Inflammatory bowel disease

Authors

  • M Boirivant,

    1. Immune-mediated Disease Section, Department of Infectious, Parasitic and Immune-mediated Disease, Istituto Superiore di Sanità, Roma, Italy
    Search for more papers by this author
  • A Cossu

    1. Immune-mediated Disease Section, Department of Infectious, Parasitic and Immune-mediated Disease, Istituto Superiore di Sanità, Roma, Italy
    Search for more papers by this author

Monica Boirivant, Department of Infectious, Parasitic and Immune-mediated Disease, Istituto Superiore di Sanità, Viale Regina Elena 299, Roma 00161, Italy. Tel: +39 649 902976, Fax: +39 649 902931, E-mail: monica.boirivant@iss.it

Abstract

Oral Diseases (2011) 18, 1–15

This review focuses on the prominent etiological and pathogenetic aspects of inflammatory bowel disease (IBD), with particular attention being paid to the mucosal immune response to commensal micro-organisms in health and disease. Pathogenetic implications for target therapy will also be discussed. The clinical presentation, diagnostic aspects, and currently recommended therapeutic options for the two main types of IBD are also taken into consideration, including manifestations of these conditions in the oral cavity.

Introduction

Inflammatory bowel disease (IBD) encompasses a range of chronic, immune-mediated inflammatory disorders of the gastrointestinal (GI) tract that are usually classified under two major relapsing conditions, Crohn’s disease (CD) and ulcerative colitis (UC) (Baumgart and Sandborn, 2007). Both are complex disorders (Cho and Weaver, 2007), resulting from the interaction of genetic and environmental factors, and, indeed, increasing evidence strongly suggests that they result from an inappropriate inflammatory response to intestinal microbes in a genetically susceptible host (Baumgart and Carding, 2007; Abraham and Cho, 2009).

The incidence of IBD reported in North America ranges from 2.2 to 14.3 cases per 100 000 person-years for UC and from 3.1 to 14.6 cases per 100 000 person-years for CD, and, in Europe, respectively, from 8.7 to 11.8 (UC) and from 3.9 to 7.0 (CD) cases per 100 000 person-years. UC is more prevalent than CD, irrespective of the geographic area considered, with an estimated number of 780 000 individuals having UC and as many as 630 000 suffering from CD in the United States (USA) (Loftus, 2004). The incidence of both diseases appears to follow a bimodal age distribution with signs and symptoms frequently manifesting in early adulthood as well as from 50 to 70 years of age. The incidence of CD and UC continues to rise in low-incidence areas such as southern Europe, Asia, and most developing countries. The observation that geographic variability in incidence parallels the gradient in economic growth and the rise in the pro capite income, together with the change observed in prevalence in some ethnic groups after migration to other geographic areas, further support the theory that environmental factors and lifestyle play a role in disease occurrence (Loftus, 2004). Long before genome-wide analysis studies (Cho and Weaver, 2007), familial clustering of cases and twin studies had already established the role of genetic factors in the occurrence of these disorders, and a positive family history is still the most important independent risk factor (Russel and Satsangi, 2004). Cigarette smoking affects these two diseases in a different way: smokers are at an increased risk of CD and tend to present more severe clinical course whereas former smokers and non-smokers are at greater risk of UC (Cosnes, 2004).

Etiology and pathogenesis

Increasing evidence suggests that IBD results from an inappropriate inflammatory response to intestinal microbes in a genetically susceptible host (Baumgart and Carding, 2007; Abraham and Cho, 2009). Evidence for the validity of this hypothesis comes from the observation that almost all the experimental murine models of mucosal inflammation depend upon the presence of the microflora. In fact, inflammation does not occur if the mice are reared in a germfree environment. This is the case for experimental models due both to defects in T-cell effector function and defects in regulatory T-cell function (Strober et al, 2002). The importance of microflora is also confirmed by findings in humans with IBD. In these patients, the disease can be ameliorated by the administration of antibiotics and is abrogated by avoiding the enteric stream coming into contact with the area of inflammation (D’Haens et al, 1998; Sartor, 2000). This inappropriate inflammatory response can occur as a result of excessive activation and differentiation of immune effector T-cell subsets against harmless antigens (so called type 1 helper T cells [Th1] and Th17 for CD; Th2 and Th17 for UC) (Duchmann et al,1995,1996; Fuss et al, 1996; Reimund et al, 1996; Ahern et al, 2008; Kobayashi et al, 2008) and/or defective counter-regulation by regulatory T cells (Tregs) (Maul et al, 2005; Saruta et al,2007; Barnes and Powrie, 2009).

Intestinal microbiota

An enormous number of micro-organisms are known to colonize the intestine and form complex communities or microbiota. The host-microbiota associations usually evolve into beneficial relationships. Bacteria in the gastrointestinal tract supply key nutrients and prevent colonization by opportunistic pathogens. Furthermore, they contribute to the anatomical development and function of the mucosal immune system (Duerkop et al,2009). In adults, the ‘normal’ microbiota has been shown to contain not only bacterial species that prevent the development of inflammation (Round and Mazmanian, 2010), but also micro-organisms that have been shown to induce inflammation under particular conditions. Therefore, the microbiota has the potential to exert both pro- and anti-inflammatory responses, and the composition of the bacterial communities, in the gut, is related to the correct functioning of the immune system (Round and Mazmanian, 2009). In patients with IBD, depletion and reduced diversity of members of the mucosa-associated phyla Firmicutes and Bacteroidetes have been observed (Frank et al, 2007). Recently, a reduction in the bacteria Faecalibacterium prausnitzii, belonging to the Firmicutes phyla, has been associated with early postoperative recurrence in patients with CD (Sokol et al, 2008), and other studies have shown that a subset of patients with CD harbors a potentially pro-inflammatory ‘adherent-invasive’ strain of Escherichia coli in the small intestine (Pineton de Chambrun et al, 2008). However, whether these modifications contribute to the disorder or merely reflect secondary changes caused by the inflammation remains to be established. Regardless of the primary or aggravating role of the alteration of microbiota in the pathogenesis of IBD, the interactions observed between the microbiota and the mucosal immune system stress the importance of the mucosal response to commensal microorganisms and how genetic factors may affect this response. In ‘non-inflammatory’ conditions, the mucosal immune response to the microbiota is controlled by the cooperation between the innate and adaptive immune response. This cooperation results in a state of ‘controlled inflammation’ defined as tolerogenic response (Coombes and Powrie, 2008).

Mucosal immune response to commensal microorganisms

Cells and their soluble products belonging to the innate response and the epithelial barrier represent the first line of host-luminal bacteria interactions. Specialized cells (globet cells and Paneth cells) within the epithelial layer, which is sealed by the tight junctions, produce mucus and antimicrobial peptides (i.e. defensins) that limit the entry of bacteria into the intestinal mucosa (Swidsinski et al, 2002; Elphick and Mahida, 2005). At the same time, the epithelial cells are able to ‘sense’ microbial components through the pattern recognition receptors (PRRs), which can be located either on both the cell surface and the intracellular compartment [Toll like receptors (TLRs)) or exclusively in the intracellular compartment [nucleotide-binding oligomerization domain (NOD) proteins]. These receptors recognize evolutionary conserved microbial components. The same receptors are present on dendritic cells (DCs) in the lamina propria and, restricted to the expression of some TLRs, on T cells (Iwasaki and Medzhitov, 2004; Strober et al, 2006). These microbial sensors are functionally intersected, as NOD2 may affect TLR signaling (Netea et al, 2004; Watanabe et al, 2004), and they cooperate in the autophagy pathway (Cooney et al, 2010; Homer et al, 2010; Travassos et al, 2010). Autophagy is a cell ‘survival’ mechanism in which cytoplasmic material is broken down, mainly to provide recycled nutrients to cell (energy-sparing function) but also serving to degradation of intracellular bacteria (antibacterial function) and delivery of degradation products (either of microbial or cell host origin) to TLRs and antigen presentation molecules (innate and adaptive immune function) (Levine and Deretic, 2007; Nedjic et al, 2008; Virgin and Levine, 2009). In the steady state, epithelial cells produce mainly molecules and cytokines [epithelial derived thymic stromal lymphopoietin (TSLP), interleukin (IL)-10, IL-25, transforming growth factor-β (TGF-β) and probably retinoic acid (RA)] (Lampen et al, 2000; Rimoldi et al, 2005; Saurer et al, 2007; Artis, 2008; Maynard and Weaver, 2008; Zaph et al, 2008) that promote in the human lamina propria macrophages and DCs [i.e. professional antigen-presenting cells (APCs)], a default state of hypo-responsiveness to several inflammatory stimuli, including commensal components (Smith et al, 2001; Smythies et al, 2005). Such APCs conditioning is likely to be related to commensal signaling in intestinal epithelial cells (IECs) through PRRs. This type of conditioning is further reinforced by the phagocytosis of apoptotic cells (epithelial cells, lymphocytes) that represent an important potential source of bioactive TGF-β capable of enhancing TGF-β production from DCs and macrophages (Chen et al, 2003; Perruche et al, 2008). All these factors are combined in generating the tolerizing milieu that normally characterize the intestinal mucosa, and consistently, lamina propria T cells show a constitutive reduced proliferation upon polyclonal stimulation and a high activation-induced apoptosis rate compared to the circulating T cells (Qiao et al, 1991; Boirivant et al, 1996).

DCs loaded with antigens migrate in secondary lymphoid organs [Peyer’s patches and mesenteric lymph nodes (MLNs)] where they trigger an adaptive response by presenting antigens to naïve T cells. At these inductive sites of immune response, APCs have been increasingly thought to possess flexibility in their function because of their ability to modulate differentiation of CD4+ T-cell subsets with opposite function and characteristic cytokine profiles [either inducible T regulatory cells (iTregs) or effector helper T cells (Th1, Th2, and Th17)] depending on the local signals received (either regulatory or inflammatory) (Mowat, 2003). For a long time, T helper responses have been defined according to the relevant cytokine production as Th1 [IL-12 driven-interferon (IFN)-γ production] or Th-2 (IL-4- production in the absence of IL-12) (Zhu and Paul, 2010). Recently, an additional subset of T helper cells producing IL-17 (Th17) has been characterized (Harrington et al, 2005; Park et al, 2005). This subset is believed to differentiate from naïve T cells in the presence of TGF-β and the inflammatory cytokines IL-1β and IL-6 and be dependent upon IL-23 for expansion (Veldhoen et al, 2006; Manel et al, 2008; Volpe et al, 2008; Valmori et al, 2010). Human Th17 response is generally recognized (and currently defined) as characterized by the production of several pro-inflammatory cytokines [namely IL-17A, IL-17F, IL-6, IL-21, IL-22, and tumor necrosis factor (TNF)-α] involved in acute inflammatory response (the recruitment of neutrophils at sites of inflammation) as well as in chronic autoimmune response and, in some circumstances, in regulatory responses. Indeed, the Th17 response is characterized by a high degree of plasticity, because IL-17 producing cells can co-express, under different circumstances, the regulatory cytokine IL-10 or the inflammatory cytokine IFN-γ (Crome et al, 2009; Zhu and Paul, 2010). Increasing evidence suggests that different subsets of DCs are present in the gut-associated lymphoid tissue, in particular a subset expressing the surface marker CD103 has been demonstrated to preferentially induce the generation of regulatory Foxp3+ cells (Coombes et al, 2007) via a TGF-β and RA-dependent mechanism. Tregs actively control autoimmunity and tissue homeostasis. TGF-β and IL-10 are the molecules associated with their development and function (Maynard and Weaver, 2008; Feuerer et al, 2009). Treg populations include thymic-derived natural Tregs, responsible for maintaining self-tolerance, and iTregs that are generated in the periphery to prevent inappropriate and exuberant immune responses to microbial or tissue antigens (Ito et al, 2008). The hallmark of nTreg cells is the expression of the Foxp3 transcription factor, which is required for maintaining Treg cell function (Williams and Rudensky, 2007). TGF-β has been shown to maintain peripheral nTreg cells (Marie et al, 2005). In the presence of TGF-β, Foxp3 can also be induced in naive T cells in the periphery, and the resulting iTregs exhibit a suppressive phenotype similar to that in nTreg cells (Wing et al, 2006). The most intriguing observation on Foxp3+ Treg cells is the increased evidence of their ability, in inflammatory conditions, to switch from a regulatory to an effector phenotype, mainly to IL-17- producing cells (Xu et al, 2007; Koenen et al, 2008; Deknuydt et al, 2009). In particular, it has been reported, in in vitro studies and in animal models, that under lymphopenic and/or inflammatory conditions, Tregs may lose Foxp3 and/or acquire different effector functions, especially in the intestine, which may contribute to uncontrolled inflammation (Murai et al, 2010). IL-23 appears to play a central role in orchestrating the mucosal immune response through modulation of both the inflammatory and regulatory arms of Th cell response. Indeed, IL-23 is able to promote the expansion of Th17 cells to induce the co-expression of IFN-γ in Th17 lymphocytes and to directly inhibit Tregs expansion (Izcue et al, 2008; Ahern et al, 2010). Thus, the IL-23-Th17 axis and its relation with regulatory cells represent currently the focus point of research aimed at elucidating the genesis of inflammation.

After priming in gut-associated lymphoid tissues, T cells enter the systemic circulation. During priming by DCs, T cells acquire the surface expression of enterotropic molecules (α4β7) that guarantee the gut homing of primed lymphocytes once they enter the systemic circulation (Johansson-Lindbom et al, 2003; Mora et al, 2003). These activated CD4+ T cells can, thereby, leave the circulation to enter the intestinal lamina propria, where they carry out their own specific functions (Mowat, 2003). In the steady state, the reactivity to mucosal antigens is finely tuned by regulatory mechanisms so that effector cells producing IFN-γ (Th1 response) or IL-4, IL-5, IL-13 (Th2 response) or IL-17 (Th17 response) are controlled by regulatory cells. In IBD, this balance is broken, resulting in chronic inflammation.

Dysregulated immunity in IBD

The inflammatory response in IBD is characterized by an increased production of inflammatory cytokines (TNF-α, IL-6, IL-1β) sustained by T cell-derived cytokines. In CD, the IL-12/23-induced IFN-γ production is recognized as having the major role in sustaining the inflammation (Th1-mediated inflammation). In UC, increased IL-5 and IL-13 production have been reported to play a significant role in the generation and perpetuation of the inflammation (Th2-like-mediated inflammation) (Fuss et al, 1996, 2004; Bouma and Strober, 2003). Irrespective of the pathway of inflammation (Th1 or similar Th2), both in CD and UC the local activation of anti-apoptotic pathways in pathogenic lamina propria T lymphocytes as a secondary consequence of inflammation (Boirivant et al, 1999; Ina et al, 1999) leads to further expansion of these cells in the lamina propria and, consequently, further increase in inflammatory mediators, inhibition of the regulatory function and perpetuation of IBD. In CD and UC, an increased expression of IL-17 and IL-23 has been shown (Fujino et al, 2003; Kobayashi et al, 2008; Rovedatti et al, 2009), although IL-23 appears to differently regulate the effector Th subsets balance in CD and UC, enhancing the production of the Th1 signature cytokine IFN-γ by lamina propria CD4+ T cells in patients with CD, and the production of IL-17 in UC (Kobayashi et al, 2008). Thus, in IBD, the increased production of IL-17 has been reported both in CD and in UC, but its role in the two diseases has not been completely clarified. Right after experimental data have linked the Th17 response to human chronic inflammatory and autoimmune diseases, and the effect of IL-23 mainly to the Th17 response, the concept of IBD as a Th17 co-mediated disease and of ‘IL-23/IL-17 axis’ as central in this pathological condition has immediately emerged. This paradigm, although well fitting with the reported genetic association of both forms of IBD with variants in the IL-23/p40 subunit, IL-23 receptor (IL-23R), and the signal transducer and activator of transcription (STAT)-3 (the predominant downstream mediator of IL-23 signaling), appears to require further confirmation. In animal models, IL-17-deficient T cells are not impaired in their ability to induce colitis (Noguchi et al, 2007; Izcue et al, 2008), suggesting that Th17 cell responses are not essential for colitis. Moreover, IL-23 seems to directly inhibit the Foxp3+ regulatory cells (Izcue et al, 2008) suggesting that, inhibition of regulatory cells, more than the increase in effector mechanisms, may represent the sine qua non condition for the development of chronic inflammation. Thus, the paradigm of the pathogenetic IL-23/IL-17 mucosal axis might change to that of IL-23/IFN-γ or IL-23/Tregs axis. With regard to regulatory cells, studies performed in IBD did not show a reduction in the percentage of regulatory cells infiltrating the mucosa. On the contrary, it has been reported that regulatory Foxp3+ T cells are increased during the active phases of the diseases and are no different from controls during remission. Furthermore, in contrast to many mouse models of IBD, regulatory T cells in human IBD appear to be completely capable of suppressing effector T-cell proliferation (Maul et al, 2005). It should be emphasized here that the ability of intestinal regulatory cells to suppress effector T-cell proliferation has been tested using, as effector cells, T cells from peripheral blood. T effector cells in an inflamed site might show a different response to suppression. Indeed, it has been shown that effector cells in active IBD and in experimental colitis present increased Smad7 expression (an intracellular inhibitor of TGF-β signaling) that might prevent the TGF-β-mediated suppressive effect of regulatory T cells (Monteleone et al, 2001; Boirivant et al, 2006). Thus, the possibility exists that regulatory function might be ineffective as a consequence of the presence, during the active phases of the disease, of effector cells resistant to inhibition by regulatory cells. In summary, the experimental findings obtained so far indicate an increased local production of pro-inflammatory cytokines from infiltrating lamina propria mononuclear cells in patients with IBD when compared to controls, but evidence of primary phenomena is still lacking. Recently, new data confirmed a previous observation that revealed the presence, in patients with CD, of a weak acute inflammatory response, resulting in the development of chronic inflammation because of partial immunodeficiency (Segal and Loewi, 1976; Marks and Segal, 2008; Smith et al, 2009). However, as the observations emerging from those studies were obtained in patients with anatomical lesions, albeit in clinical remission, it is not possible to rule out the possibility that they were secondary to the presence of chronic inflammation.

The genetically susceptible host

In IBD, several studies have shown the presence of a disorder of immune tolerance to luminal antigens (Duchmann et al,1995,1996; D’Haens et al, 1998; Sellon et al, 1998). The conditions leading to dysregulation of the immune response toward the microbiota remain unknown. However, following the initial experimental studies showing the presence of a loss of immune tolerance, genetic studies revealed an increased risk of CD or UC associated with polymorphisms in genes encoding components of innate and adaptive immune response, the latter involving mainly the crosstalk between innate and adaptive immune responses. In CD, a meta-analysis of three CD genome-wide association studies (GWAS) has identified more than 30 loci associated with the disease, with odds ratio ranging from 1.08 to 3.99 (Barrett et al, 2008). Prominent among these genetic studies, for the strength of the association and the insights into the mechanism of the disease, are those related to polymorphisms in genes related to adaptive immunity (IL-23R, STAT-3; associated with both CD and UC) (Duerr et al, 2006; Wellcome Trust Case Control Consortium, 2007) or to genetic mutations related to a disturbed surveillance of bacteria of the microflora by the intestinal mucosa (NOD2) (Lesage et al, 2002; Economou et al, 2004; Abraham and Cho, 2006; Hedl et al, 2007), or to deficient autophagy (ATG16L1, IRGM) (Hampe et al, 2007; Homer et al, 2010; Travassos et al, 2010). It is important to stress here that these variants do not appear to be either unique or necessary for the disease to express. Attempts have been made to link the associated genetic variants with the classic clinical CD sub-phenotypes. So far, the association of NOD2 variants with ileal disease location has been described (Ahmad et al, 2002; Cuthbert et al, 2002; Hampe et al, 2002; Vermeire et al, 2002) but data from a recent study indicate a poor relationship between the genetic-based subgroups and the clinical sub-phenotypes used (Cleynen et al, 2010).

Implications for target of therapy

The better understanding of the underlying immunopathogenetic mechanisms of IBD has provided the rational basis for the development of biological therapeutic approaches, selectively targeting those inflammatory mediators which appeared to be involved in initiation and/or perpetuation of the uncontrolled mucosal immune response to resident intestinal microflora. This pathophysiology-orientated therapeutic strategy was focused mainly on the inhibition of several pro-inflammatory cytokines reported to be increased in IBD lesions and known (or suggested) to be, in some way, involved in the intestinal expansion of the pathogenetic effector T-cell subsets (by means of their differentiation, proliferation or survival).

The pro-inflammatory cytokine TNF-α, mainly produced by activated macrophages and lymphocytes, and shown to be elevated in the intestinal mucosa and serum of patients with IBD (MacDonald et al, 1990; Breese et al, 1994; Reimund et al, 1996), was first selected as a target of therapeutic neutralization through monoclonal antibodies.

Infliximab (IFX), a chimeric IgG1 anti-TNF-α monoclonal antibody, is currently an important option used to induce remission in steroid-refractory patients with CD and in the treatment of fistulizing disease. Although its precise molecular mechanism of action remains to be fully elucidated, it has been demonstrated to induce apoptosis in lamina propria T cells of patients with active CD, and, even more significantly, induction of intestinal cell apoptosis was related to clinical efficacy of anti-TNF-α treatment in these patients (Van den Brande et al, 2007). These observations, besides providing evidence that induction of intestinal T-cell apoptosis is a central molecular mechanism of action of at least this anti-TNF-α agent in IBD, also further support the general concept of a link between clinical efficacy and drug ability in the restoration of activated effector T-cell susceptibility to apoptosis in the gut. The beneficial effect of azathioprine (AZA), in IBD, has consistently been shown to be more likely related to induction of T-cell apoptosis rather than to inhibition of cell proliferation (Tiede et al, 2003).

The involvement of the pro-inflammatory cytokine IL-6 in chronic intestinal inflammation has been well documented, although its function, in these diseases, remains to be clearly defined (Isaacs et al, 1992; Reinecker et al, 1993). In IBD lesions, macrophages and T cells are likely to be the main producers of IL-6, and the IL-6 signal is received by T cells through its alternative pathway [also termed IL-6 trans-signaling: membrane binding of the complex IL-6/soluble IL-6R by the signal-transducing glycoprotein (GP)-130]. IL-6 trans-signaling then leads to translocation of STAT-3 and thereafter induction of the anti-apoptotic genes Bcl-2 and Bcl-xl, thereby conferring to intestinal T cells resistance against apoptosis (Atreya et al, 2000). These findings suggested a novel pathophysiological mechanism underlying the uncontrolled intestinal inflammatory process in IBD, providing the basis for development of potentially therapeutic strategies. Treatment with antibodies against the IL-6R suppressed or reduced the inflammatory activity in murine models of chronic intestinal inflammation, and the curative impact of the antibody was based upon induction of intestinal T-cell apoptosis, confirming the pathogenic role of IL-6 trans-signaling in vivo (Atreya et al, 2000; Yamamoto et al, 2000). Subsequently, in an exploratory placebo-controlled trial, a humanized anti-IL-6R monoclonal antibody, Tocilizumab, was shown to be effective in patients with active CD (Ito, 2005). Thus, the anti-IL-6R antibody may represent another therapeutic option for the management of IBD.

The heterodimeric (p35/p40) pro-inflammatory cytokine IL-12 has been considered an attractive target for the treatment of IBD, on account of its pivotal role in controlling the Th1 T-cell differentiation (Zenewicz et al, 2009) and of the observation that anti-IL-12/p40 treatment was shown to abrogate established colitis and to regulate apoptosis of Th1 T cells in experimental colitis in mice (Neurath et al, 1995; Fuss et al, 1999). This therapeutic approach is encouraged by the results of two pilot trials in patients with CD, showing the clinical efficacy of anti-p40 treatment (Mannon et al, 2004; Sandborn et al, 2008). However, it is still necessary to establish the extent to which the therapeutic effect of this compound in vivo can be attributed to neutralization of IL-12 or IL-23, a heterodimeric cytokine composed of the specific p19 subunit and the p40 chain which is shared with IL-12 (Oppmann et al, 2000). These latter considerations, together with the above-mentioned genetic and experimental findings regarding the IL-23 contribution to disease, clearly have important implications for the design of new therapeutic strategies in IBD. Blockade of IL-23p19 may be of clinical efficacy, thereby supporting the therapeutic importance of enhancing regulatory function, which has, so far, clearly been underestimated.

Clinical aspects and diagnosis

Ulcerative colitis

UC affects the mucosal layer of the large bowel, and it involves invariably the rectum and may extend proximally to part or even to the entire colon, in an uninterrupted fashion (Judge and Lichtenstein, 2003). Depending upon the anatomic extent of involvement, patients can be classified as having proctitis (involvement limited to the rectum), left-sided colitis (involvement limited to the portion of the colon distal to the splenic flexure), or extensive colitis (Silverberg et al, 2005). The extent of UC influences not only the medical treatment (oral and/or topical therapy) (Stange et al, 2008), but also the schedule for colonic cancer surveillance, because the contribution of the disease extent at the time of diagnosis to the risk of malignancy has been confirmed (Katsanos et al, 2007).

Symptoms of UC depend upon the extent and severity of the disease, and usually include bloody diarrhea, rectal bleeding, rectal urgency, passage of pus, mucus or both, abdominal pain during bowel movements, weight loss, and fatigue. Systemic symptoms of malaise, anorexia, or fever are features of a severe attack (Friedman and Blumberg, 2008). In clinical practice, disease activity is typically described as mild (up to 4 bloody stools per day and no systemic toxicity), moderate (4–6 bloody stools per day and minimal toxicity), or severe (more than 6 stools per day and signs of toxicity, such as fever, tachycardia, anemia, increase in erythrocyte sedimentation rate) (Carter et al, 2004; Kornbluth and Sachar, 2004; Baumgart and Sandborn, 2007). Fulminant colitis is a rare and severe form of the disease, still lacking a formal definition, commonly referred to patients presenting extensive bleeding and anemia requiring blood transfusion, high fever, raised biochemical markers of inflammation, abdominal tenderness, and colonic dilation on plain abdominal film, a condition often leading to potentially fatal outcomes, such as toxic megacolon and colonic perforation (Sands, 2008).

With regard to disease behavior over the years, two cohort studies showed that, by 25 years, more than half of the patients with proctitis will progress to left-sided colitis and patients with more extensive disease will regress in approximately 75% of cases (Langholz et al, 1996). The most aggressive phase of the disease was frequently seen in the first 3 years after diagnosis (Langholz et al, 1994). In the first 3–7 years after diagnosis, 25% of patients were in remission, 18% experienced disease activity every year, whereas 57% had intermittently occurring relapses. After 10 years, the cumulative probability of colectomy was 24%. Reasons for procto-colectomy are mainly represented by severe colitis refractory to intensive treatment or rescue therapy (see below), massive bleeding, impending or toxic megacolon, non-severe chronic active disease despite therapy with immuno-suppressive and/or immuno-modulating agents, and dysplasia or cancer. Development of dysplasia or adenocarcinoma is associated with longstanding and extensive UC and insufficient suppression of chronic inflammation (Gupta et al, 2007; Viennot et al, 2009; Langholz, 2010). As far as concerns duration of disease, the reported cumulative probability of colorectal cancer (CRC) appears to be decreased, in the last years, from 8% to 18% after 20 and 30 years of disease (Eaden et al, 2001), to 2.5% and 7.6%, respectively (Rutter et al, 2006). In some recent studies, the risk of CRC, in patients with UC, has been reported to be comparable to that in the general population (Winther et al, 2004; Jess et al, 2006). This has been held to be due primarily to the more widespread use of surveillance colonoscopy, the more frequent use of chemo-prevention and the more frequent use of surgery in the UC treatment strategy. CRC occurs primarily in patients with extensive colitis or in those with a history of extensive colitis. These patients have a relative risk (RR) of CRC of 14.8, vs 2.8 in the case of left-sided colitis (population-based study; 12 000 person-years of mean follow-up). The risk of CRC is virtually inexistent in the case of proctitis and very low (RR = 1.7) when disease is limited to the portion of the colon distal to the splenic flexure (Ekbom et al, 1990). The association of persistent chronic inflammation and risk of CRC has long since been controversial; nevertheless, recent clinical and experimental data support its contribution to colon carcinogenesis (Viennot et al, 2009). Furthermore, features indicative of previous severe inflammation, such as pseudopolyps, and features indicative of chronically active colitis, such as a shortened or tubular colon and stricture formation, are all associated with a significant increase in the risk of CRC (Rutter et al, 2004).

Medical treatment with 5-aminosalicylates (5-ASA) has proved, in most studies, to prevent the onset of CRC (Velayos et al, 2005). Endoscopic surveillance is based upon the possibility to reveal the presence of dysplasia. Several questions still remain to be answered, such as, for instance, how dysplasia should be managed and how best to detect dysplasia in the flat mucosa. The relevance of the latter question emerges from the greater frequency of dysplasia in the flat mucosa, in IBD, compared to sporadic CRC. An analysis of the efficacy of surveillance strategies is difficult, on account of several unavoidable methodological limitations, especially the lack of control studies. However, some studies suggest that it reduces mortality from CRC in regularly monitored patients (Langholz, 2010). Overall, patients with UC have a normal life expectancy (Winther et al, 2003).

The diagnosis of UC relies upon a combination of medical history, endoscopic findings, histological features on multiple colonic biopsies, and negative stool tests for infectious agents. Endoscopic findings include loss of normal vascular pattern, erythematous and granulous appearance of the mucosa, mucosal friability, and ulceration. Histological lesions affect primarily the mucosa and show a continuous pattern. Histological changes in the mucosa of active disease include ulcerations, crypt abscesses, loss of globet cells, and infiltration of lymphocytes and granulocytes (Stange et al, 2008). Quiescent disease may show only architectural distortion of the mucosal crypts. These features may help in ‘delayed’ (i.e. once the acute episode is overcome) differential diagnosis with bacterial acute colitis, that results, as a rule, in complete restitution ad integrum. Stool specimens should be cultured for common pathogens including specific assays for Clostridium difficilis toxin A and B, Campylobacter spp and Escherichia coli 0157:H7 (Stange et al, 2008).

Crohn’s disease

Crohn’s disease involves the ileum and large bowel in more than 90% of patients, but it can affect any part of the GI tract from the mouth to the anus.

Unlike UC, anatomical lesions are discontinuous and chronic inflammation is transmural (Judge and Lichtenstein, 2003). At diagnosis, the disease involves the terminal ileum in 47% of patients, the large bowel in 28%, the small and large bowel in 21%, and the upper GI tract in 3% (Louis et al, 2001). Perianal disease and, in particular, perianal fistulas are often associated with CD localized elsewhere. It is more often associated with colonic disease involving the rectum, followed by colonic disease with rectal sparing and by ileocolonic disease. The lowest frequency of perianal disease occurs in patients with isolated ileal disease (Hellers et al, 1980). Perianal disease often precedes, or appears simultaneously with, intestinal symptoms (Hellers et al, 1980; Schwartz et al, 2002). The cumulative frequency of perianal fistulas occurrence appears to increase with the duration of disease (12% at 1 year, 26% at 20 years) (Hellers et al, 1980).

The Vienna classification recognizes distinct clinical phenotypes of CD with respect to disease location and development of parietal (strictures) or extra-parietal (fistulas or abscesses or both) complications, the occurrence of which are, respectively, referred to as stricturing and penetrating behavior (Gasche et al, 2000).

Although the anatomical involvement of CD is fairly stable over time, the behavior of the disease varies substantially with time. In a French study, behavior was classified, at the diagnosis, as non-stricturing and non-penetrating in 70% of patients, stricturing in 17%, and penetrating in 13% of patients. After 10 years of follow-up, there was a change from non-stricturing non-penetrating to either stricturing, in 27%, or penetrating disease, in 29%, of patients (Louis et al, 2001).

Symptoms of CD are heterogeneous and vary depending upon the location, behavior, and severity of disease, as well as the presence of extra-intestinal manifestations, but usually include diarrhea for more than 6 weeks, abdominal pain, and weight loss. These symptoms should give rise to the suspicion of CD, especially in young patients. Systemic symptoms of malaise, anorexia, or fever are also common (Stange et al, 2006). Although in clinical trials, clinical or endoscopic disease activity can be measured with a variety of disease activity indices (Sandborn et al, 2002), in clinical practice, disease activity is typically described as mild (ambulatory patients able to tolerate oral alimentation without manifestations of obstruction, dehydration, fever, abdominal mass or tenderness, or >10% weight loss), moderate (treatment for mild disease ineffective, more prominent symptoms of abdominal pain and tenderness, intermittent nausea and vomiting without overt obstruction, >10% weight loss or significant anemia), and severe disease (persisting symptoms despite intensive treatment, persistent vomiting, evidence of intestinal obstruction or abscess, rebound tenderness, BMI <18 kg−2) (Stange et al, 2006; Baumgart and Sandborn, 2007). The clinical course is typically chronic intermittent with periods of clinical activity and remission. It has been reported that 1 year after diagnosis, 10–30% of patients with CD have a relapse or exacerbation of disease, 15–25% experience low disease activity, and 50–65% are in remission. Long-term follow-up (10–15 years) showed that the majority of patients (up to 73%) experienced a chronic intermittent course of the disease, 10% experienced prolonged remission, while 20% showed a chronic course with continuous activity (Munkholm et al, 1995; Loftus et al, 2002). During the course of the disease, as a consequence of transmural inflammation, local complications such as strictures, fistulas with bladder and vagina, intra-abdominal abscesses, may occur and, in many cases, may require surgery. Because of the frequent occurrence of local complications and/or the presence of a chronic course of the disease with continuous activity not responding to medical treatment, the majority of patients with CD require surgery with time (Munkholm et al, 1993, 1995; Louis et al, 2001; Schwartz et al, 2002; Carter et al, 2004). After 20 years, most patients with CD have been operated on at least once (Cosnes et al, 2002). Following surgical resection, recurrence of CD is virtually inevitable (Rutgeerts et al, 1990). Risk factors for recurrence of CD after surgery include penetrating/fistulizing disease behavior, young age, short duration of disease before surgery, ileocolonic disease and cigarette smoking (Langholz, 2010). The latter is an accepted risk for a more severe disease course, in terms of both clinical activity and need of either first or further intestinal surgery (Papay et al, 2010).

Because of the multiple surgical resections and/or to the extensive inflammatory involvement of the intestine, progressive structural damage to the bowel can occur, with intestinal function becoming irreversibly lost (D’Haens, 2010). Manifestations of loss of function include bile-salt diarrhea, steatorrhea, vitamin and mineral deficiencies, anemia, short bowel syndrome with dehydration and malnutrition, and, in case of dysfunction of internal mechanisms, loss of continence (Judge and Lichtenstein, 2003).

In CD patients with colonic involvement, an overall RR of CRC of 2.5% was estimated, whereas such risk in patients without colonic involvement was similar to that in the general population (Canavan et al, 2007). Patients with small intestinal CD are at increased risk of small bowel adenocarcinoma, with a reported pooled RR of 31.2 compared with the general population (Canavan et al, 2006). Although this difference is highly significant, the real risk is low because small bowel adenocarcinoma is a rare cancer and accounts for <5% of all GI cancers (Langholz, 2010). No single evaluation can establish the diagnosis of CD. Instead, the diagnosis is made on the basis of clinical history and physical examination, together with objective findings from endoscopic, radiological, laboratory, and histological studies. When CD is suspected, hematological tests, such as complete blood count, C-reactive protein (CRP), erythrocyte sedimentation rate, routine stool examination and fecal calproctectin measurement (when available), should be performed (Stange et al, 2006; Benevento et al, 2010). The value of routine stool examination, in patients with suspected CD or exacerbations of disease, arises both from the need to rule out enteric infections and to definitely exclude C difficilis colitis that, in its recurrent form, may mimic CD (Thomas et al, 2003; Thielman and Guerrant, 2004). Ileo-colonoscopy and biopsies from the terminal ileum, as well as from each colonic segment to look for microscopic evidence of CD, are first-line procedures to establish the diagnosis. The most useful endoscopic features of CD include skip lesions, cobblestone mucosa, and ulcerations with clear margins, usually surrounded by normal mucosa. Focal (discontinuous) chronic (lymphocytes and plasma cells) inflammation and patchy chronic inflammation, focal crypt irregularity (discontinuous crypt distortion), and granulomas (not related to crypt injury) are the generally accepted microscopic features that allow a diagnosis of CD (Stange et al, 2006). Interestingly, recent studies have reported that non-invasive procedures, such as abdominal and pelvic US, provide support to the clinical suspicion (Benevento et al, 2010); thus, in the presence of an experienced operator, abdominal US may precede ileo-colonoscopy that will be performed if the findings obtained by laboratory and US tests support the suspicion of CD. Once endoscopic and histological findings confirm the diagnosis of CD, then the location, extent of inflammatory lesions, and the presence of complications should be carefully assessed (Stange et al, 2006). To this end, the patient should undergo an imaging study of the bowel to define extent, severity, and type of disease (inflammatory vs fibrotic vs penetrating), as well as to exclude the presence of septic complications (abscesses) of CD. Several instrumental techniques are used to investigate the small intestine including fluoroscopic examinations [small bowel follow through (SBFT, a non-intubation study), and small bowel enema, (SBE, intubation study)], trans-abdominal US, and magnetic resonance (MR) – or computed tomography (CT)-enteroclysis (intubation study) or enterography (non-intubation study). SBE is gradually being replaced by other techniques. Major pitfalls are the high dose of radiation, the poor capacity, compared with other techniques, of quantifying the degree of inflammation of the bowel wall and the poor acceptance rate of patients. It is still considered relatively reliable for evaluating strictures and internal fistulas, before surgery (Angriman et al, 2008). Compared with SBE, SBFT has similar sensitivity and specificity and is associated with less radiation exposure, is less expensive and less time-consuming (Bernstein et al, 1997). Furthermore, mucosal detail is usually better identified with SBFT.

Regardless of performance, the current preference for one technique with respect to another often depends largely on institutional standards, availability of local expertise and individual experience. CT scan or MRI should be preferred to traditional techniques, however, when an abscess or additional pathological findings are suspected (Gore et al, 1996; Gasche et al, 1999; Rieber et al, 2000, 2002; Potthast et al, 2002; Rohr et al, 2002; Stange et al, 2006; Benevento et al, 2010).

Extra-intestinal manifestations

Up to 47% of patients with IBD may suffer from non-intestinal disorders that are commonly referred to as extra-intestinal manifestations (EIMs) of IBD, because they often show an epidemiological, clinical, and response to therapy profile constituting a true parallel extra-intestinal disease. The organs most commonly involved include the skin, eyes, joints, biliary tract, and lungs. Some EIMs, such as oral lesions and amyloidosis, occur more often in association with CD than with UC. Other manifestations, such as those affecting skin and eyes, are seen equally in CD and UC. The development of one EIM appears to increase the susceptibility of developing other EIMs. The high concordance in EIMs, in siblings and first degree relatives with IBD, suggests a relevant genetic contribution (Rothfuss et al, 2006).

Extra-intestinal manifestations of IBD often respond to systemic therapy with anti-inflammatory drugs, immuno-suppressive and biological agents directed to the underlying intestinal disease, the activity of which they usually parallel, with the exception of primary sclerosing cholangitis and ankylosing spondylitis (Rothfuss et al, 2006; Fatahzadeh, 2009).

Oral manifestations

Of particular importance to oral health care specialists are muco-cutaneous manifestations that may occur in IBD.

Mucosal changes in the oral cavity of patients with UC include stomatitis, glossitis, cheilitis, aphthous ulceration, and pyostomatitis vegetans (Boh and al-Smadi, 2002).

Pyostomatitis vegetans, a lesion typically associated with UC, appears as multiple, friable pustules, erosions, and ulcerations of the oral mucosa. It reflects the presence of active intestinal disease and often appears before the diagnosis of UC (Hansen et al, 1983; Neville et al, 1985; Chan et al, 1991).

The considerable variability of reported prevalence of oral lesions associated with CD (between 0.5% and 80%) is most probably attributed to inclusion of non-specific CD oral changes (Lisciandrano et al, 1996; Greenberg and Pinto, 2003; Fatahzadeh, 2009; Rowland et al, 2010).

Distinct disease-specific oral lesions have been described in patients with intestinal CD, including swelling of the lips, buccal mucosal swelling or cobble-stoning, muco-gingivitis, deep linear ulceration (usually in the buccal sulci), and mucosal tags, which may cause pain, impair oral functions, or lead to disfigurement and psychological distress (Plauth et al, 1991; Scheper and Brand, 2002; Fatahzadeh, 2009; Rowland et al, 2010). Oral CD, as conventionally referred to patients with intestinal CD who present involvement of the mouth (Rowland et al, 2010), shows a young age of presentation (Plauth et al, 1991). Diagnostic work-up of suspected oral CD involves tissue sampling and histopathological examination for the presence of non-caseating granulomas similar to those identified in intestinal CD, together with complete GI evaluation (Fatahzadeh, 2009). In two prospective studies, from the same group, in children presenting with suspected oral CD, typical oral lesions (with the presence of granuloma in 100% of biopsy specimens, when taken) were found in 41.7% of patients, but only 30% of them continued to show oral manifestations at follow-up (mean follow-up 55 months) (Harty et al, 2005; Hussey et al, 2007). These findings indicate that, at least in children, examination of the mouth for oral CD is of value only at the initial presentation, and more important the oral cavity provides an easily accessible source of diagnostic material when IBD is suspected. As in bowel disease, the presence of granuloma is neither highly specific (causes of granuloma to be ruled out include fungal and mycobacterium infection) (Palamaras et al, 2008), nor is it sensitive (because its absence may be related to sample error, a variable incidence of this histological lesion or patchy distribution of CD) (Plauth et al, 1991; Scheper and Brand, 2002).

Oro-facial findings such as stomatitis, glossitis, aphthous ulceration, cheilitis, or peri-oral dermatitis may also be secondary to deficiencies of albumin, zinc, folic acid, niacin, vitamin B, and other essential nutrients, as a consequence of disease-related lesions of the intestinal mucosa or reduced intake of foods, luminal malabsorption (overgrowth of intestinal bacteria), bowel resection or drugs used to treat IBD (Judge and Lichtenstein, 2003; Fatahzadeh, 2009).

Management

Ulcerative colitis

5-ASA (mesalazine) is currently the recommended first-line inductive therapy for mild-to-moderate UC. Proctitis and colitis, regardless of extent, belonging to this class of activity, should be first treated, respectively, with topical mesalazine and with the combination of oral and topical mesalazine. The same combination therapy, with a topical steroid also being added or even replacing oral mesalazine, should be the second choice for proctitis not responding to topical 5-ASA (Travis et al, 2008). This topical 5-ASA-based treatment is effective for most patients with ulcerative proctitis (UP), as they show an unsatisfactory response and are therefore considered for escalation of treatment in less than 5% of cases (Lakatos and Lakatos, 2008). If symptoms of active disease do not respond rapidly to mesalazine, oral corticosteroids (CSS) are indicated both in UP and non-severe colitis of any extent. For patients with non-severe steroid-refractory UC, alternative options are immuno-suppressive thiopurines [AZA and 6-mercaptopurine (6-MP)], IFX, cyclosporin A (CsA) and surgery (Lichtenstein et al, 2006; Travis et al, 2008).

Patients with severe UC should be treated with intravenous corticosteroids, together with additional measures of nutritional support, hydro-electrolytic replacement, and blood transfusion, if necessary. In the absence of a rapid response to intravenous steroids (as can be objectively assessed by stool frequency, CRP and abdominal radiography), a rescue therapy with CsA or IFX should be the only short-term option alternative to total procto-colectomy (Kornbluth and Sachar, 2004; Travis et al, 2008).

The profoundly different role of surgery in CSS-refractory non-severe chronic active UC and severe UC should be stressed. Given the high-potential fatal outcome of the latter, surgery, in such cases, should never be considered the last chance once rescue therapy has failed, and medical rescue treatment should be chosen adopting safety as the exclusive criterion. On the contrary, in CSS-refractory chronic active non-severe UC, additional considerations should be taken into account in decision making, besides severity of symptoms, including patient’s will. Moreover, although procto-colectomy will cure UC, more than 50% of patients who undergo surgery develop some form of pouchitis and the quality of life achieved with medically induced remission is usually superior to that achieved with surgery (D’Haens, 2010).

As far as concerns maintenance therapy, mesalazine is the first-line therapy for the maintenance of remission. Both AZA and IFX are effective for this purpose in patients who relapse while on oral 5-ASA compounds, and both of these are effective in maintaining remission of CSS-dependent UC (steroid-sparing agents), although AZA should be the first choice in apparent steroid dependence. IFX is also effective in maintaining remission and sparing steroids in patients who remain CSS-dependent despite treatment with 5-ASA and thiopurines (Rutgeerts et al, 2005; Travis et al, 2008).

Although comparable to that in adults, the management of pediatric UC patients must carefully take into consideration several important consequences both physical and psycho-social (both of the disease and the respective therapy), which are for some reason unique to evolutive age. Therefore, the ECCO consensus states that colectomy is indicated for persistently active disease, when CSS dependency exists, despite concomitant therapy with thiopurines, and also when there is evidence of growth failure despite apparently adequate maintenance therapy (Biancone et al, 2008). The relevance of these considerations arises from the predominance of pancolitis in children (70–80% of cases) at the time of diagnosis (Griffiths, 2004).

Crohn’s disease

As for UC, effective treatment options for CD depend upon disease location and activity and the presence of complications, and evidence-based treatment principles aimed at maximizing benefits against risks have been established (standard ‘step-up’ treatment). According to these recommendations, CD remission should be induced with systemic CSS, using first ‘topically acting’ steroids (budesonide) only in mild-to-moderate localized ileo-cecal CD (Travis et al, 2006). Thiopurines (or, if intolerant, methotrexate; all these drugs are traditionally referred to as immuno-suppressive) should be added for those who have relapsed, because of their efficacy in steroid-sparing and maintaining remission, but should be started simultaneously in extensive, moderate–to-severe small bowel disease and are drugs of first choice in CSS-refractory disease (Travis et al, 2006; Prefontaine et al, 2009, 2010).

Infliximab should be reserved for steroid and immuno-suppressive refractory disease, or intolerance, or steroid dependence (in this latter case, in addition or as an alternative to an immuno-suppressive, upon its failure), albeit surgical options should also be considered and discussed in these more complex cases. IFX is also an effective maintenance agent, as it is effective in fistulizing CD (Lichtenstein et al, 2006; Travis et al, 2006). In this latter behavior of CD, antibiotics, thiopurines, or combined immuno-suppression (biological plus conventional immune-suppressive agent), are commonly the medical treatment of progressive choice (Hanauer et al, 2001; Sandborn et al, 2003; Travis et al, 2006). The optimum therapy for postoperative maintenance of remission in patients with CD remains to be determined (Doherty et al, 2009).

Although surgery may be necessary to induce remission or to treat complications in some patients, it will not cure CD. Resection is generally postponed for as long as possible, also in consideration of its functional consequences and of the high likelihood of recurrence of CD (Rutgeerts et al, 1990) with similar features at recurrence to those before surgery as far as concerns location and behavior (Greenstein et al, 1988). As surgical bowel resection is usually performed in patients with terminal ileal distribution and fibrostenotic and/or perforating complications (Larson and Pemberton, 2004; Ramadas et al, 2010), and these same disease location and behavior associated with high risk of intestinal surgery are likely to recur, an increased long-term probability of further surgery could be inferred in these patients if they were treated with the same standard therapeutic approach. In patients with newly diagnosed CD, early combined immuno-suppression therapy achieved, as compared to conventional step-up therapy, higher clinical remission rates, median time to relapse and, more important, higher frequency of mucosal healing (D’Haens et al, 2008), which has been shown to predict sustained clinical response and to be associated with significantly less need for hospitalization and surgical intervention (Rutgeerts et al, 2004; Baert et al, 2010). Moreover, a marked reduction in surgery rate has been reported to be temporally associated with increased and earlier thiopurine use (Ramadas et al, 2010). These findings suggest the ability of several drugs to alter the natural tendency of CD to evolve, over time, from an inflammatory to a penetrating/stricturing disorder (Cosnes et al, 2002).

It thus appears that sufficient control of inflammation, in the very early phase of CD, should reduce the number of relapses by achieving and maintaining remission and thereby prevent complications and the need for surgery (D’Haens, 2010). A change in the therapeutic approach, not symptom-based and not dominated by symptomatic drugs as first choice (CSS), in a disease appearing to be naturally, sometime silently, progressive over time from the inflammatory stage to complicated tissue lesions, the latter often requiring surgical and not-curing intervention, is increasingly advocated (D’Haens, 2010) and seems now to have some requisite of evidence. Also, in favor of a ‘top-down’ therapy are some observations suggesting a somehow ‘structural’ change over years of CD toward an immuno-inflammatory trait less responsive to more potent therapeutic compounds. However, as possibly associated with more adverse effects (including agranulocytosis, venocclusive liver diseases, lymphoma, severe viral and opportunistic infections, demyelinating disease, and exacerbations of heart failure) (Colombel et al, 2004; Blonski and Lichtenstein, 2006; Siegel et al, 2006; D’Haens, 2007; Gisbert et al, 2007; Gisbert and Gomollón, 2008; Beaugerie et al, 2009), a more aggressive therapy could be reserved for selected patients whose unfavorable disease course was possibly predicted, along with other parameters, by genetic (Guagnozzi et al, 2004; Alvarez-Lobos et al, 2005; Weersma et al, 2009) or even immuno-phenotypic tests, and should, in any case, be guided by objective, rather than clinical, evaluation of inflammation, which can be assessed by endoscopy, histology, and imaging techniques (Benevento et al, 2010).

Unfortunately, to date, we are only at the beginning of single patient disease course stratification and individualized therapy, and clearly further research is needed in this field.

Author contributions

MB and AC equally contributed to the paper.

Ancillary