New pathophysiological insights and modern treatment of IBD
- First Online:
- Cite this article as:
- Engel, M.A. & Neurath, M.F. J Gastroenterol (2010) 45: 571. doi:10.1007/s00535-010-0219-3
Inflammatory bowel disease (IBD), which comprises two main types, namely, Crohn’s disease and ulcerative colitis, affects approximately 3.6 million people in the USA and Europe, and an alarming rise in low-incidence areas, such as Asia, is currently being observed. In the last decade, spontaneous mutations in a diversity of genes have been identified, and these have helped to elucidate pathways that can lead to IBD. Animal studies have also increased our knowledge of the pathological dialogue between the intestinal microbiota and components of the innate and adaptive immune systems misdirecting the immune system to attack the colon. Present-day medical therapy of IBD consists of salicylates, corticosteroids, immunosuppressants and immunomodulators. However, their use may result in severe side effects and complications, such as an increased rate of malignancies or infectious diseases. In clinical practice, there is still a high frequency of incomplete or absent response to medical therapy, indicating a compelling need for new therapeutic strategies. This review summarizes current epidemiology, pathogenesis and diagnostic strategies in IBD. It also provides insight in today’s differentiated clinical therapy and describes mechanisms of promising future medicinal approaches.
KeywordsCrohn’s diseaseIBD diagnosticsInflammatory bowel diseasesPathogenesisTherapeutic strategies and optionsUlcerative colitis
Epidemiology and risk factors of IBD
Crohn’s disease (CD) and ulcerative colitis (UC) are autoimmune chronic inflammatory disorders affecting an estimated 1.4 million persons in the USA and 2.2 million persons in Europe . The past years have seen a stabilization of incidence and prevalence of UC and CD in high-incidence areas, such as northern Europe and the USA, but at the same time, there has been an increase in low-incidence areas, such as southern Europe and Asia. In particular, the incidence of UC is rising in areas with a previously low incidence, such as Japan, South Korea and Singapore [2–4]. In the majority of cases, inflammatory bowel diseases (IBD) are diagnosed in late adolescence and early adulthood. In a systematic review of population-based cohorts of CD from North America, the mean age at diagnosis ranged from 33 to 45 years . Based on statistics, the diagnosis of UC appears 5–10 years later than CD [6, 7]. Analyses of epidemiologic data have revealed gender-related differences in IBD incidence, with a slight female predominance in CD suggesting a hormonal element in disease expression. In contrast UC predominantly affects males. Among the diverse putative risk factors for IBD, cigarette smoking and appendectomy are both independent risk factors for developing CD while, intriguingly, they are protective for UC. Controversial data exist on whether measles or Mycobacterium paratuberculosis infection as well as high sugar or fat diets putatively increase IBD prevalence .
Lessons from human and mouse genetics
The human genome comprises an estimated 25000 genes contained within 3 billion DNA nucleotides. Interindividual variations in DNA sequences, such as large-scale copy number polymorphisms, repetitive sequences of varying length and single nucleotide polymorphisms (SNPs), occur naturally . About 10 million SNPs mirror their multiplicity within the human genome; however, only a minority of these SNPs may have functional consequences. The most studied and well-known gene mutations associated with an increased susceptibility of CD concern the NOD2 gene [9, 10]. The NOD2 gene is located on chromosome 16q12, and its encoded intracellular protein activates nuclear factor κB and mitogen-activated protein kinase pathways in response to stimulation by peptidoglycan components in the bacterial cell wall [11–13]. Three variants of this gene (Arg702Trp, Gly908Arg, Leu1007fsinsC) each independently increase the risk for UC and ileal-only and ileocolonic CD but, interestingly, not for colonic-only CD [9, 10, 14]. Genetic susceptibility, however, seems to form regional clusters, since all three major CD mutations are not observed in Asian patients with CD, and recently reported associations in European CD cohorts, such as increased CD susceptibility in association with IBD5 or interleukin (IL) 23R, have not been replicated in Asian cohorts [15–18].
Animal models of intestinal inflammation are essential to gain a better understanding of the pathogenesis of IBD. Induced or spontaneous mutations in a diversity of genes have identified a number of pathways that can lead to IBD in the mouse . These models implicate a pathological dialogue between the intestinal microbiota and components of the innate and adaptive immune system that ultimately misdirects the immune response to finally attack the colon. A primary component of disease expression is unrestrained effector CD4+ T cell reactivity to enteric bacteria. Insofar as CD4+ T cell responses are induced and directed by the innate immune system, the interplay among the microbiota, the intestinal epithelial barrier and innate immune cells are critical determinants of mucosal CD4+ T cell responses. Accordingly, dysregulation at any of these levels may lead to IBD.
For most human IBD susceptibility genes, functional polymorphisms do not result in gene defects of the cellular immune system. Thus, many of the mouse models appear to be more important for identifying pathways of susceptibility rather than for establishing strict pathogenetic associations. For example, on the basis of the observation that C3H/HeJ are highly susceptible colitis mouse strains and C57BL/6 strains, in contrast, are relatively resistant, quantitative trait locus analyses in IL-10-deficient mice identified a colitis susceptibility gene locus on murine chromosome 3 (Cdcs1) that appears to regulate the magnitude of innate responses to Toll-like receptor (TLR) ligands [20, 21]. Moreover, investigation of the spontaneously ileitic SAMP1/YitFc mouse identified a susceptibility locus on chromosome 9 that encompasses candidate genes encoding IL-18 and the IL-10 receptor-chain . Only recently, another susceptibility locus in mice that includes the inflammation-modulating transcription factor peroxisome proliferator-activated receptor was reported for which susceptibility alleles have also been identified in humans .
Microbial influences on IBD pathogenesis
Pathologic interaction of the immune system and the commensal enteric bacteria provides the basis for the chronic autoimmune intestinal inflammatory process [24–27]. Multiple mechanisms may underlie this process. Classically, a pathogenic germ may lead to a state of chronic stimulation of innate and adaptive immune responses. Alternatively, a decreased number of protective-acting bacteria producing short-chain fatty acids could enhance mucosal permeability and/or—to the contrary—a preponderance of aggressive bacteria could lead to an increase in toxic metabolites, such as hydrogen sulfide, consequently also increasing mucosal permeability. The increase in mucosal permeability may result in an overwhelming exposure of bacterial TLR ligands and antigens that activate pathogenic innate and T cell immune responses. Malsecretion of secretory immunoglobulin (Ig) A can lead to mucosal bacterial overgrowth, whereas defective killing of phagocytosed bacteria may serve as a continuous antigen stimulus, leading to an exaggerated immune response. Finally, dysfunction of regulatory T cells or antigen-presenting cells can lead to a loss of tolerance to ubiquitous microbial antigens or the induction of cross-reactive autoimmune responses because of molecular mimicry between host and microbial antigens . Increased mucosally associated bacteria are evident in patients with IBD . Although this may represent a secondary phenomenon, disease-related genetic polymorphisms in the intracellular bacterial recognition receptor, NOD [2, 9, 10], TLR2 and -4, [5, 30] the autophagy gene ATG 16L1, which regulates intracellular microbial processing and killing [31, 32] and NCF4  (mediating NADPH-dependent bacterial killing in phagocytic cells) further support the role for a defective immune response against microbial antigens. The important role of gut microbial flora in IBD pathogenesis is highlighted still further by the fact that genetically engineered mice with systemic immunoregulatory defects developing IBD demonstrate an inflammation pattern that is limited to the colon that harbours the highest concentration of microbial species in the gut, and only rarely are other parts of the gastrointestinal (GI)-tract involved .
A synopsis of clinical, radiologic, endoscopic and histologic results mostly enable the two subforms of IBD to be differentiated. In patients with exclusive colonic manifestation and without typical endoscopic or histologic findings, however, the distinction between CD and UC can be impeded. Diarrhea is the leading symptom that results in the initial diagnosis of IBD. Acute phases of UC mostly go along with hematochezia, a symptom which is hardly seen in CD patients. Due to the distribution pattern of the disease, UC patients often feel pain in the area of the lower left quadrant whereas three-fourths of CD patients suffer from terminal ileitis and therefore complain about abdominal cramps in the lower right quadrant .
About half of all IBD patients suffer from extraintestinal manifestations, with about one fourth having more than one. Cutaneous lesions can occur in IBD. The most frequent of these is erythema nodosum, which manifests as an erythematous, nodular, extremely painful lesion, typically located pretibially. Pyoderma gangraenosum presents as a deep, sterile purulent ulcer or as extremely painful pustules and, similar to erythema nodosum, can occur independently of intestinal inflammatory activity in IBD. Ophthalmologic manifestations include uveitis and iritis and occur in <10% of cases, with a predominance of these in CD patients . Affections of the joints can be divided into peripheral arthritis, sacroiliitis and ankylosing spondylitis. Approximately 60% of patients with IBD develop ankylosing spondylitis if they are seropositive for HLA-B27. Biliary and liver manifestations include fatty liver degeneration, which can be followed by primary sclerosing cholangitis, cholelithiasis and hepatitis. Finally, as CD patients with affection of the small bowel may suffer from malabsorption, they are prone to developing osteopenia/-porosis, which is further complicated if those patients frequently use corticosteroids. It appears that the inflammatory activity of IBD per se is a risk factor for bone loss . Owing to ileal inflammation, chologenic diarrhea and gallstones are also more frequent in patients with CD.
To evaluate the patient’s clinical situation, clinicians commonly use indices, such as the Mayo score or colitis activity index for UC and the Crohn’s disease activity index for CD [37, 38]. These scores are calculated on the basis of symptoms, such as the number of diarrhea episodes, pain or general well-being, and they are used to define an acute episode of the disease.
The major procedure in the diagnosis of IBD is the conventional ileocolonoscopy with systematic biopsies from each anatomic segment, except in cases of fulminant colitis where a limited flexible sigmoidoscopy with biopsies may be more appropriate because of an increased risk of perforation. Patients with UC present with a uniform, continuous inflammation extending proximally, beginning from the rectum and limited to the colon; only seldomly can a “backwash ileitis” be seen. At diagnosis, in 25–55% of the cases, the disease is limited to the rectum; 50–70% of UC patients suffer from left-sided disease (inflammation reaching to the splenic flexure) . In contrast, CD can involve the entire GI tract from the lips to the anus. Inflammatory lesions are limited to the colon in one of every five patients, while about half of the patients show activity in the terminal ileum and the colon. In 25–40% of patients, the ileum is exclusively involved—a fact from which the Latin name for CD “ileitis terminalis” has been derived. Involvement of the esophagus, stomach and proximal parts of the small bowel can be observed in 1–10% of patients suffering from CD. Characteristic macroscopic features of CD are inflamed mucosal areas alternating with healthy surfaces (“skip lesions”) and small but deep “aphthous” ulcers or longitudinal (“snail-track”) ulcerations. A common additional feature of chronic disease is the appearance of a patchy, cobblestone-pattern in the terminal ileum. Extensive regenerative hyperplasia (“pseudopolyps”) of the colonic mucosa can be present in both subforms of IBD (CD < UC) and may be associated with disease duration .
Histopathologic assessment is important for confirmation of the endoscopic or clinical diagnosis of the respective IBD subform. UC is characterized by crypt infiltration, predominantly with neutrophil granulocytes, mucous depletion from the goblet cells, cryptitis with crypt abscesses and disturbed crypt architecture. In contrast, colonic goblet cells are usually intact in CD, and the mononuclear infiltrate of the mucosa shows mainly T cells and monocytic cells. Granulomas, which are characteristic for CD, are detected in the resection specimen in only 40–60% of the cases and much less frequently in biopsy samples (15–36%) [39, 40]. Another classical discriminator, the transmural (CD) versus the mucosal inflammation (UC), requires operative specimens, whereas colonic biopsies should not extend beyond the lamina propria.
We and others advocate that at the time of diagnosis of IBD the anatomic pattern of the disease should be clarified even if UC is highly probable in the clinical context. To complete staging procedures a gastroduodenoscopy should be performed complemented by a radiologic examination of the small bowel. A highly sensitive, safe and well-compliant method is the use of magnetic resonance imaging (MRI) enteroclysis, and its sensitivity in detecting stenoses in CD patients with small bowel involvement is comparable to the results with conventional small bowel follow-through . Moreover, MRI is valuable diagnostic tool for differentiating active inflammation from fibrosis as it can detect differential water content within the gut wall. This capability enables the clinician to distinguish between inflammatory and fixed fibrostenotic lesions in CD and has, therefore, important therapeutic implications . High-resolution MRI has been demonstrated to be more efficient than computed tomography (CT) scanning, but the former requires the availability of high-resolution equipment. Transabdominal sonography, due to a steadily improving resolution, can detect inflamed areas of the small bowel as well as of the colon with high diagnostic precision . The diagnosis of inflammation is based on an assessment of the bowel-wall diameter and blood flow. The diameter of the bowel wall has a good predictive value for differentiating quiescent from moderate to severe disease in the terminal ileum and in the colon proximal to the rectum, which can be easily assessed by ultrasound [44, 45]. Other pathologic findings, such as enlarged lymph nodes, abscesses, stenoses and even fistulae, can be detected by sonography. Depending on the operator, sensitivity for the primary detection of active CD has been determined to be 75–94%, with a specificity of 67–100% . The high sensitivity (74%) of transabdominal ultrasound to detect strictures in patients with CD having suggestive symptoms can be further increased through bowel preparation with polyethylene glycol (PEG; 89%) to levels comparable with barium enteroclysis and ileocolonoscopy .
Diagnosis of infectious complications
All patients with a new possible IBD diagnosis or an exacerbation of disease formerly in remission should be tested for routine bacterial enteric pathogens (e.g. Shigella, Campylobacter, Escherichia coli O157:H7, Salmonella) and, depending on the individual anamnesis, testing for parasitic infections may also be appropriate. Additionally, the presence of cytomegalovirus (CMV) infections should be excluded in exacerbated or refractory IBD patients with colitis. The frequency of symptomatic CMV infection is estimated to range between 0.53 and 3.4% in all IBD patients, and the extent of colitis in affected patients may range from proctitis to pancolitis [48–50]. The identification of CMV infection has a dramatic impact on the patient’s disease course. Treatment of CMV colitis involves tapering off steroids and other immunosuppression as rapidly as possible while treating the patient with an antiviral agent. Clinical trials have demonstrated that patients treated with ganciclovir avoided colectomy and entered clinical remission [48, 51]. A flexible sigmoidoscopy with biopsies is usually sufficient to diagnose CMV colitis, and the diagnostic gold standard remains the histologic diagnosis of CMV [48, 50]. Microscopic examination reveals characteristically enlarged cells with intranuclear inclusion bodies; if applicable, immunohistochemical staining should also be performed to increase diagnostic sensitivity. Another important pathogen that has been associated with exacerbations of IBD is Clostridium difficile [52–55], and C. difficile infections among IBD patients with flares may be as high as 28% [53, 54]. Interestingly, C. difficile infection may be detected in IBD patients even in the absence of recent antibiotic use [53, 54]. The best diagnostic test is the assay for both C. difficile toxins A and B in the stool [56, 57]. Treatment consists of oral metronidazole or vancomycin, with intravenous (i.v.) metronidazole as an alternative.
Innovations in the diagnostics of IBD
Novel serologic markers may improve the diagnosis of IBD, particularly in difficult cases. Anti-Saccharomyces cerevisiae antibodies (ASCA), which are directed against Candida albicans, and perinuclear anti-neutrophil cytoplasmic antibodies (pANCA) are used to increase the diagnostic precision and discriminative ability to differentiate between UC and CD . In a meta-analysis of 60 studies involving 3841 UC and 4019 CD patients, the positive detection of ASCA in combination with the lack of pANCA resulted in 55% sensitivity for the detection of CD (specificity, 93%); the sensitivity and specificity of pANCA detection for diagnosis of UC were 55 and 89%, respectively. A further large study demonstrated a high specificity of an ASCA/pANCA panel, but the sensitivity was low . Population-representative studies have revealed a particular strength of ASCA in the prediction of a complicated, severe course of disease . Novel antimicrobial antibodies in the serum of IBD patients include the CD-related protein from Pseudomonas fluorescens (anti-I2), the flagellin-like antigen (anti-Cbir1) and E. coli outer membrane porin C (anti-OmpC) [61–64] I2 antibodies; these can be detected in about 50% of CD patients but in only 10% of UC patients [65, 66]. Anti-CBir1 expression is associated independently with small bowel, intestinal penetrating and fibrostenosing disease. Antibodies against OmpC can be found in about 50% of CD patients but in only a small percentage of UC patients [66–70].
Markers of mucosal function may also be used to assess disease activity and predict the clinical course; however, controversial data and their impracticability in clinical practice limit their use. One example is mucosal tumor necrosis factor (TNF) release, which has been shown to be a predictor of relapse while CD patients are in remission . Another useful marker may be fecal calprotectin, which is an antibiotic, cytoplasmatic protein released by activated neutrophilic polymorphonuclear cells and/or monocytes/macrophages or during cell death. Fecal calprotectin is a stable marker that reflects the degree of neutrophil-driven intestinal inflammation and therefore allows differentiation between IBD and irritable bowel syndrome . In remission, fecal calprotectin levels are elevated in patients with IBD in comparison with those in normal healthy individuals, as has been confirmed by a meta-analysis of 5983 patients with IBD, colorectal cancer and healthy normal controls from 30 studies. Elevated calprotectin levels were shown to precede an imminent relapse [73, 74].
Since intestinal permeability in patients with CD is increased and correlates with inflammatory alterations of small bowel mucosa, permeability tests with macromolecular sugars may have prognostic implications [75, 76]. Increased intestinal permeability during remission has been demonstrated to precede relapse in CD ; however, results from PEG-based assessments of permeability are conflicting [77–79]. In summary, to date, the diagnostic strategy to use the additional diagnostic markers discussed here has not yet been satisfactorily elucidated.
Medical treatment of IBD
Individual treatment of patients with IBD is dependent on several distinct factors, including disease location and severity. Therapy is sequential, with the first priority to treat acute disease, followed by the maintenance of remission.
Aminosalicylates and corticosteroids
Topical corticosteroids have been found to be less effective than topical aminosalicylates for inducing the remission of distal UC [80, 81]. However, the combination of topical corticosteroids with topical salicylates is often more efficacious than either alone in the short-term treatment of patients with distal UC . Foam preparations are often better tolerated by patients and may be easier to retain . Some studies have shown topical budesonide to be as efficacious as prednisolone enemas , mesalamine enemas  and systemic corticosteroids  in the treatment of patients with distal UC. Oral controlled ileal release (CIR) budesonide was shown to be ineffective in patients with distal UC . The use of topical plus oral salicylates seems to be appropriate in cases of extensive colonic affection beyond the left colonic flexure. For refractory cases, further therapeutic exaggeration comprises the use of systemic steroids, which is indicated in the initial phase of severe cases. When sulfasalazine 8 g/day was compared with prednisolone 40 mg/day, the steroid was found to induce higher remission rates in patients with moderate to severe UC . After a successful induction of remission, aminosalicylates serve as maintenance therapy of UC and should be given for a minimum of 2 years . In cases of mild to moderate CD, an initial attempt with oral aminosalicylates (3–4 g/day) is justified according to current data. If therapeutic response is lacking, the early use of corticosteroids is indicated. CIR budesonide (9 mg/day; Entocort)—instead of systemic steroids—is indicated for inducing remission in patients with mild to moderate ileal and right-sided colonic CD. CIR budesonide (9 mg/day) has been shown to be more effective than mesalamine (at a dosage of 4 g/day; 69 vs. 45% at 8 weeks) for induction of remission of ileocolonic CD . Trials comparing oral budesonide with prednisolone show a comparable efficacy for inducing remission in active CD [91–94]. However, in the cohort of patients with moderate to severe active CD, budesonide was found to be slightly less active than prednisolone . The addition of sulfasalazine to corticosteroid therapy does not result in a better efficacy compared with only corticosteroid treatment . Corticosteroids, however, are not effective for the treatment of the subpopulation of patients with perianal fistulas, for whom treatment with antibiotics, azathioprine (Aza), methotrexate (MTX), cyclosporine or infliximab or, ultimately, surgery is superior. Dosages in the range of 40–60 mg/day or 1 mg/kg/day of prednisone or equivalent are effective for inducing remission. Following the induction of remission, steroids should be tapered; for prednisolone, a tapering range between 5 mg/week and 10 mg every 10 days (slower <20 mg/day) seems to be reasonable based on studies and anecdotal experience. Corticosteroids have not been shown to be effective for the maintenance of remission, and the risks of long-term therapy outweigh their benefits [97–99]. Patients with severe to fulminant IBD who fail to respond to oral steroid medication require hospitalization and should be given i.v. steroids. In patients with severe UC, prednisolone i.v. at a dosage of 60 mg/day was found to induce remission in 64% of patients; an additional 13% showed improvement, while 23% had no response and required colectomy [100, 101]. Randomized controlled trials of corticosteroids in patients with severe CD have not been performed, but it is likely that the remission rate in CD for parenteral corticosteroids is similar to that in UC. The prolongation of therapy with parenteral corticosteroids beyond 7–10 days has not been proven to add any benefit and may in fact be deleterious [100–103].
The immunosuppressives Aza and its metabolite 6-mercaptopurine (6-MP) are used to treat patients with CD and UC with the aim to withdraw corticosteroids in cases refractory to or dependent on corticosteroids. However, it has to be kept in mind that the onset of full activity of Aza and 6-MP is slow and takes up to 3 months. The most common side effects are bone marrow suppression and pancreatitis [104–107]. Whereas bone marrow suppression may occur at any time during treatment, pancreatitis usually occurs within the first 4 weeks of initiating therapy. The fourfold increased risk of developing Aza/6-MP-induced lymphoma remains one of the most common concerns [108, 109]. Whether the dose should be escalating or immediately at the weight-calculated dose upon treatment initiation is not sufficiently clear based on currently available experimental data. A dose-escalating method, however, offers the possibility for an early recognition of leukopenia and other potential adverse events and is thus commonly used in clinical daily routine. Aza may be started at 50 mg daily and the dose increased by 25 mg every 1–2 weeks to a target dose of 2.0–3.0 mg/kg (6-MP 1.0–1.5 mg/kg/day, respectively). Subjects who are homozygous for the Aza degrading enzyme thiopurine methyltransferase (TPMT) mutation should be placed on other forms of therapy due to the high risk of myelosuppression.
Current evidence supports the use of MTX for the same indications as for Aza/6-MP. In contrast, MTX is indicated for the induction of remission in patients with active CD and for the maintenance of remission in patients with inactive CD, but not for the induction or maintenance of remission in UC. Due to bone marrow- and hepato-toxicity, routine monitoring of laboratory parameters, including complete blood counts and liver-associated laboratory chemistries, is recommended. MTX may induce hepatic fibrosis, thus patients with persistently abnormal liver-associated chemistries should either discontinue therapy with MTX or undergo liver biopsy. Due to its embryotoxicity it is absolutely contraindicated in pregnancy, thus limiting its use in young women .
Cyclosporine is effective in the treatment of severe UC; concomitant administration of i.v. corticosteroids is recommended, but not required. While cyclosporine has not proven to be efficacious in patients with luminal CD, it is effective for the treatment of patients with fistulizing CD. Response or remission induced with i.v. cyclosporine typically requires a continuation of therapy with oral cyclosporine for a few months, along with a tapering dose of corticosteroids, initiation of AZA or 6-MP therapy and prophylaxis against Pneumocystis carinii. AZA or 6-MP should be continued as maintenance therapy .
Tumor necrosis factor-α blockers
Tumor necrosis factor alpha (TNF-α) is a proinflammatory cytokine with a central role in the pathogenesis of CD . Infliximab is a chimeric monoclonal antibody directed against TNF-α that was introduced into clinical practice in the USA almost 12 years ago. Infliximab was the first TNF-blocker used for the treatment of moderately to severely active CD and UC in patients who do not respond despite complete and adequate therapy with a corticosteroid or an immunosuppressive agent. In the case of nonresponse to the standard three infusions at time 0, 2 and 6 weeks, further treatment with infliximab is not recommended. If remission is achieved, an attempt to withdraw or taper any concomitant corticosteroid therapy is appropriate. Treatment with corticosteroids, AZA, 6-MP, or MTX prior to and concomitantly with infliximab can reduce the formation of antibodies against infliximab. After successful remission, maintenance therapy is recommended, comprising an infusion every 8 weeks. In the case of a loss of effect, these intervals should be shortened up to 4 weeks, and if the patient is still therapy refractory, an increase in dose to 10 mg/kg daily can be considered . The use of infliximab is also recommended for the treatment of CD with fistulas. Before the administration of infliximab, delineation of the fistula anatomy is useful to exclude the presence of an abscess . Abscesses should be drained adequately before treatment with infliximab. Due to an increased risk of infection under immunomodulatory therapy with infliximab, it is contraindicated in patients with active tuberculosis and other serious infections or opportunistic infections.
Two additional anti-TNF molecules are currently used to treat IBD: adalimumab and certolizumab pegol . To date, evidence for efficacy presented in clinical trials has been more robust for infliximab than for the other anti-TNFs. Sequential use of different anti-TNF agents in patients who have responded well to a first agent is applicable. Further clinical data is needed to decide whether early therapy with TNF antibodies is generally efficient.
Mycophenolatemofetile and Tacrolimus
Mycophenolatemofetile (MMF) inhibits lymphocyte proliferation by selectively blocking the synthesis of guanosine nucleotide in T cells . At the present time, there are inadequate data to suggest that individuals with IBD should be treated with MMF, and the safety concerns, including reports of severe infections and induction of a colitis with histologic changes similar to those seen in graft-versus-host disease or CD, make MMF dispensable [115–117]. Data received from clinical trials using tacrolimus for CD and MC is controversial, and larger controlled studies are needed before its use can be recommended.
Future therapeutic options
Targeting CD4+ T cell cytokines
Naive CD4+ T cells differentiate into several functional lineages characterized mainly by their dependent cytokines. CD4+ T cell phenotypes include T-helper (Th) 1, Th2, Th17 and CD4+ T-regulatory cells. Their differentiation and survival depends on the relative abundance of key regulatory cytokines produced mainly by macrophages and dendritic cells. IL-12 and IL-23 have been implicated in the pathogenesis of CD. In the presence of IL-12, which is a heterodimer of p40 and p35 subunits, naive CD4+ T cells adopt a Th1 phenotype and produce interferon (IFN)-γ to mediate cellular immunity. IL-12 and IFN-γ act in a positive feedback loop . In the presence of IL-6, transforming growth factor-β and IL-23 (a heterodimer of the same p40 subunit as IL-12 and a unique p19 subunit), naive human CD4+ T cells adopt a Th17 cell profile, characterized by the unique production of IL-17A, IL-17F, IL-22 and IL-21 to mediate cellular immunity . It has been shown that IL-23p19-deficient mice, but not IL-12p35-deficient mice, are protected against the onset of intestinal inflammation in the IL-10 −/− model . In accordance with these data, IL-17R-deficient mice were found to be significantly protected against TNBS-induced colitis . Underlining the prominent role of CD4+ Th17 cells in the pathogenesis of IBD, the transfer of bacterial reactive CD4+ Th17 cells was shown to induce a colitis in severe combined immunodeficient mice that was much more severe than that induced after the transfer of CD4+ Th1 cells . Human data reveal a significant increase in intestinal IL-23 mRNA and IL-17 serum and mucosal levels in IBD patients [121, 122]. Further, in CD patients receiving the α-IL-12p40 antibody, lamina propria mononuclear cells produced IL-23, and T cell-derived IL-17 concentrations were considerably suppressed after the treatment . A better understanding and characterization of the IL-12/IFN-γ and the IL-23/IL-17 axis would help in developing novel therapeutic strategies. Both pathways have been targeted by selective biological agents with large trials being conducted currently.
IL-6, which is secreted by lamina propria T cells and macrophages, is a double-edged sword since it acts as both a proinflammatory and anti-inflammatory cytokine [123–125]. Serum IL-6 and mucosal IL-6 mRNA levels are considered to be clinically relevant parameters, as they correlate with the inflammatory activity and the frequency of relapses in CD patients during remission [126–129]. This implies that IL-6 affects systemic events while simultaneously being locally involved in mucosal disease pathogenesis. Interestingly, the biological function of IL-6 in IBD is not mediated through the membrane-bound receptor for IL-6 (IL-6R)  but through binding to a soluble form of its receptor (sIL-6R), which is generated by shedding from the surface of macrophages [130, 131]. Intriguingly, the complex of IL-6/sIL-6R then activates gp130-positive T cells lacking the membrane-bound IL-6R, a phenomenon called trans-signaling . In IBD, IL-6 trans-signaling leads to the translocation of the signal transducer and activator of transcription (STAT)-3 and the subsequent induction of the antiapoptotic genes Bcl-2 and Bcl-xl in lamina propria T cells . This pathway has been shown to confer resistance against intestinal T cell apoptosis in experimental models of colitis as well as in IBD patients [123, 132]. In the experimental trinitrobenzene sulfonic acid (TNBS) colitis model, sgp130Fc (which selectively binds the sIL-6R) shows a therapeutic effect that is similar to that of the IL-6R antibody, indicating that blockage of the sIL-6R is pivotal for therapeutic efficacy . Since activated intestinal T cells in IBD normally do not express the membrane-bound form of the IL-6R, blocking IL-6 trans-signalling may be a novel therapeutic approach . Tocilizumab is a humanized monoclonal antibody that blocks both the membrane-bound and the soluble IL-6 receptor. The results of a first clinical trial are promising: 80% of the patients given tocilizumab infusions biweekly for 12 weeks showed a significantly higher clinical response rate than the placebo group . There is a need for a large placebo-controlled trial evaluating this therapy in patients with IBD.
In contrast to the cytokine-targeted therapies described above, the administration of IL-10, IL-11 and IFN-α and -β has shown no success in clinical trials of IBD . Reagents blocking the CD4 T cell receptor were found to cause long-term depletion of lymphocytes in some patients and, therefore, are not applicable as a true therapeutic alternative . Anti-CD3 monoclonal antibodies showed promising results in phase I and II trials; however, a phase III, randomized, double-blind, placebo-controlled, multicenter study in subjects with i.v. steroid-refractory UC was withdrawn because an interim analysis showed no difference in colectomy rates for the anti-CD3 monoclonal antibody visilizumab vs. placebo. The results from pilot trials with blockers of the IL-2 receptor CD25 do not suggest that there is a role for anti-CD25 antibodies in the treatment of IBD .
Blocking leukocyte adhesion
Natalizumab is a member of a new class of molecules known as selective adhesion molecule inhibitors and is currently approved for clinical use in the USA. It is a recombinant humanized IgG4 monoclonal antibody to α4 integrin that blocks adhesion and subsequent leukocyte migration into the gut. Another seminal substance, the humanized anti-α4β7-integrin antibody vedolizumab, inhibits only gut-specific Madcam-1-mediated leukocyte adhesion . The results from the first clinical trials are promising for both substances, and they are currently being tested in controlled large clinical trials.
A large number of stem cells are present in intestinal crypts and may thus be an interesting target for therapeutic gene transfer. The successful genetic modification of intestinal stem cells has therefore outstanding clinical potential. In vitro and in vivo studies with human and rodent cell lines and animal models using reporter genes have demonstrated that transduction of the intestinal mucosa by local administration of liposomal [135–137] retroviral [138, 139], lentiviral , adeno-associated (AAV) viral [141, 142] and adenoviral [143–147] vector systems is feasible. It has been nicely shown that oral administration of AAV vectors yields the stable transduction of the gut epithelium more than 6 months postinfection . Transplantation of in vitro-transfected intestinal stem cells may be an important method of delivering therapeutic genes to the gut [148–150]. An earlier study demonstrated that the treatment of T cell-dependent experimental colitis in SCID mice by local administration of an adenovirus expressing IL-18 antisense mRNA is effective. Moreover, TGF-β gene transfer may prevent the formation of murine experimental colitis; this therapy was effective in ameliorating established disease . Three studies have demonstrated the prevention of colonic inflammation in experimental TNBS colitis in rats, mice or IL-10-deficient mice after systemic administration of recombinant Ad5 encoding IL-10 [152–154]. Adoptive transfer of IL-10-transduced T cells has recently been used successfully to treat colitis in the murine CD4+ CD45RBhigh SCID transfer model . Since systemic injection of recombinant IL-10 has shown only a modest potency in patients with active CD [156, 157] it is interesting to speculate on whether the local delivery of immuno-regulatory T cells stably producing IL-10 would result in a better outcome.
With growing knowledge about pathophysiological mechanisms in IBD, therapeutic diversity is constantly growing with increasing success. However, disappointing results from different clinical trials suggest that a monotherapeutic concept often does not live up to the complexity of the multifactorial inflammatory process in IBD. Thus, combinations of drugs that alter the aberrant immune response in the gut through different mechanisms may have a valuable therapeutic potential by exerting synergistic effects and, hopefully, at the same time improve the safety profile of such combined approaches. Testing combinatory approaches should thus be intensified in basic research as well as in clinical trials.