regulation of Smad2 and Smad3 occurs via Smad7, which prevents binding of the Smad2/3 complex to its receptor complex. Knowledge of the molecular mechanisms of TGFβ signaling has proven crucial in the development of potential specific antifibrotic therapies for IBD, more of which will be detailed further in the chapter.

Before ending discussion on TGFβ, it must be noted that this growth factor has significant functions outside of fibrogenesis and wound healing. The immunoregulatory functions of TGFβ are well described, particularly with regards to immune tolerance, innate immunity, and adaptive immunity. The fact that TGFβ has such a wide array of effects will have significant implications in regards to targeting this molecule for specific antifibrotic therapy.

TNF-α is perhaps the most well-studied cytokine in regards to inflammation in IBD; however, its role in fibrostenosis is less clear. TNF-α has been shown to have pro-fibrotic effects, such as induction of TIMP-1 expression, reduction in MMP-2 activity, and reduction in collagen degradation. Somewhat paradoxically, TNF-α has also been found to have antifibrotic effects as well, for example, reducing collagen deposition and TGFβ through NFKB (nuclear factor kappa-light-chain-enhancer of activated B cells) induction of Smad7. These differential effects may be dependent on the specific cell-type and TNF receptor engaged, as TNFR1-deficient mice have shown accelerated wound healing and increased collagen deposition and angiogenesis at wound sites, whereas TNFR2-deficient intestinal myofibroblasts show decreased collagen synthesis and cell proliferation. The multiple cellular effects of TNF-α will have significant implications on the use of anti-TNF biologics on fibrostenotic disease.

Related to TNFα is TNF-like ligand 1A (TL1A), a protein encoded by the TNFRSF15 gene. The immunologic function of TL1A has been well described, in particular its ability to bind to death domain receptor 3 (DR3), thereby promoting Th17 cell expansion and cytokine production in immune responses. More recently however, the pro-fibrotic effects of the TL1A–DR3 pathway have come to light. DR3 has been found to be expressed on both human and murine intestinal fibroblasts. DR3-deficient mice have displayed decreased numbers of colonic fibroblasts, reduced fibroblast activation, and reduced TL1A-induced collagen deposition. Conversely, TL1A overexpression in mice leads to spontaneous ileitis with increased collagen deposition. Perhaps most strikingly, a TNFRSF15 haplotype that increases TL1A expression has been described in humans, and this population exhibits increased risk of CD, intestinal fibrostenosis, and greater need for surgery. TL1A blockade is a potentially exciting modality of treating fibrostenotic IBD, and present studies on this treatment modality will be described later in the chapter.

A number of cytokines released by certain T cell subpopulations have also been heavily implicated in the development of fibrostenotic strictures. Th2 cytokines, in particular IL-4 and IL-13, are both elevated in fibrostenotic disease and promote fibroblast activation, fibroblast proliferation, and collagen synthesis. IL-17, secreted by Th17 cells, is a well-known pro-inflammatory cytokine that not only activates granulocytes but also stimulates human colonic myofibroblast activation. Moreover, in vitro intestinal samples from patients with fibrostenotic CD express higher levels of IL-17A, and IL-17-stimulated myofibroblasts isolated from fibrostenotic strictures generate more collagen and TIMP-1.

Conversely, there are T cell subpopulations whose activities are ultimately antifibrotic in nature. Th1 cells, for example, secrete interferon-γ (IFNγ), which has been shown to inhibit fibroblast migration and proliferation through suppression of TGFβ production. In addition, regulatory T cells (Tregs) secrete IL-10, which has been found to inhibit fibrosis in liver, lung, and kidney pathologies. Although the fact that IL-10-knockout mice are routinely used in murine models of colitis, along with the fact that certain IL-10 polymorphisms are associated with IBD, certainly implicates IL-10 in IBD pathogenesis, the utility of IL-10 blockade has unfortunately not borne out in clinical trials.

As stated earlier, the final common pathway for the myriad cytokines and factors described previously involves the recruitment and activation of mesenchymal cells, mainly fibroblasts, myofibroblasts, and smooth muscle cells. In the normal homeostatic state, fibroblasts continuously secrete ECM to maintain the baseline structural integrity of tissues. Additionally, there exists a balance between ECM-degrading MMPs and their inhibitors, known as TIMPs. In response to inflammation however, mesenchymal cells rapidly proliferate, follow chemotactic factors to home into sites of injury, and secrete increased amounts of ECM proteins in an attempt at tissue repair. Moreover, the balance between MMPs and TIMPs shifts to favor TIMPs, ultimately leading to increased ECM deposition. Myofibroblasts and smooth muscle cells also have the ability to contract the wound area with the aim of limiting the area of tissue damage. Although the mesenchymal cell activities just described are physiological in the context of acute minor injury, they can become grossly disordered in the setting of chronic inflammation.

Furthermore, a process called EMT has also been heavily implicated in the development of both IBD-associated strictures and fistulae. EMT is normally a physiological process, especially in the context of embryogenesis and organ formation. However, in response to injury and inflammation, epithelial cells can lose their key characteristics such as apicobasal polarity and cell–cell contact and come to function like myofibroblasts, secreting ECM, TIMPs, and contracting areas of injury. At the molecular level, cytokines and growth factors such as TNF, TGFβ, and IL-13 have been found to be drivers of EMT, ultimately causing nuclear localization of β-catenin and of the related transcription factors SNAIL and SLUG. Resulting events include the downregulation of epithelial-specific markers such as E-cadherin, and the upregulation of mesenchymal markers like vimentin. Evidence for the involvement of EMT in IBD-associated strictures was found via immunohistochemical staining of fibrotic colon samples from CD patients, which discovered cell populations with colocalization of several epithelial and mesenchymal markers, increased expression of TGFβ, and nuclear localization of β-catenin and SNAIL transcription factors. Of note, EMT is not only implicated in fibrostenosis but tumorigenesis as well, making this process a highly attractive target for future therapies.

The end result of both mesenchymal cell activity and EMT is the deposition of ECM and contraction of damaged areas with the aim of tissue repair ( Fig. 3.1 ). As stated earlier, in more acute and minor insults, this is an appropriate physiological response that can lead to proper wound healing. However, with increasing severity and chronicity of injury, such as the chronic inflammation seen in IBD, these tissue repair mechanisms can perpetuate, ultimately leading to clinically significant fibrosis and stricturing disease in the gastrointestinal tract.