European evidence-based Consensus on the management of ulcerative colitis: Special situations

8.1.1. Symptoms 

After total proctocolectomy with IPAA, median stool frequency is 4 to 8 bowel movements1, 2, 3, 4, 13, 14 with 700 mL of semiformed/liquid stool per day2, 13, 14. Symptoms related to pouchitis include increased stool frequency and liquidity, abdominal cramping, urgency, tenesmus and pelvic discomfort (2, 15). Rectal bleeding, fever, or extraintestinal manifestations may occur. Rectal bleeding is more often related to inflammation of the rectal cuff (“cuffitis”),16 than to pouchitis. Poor faecal incontinence may occur in the absence of pouchitis after IPAA, but is more common in patients with pouchitis. Symptoms of pouch dysfunction in patients with IPAA may be caused by conditions other than pouchitis, including Crohn's disease of the pouch,17, 18, 19 cuffitis,16 and an irritable pouch.20 This is why the diagnosis depends on endoscopy and biopsy in conjunction with symptoms.

8.1.2. Endoscopy (‘pouchoscopy’) 

Pouchoscopy and pouch mucosal biopsy should be performed in patients with symptoms compatible with pouchitis, in order to confirm the diagnosis.15 Patients with an ileoanal pouch occasionally have a stricture at the pouch-anal anastomosis, so a gastroscope rather than a colonoscope may better be necessary for pouchoscopy. Endoscopic findings compatible with pouchitis include diffuse erythema caused by inflammation of the ileal pouch, which may be patchy, unlike that observed in UC. Characteristic endoscopic findings include oedema, granularity, friability, spontaneous or contact bleeding, loss of vascular pattern, mucous exudates, haemorrhage, erosions and ulceration.17 Erosions and/or ulcers along the staple line do not necessarily indicate pouchitis. Characteristically, these findings are non-specific and lesions may be discontinuous, unlike the colorectal lesions in UC.18, 21, 22 Biopsies should be taken from the pouch mucosa and from the afferent limb above the pouch, but not along the staple line.

8.1.3. Histopathology of pouchitis 

Histological findings of pouchitis are also non-specific, including acute inflammation with polymorphonuclear leukocyte infiltration, crypt abscesses and ulceration, in association with a chronic inflammatory infiltrate.21, 22 There may be discrepancy between endoscopic and histologic findings in pouchitis, possibly related to sampling error.23, 24 Morphological changes of the epithelium lining the ileal pouch normally develop in the 12–18 months after ileostomy closure, characterised by flattening and a reduced number, or disappearance of the villi, leading to villous atrophy (“colonic metaplasia”).22, 23, 24 Although the aetiology of pouchitis remains unknown, it can be inferred from the predeliction for patients with UC and the response to antibiotic therapy that the bacterial flora and whatever predisposes to UC itself are involved in the pathogenesis of tissue damage in the ileoanal pouch.25, 26 Pouchitis tends to occur only after colonic metaplasia has developed in the pouch, although a causal association is unproven.

8.1.4. Differential diagnosis 

The clinical history and biopsies help discriminate between pouchitis, ischaemia, Crohn's disease (CD) and other rare forms of pouch dysfunction such as collagenous pouchitis, Clostridium difficile or cytomegalovirus pouchitis.27, 28, 29 Secondary pouchitis, caused by pelvic sepsis, usually causes focal inflammation and should be considered. Biopsies taken from the ileum above the pouch may reveal pre-pouch ileitis as a cause of pouch dysfunction, although this usually causes visible ulceration that may be confused with Crohn's disease.30 The possibility of non-specific ileitis caused by NSAIDs should be considered.31

8.1.5. Risk factors for pouchitis and pouch dysfunction 

Reported risk factors for pouchitis include extensive UC,1, 32 backwash ileitis,32 extraintestinal manifestations (especially primary sclerosing cholangitis),5, 19, 33 being a non-smoker34 and regular use of NSAIDs.31, 35 Interleukin-1 receptor antagonist gene polymorphisms36 and the presence of perinuclear neutrophil cytoplasmic antibodies37 are also associated with pouchitis. Not surprisingly studies are discordant with regard to the role of each risk factor. Some of the best data on risk factors come from the Cleveland Clinic.38 240 consecutive patients were classified as having healthy pouches (n=49), pouchitis (n=61), Crohn's disease (n=39), cuffitis (n=41), or irritable pouch syndrome (n=50). The risk of developing pouchitis was increased 3–5 fold when the indication for IPAA was dysplasia (OR 3.89; 95% CI 1.69–8.98), or when the patient had never smoked (OR 5.09; 95% CI 1.01–25.69), or used NSAIDs (OR 3.24; 95% CI 1.71–6.13), or (perhaps surprisingly) had never used anxiolytics (OR 5.19; 95% CI 1.45–18.59). The risk of turning out to have Crohn's disease of the pouch was greatly increased by being a current smoker (OR 4.77; 95% CI, 1.39–16–25), and modestly associated by having a pouch of long duration (OR 1.20; 95% CI 1.12.–1.30). Cuffitis was associated with symptoms of arthralgia (OR 4.13; 95% CI 1.91–8.94) and a younger age (OR 1.16; 95% CI 1.01–1.33). Irritable pouch syndrome is probably under-recognised, although is a common cause of pouch dysfunction when other causes (including a small volume pouch, incomplete evacuation and pouch volvulus) have been excluded and investigations are normal. The principal risk factor is the use of antidepressants (OR 4.17; 95% CI 1.95–8.92) or anxiolytics (OR 3.21; 95% CI 1.34–7.47), which suggests that these people may have had irritable bowel syndrome contributing to symptoms of colitis before pouch surgery.38

These risk factors should not preclude proctocolectomy if surgery is appropriate, but should inform pre-operative discussions with the patient and family. In particular the possibility that IBS may be contributing to symptoms of refractory UC should be considered and objective evidence of treatment refractory colitis obtained before surgery. If there is a disparity between preoperative and endoscopic appearance, or if the patient is on antidepressants, then the risk of pouch dysfunction after IPAA needs particularly careful consideration. Similarly, if a patient has primary sclerosing cholangitis, then it is appropriate to discuss the higher risk of pouchitis. This is appropriate management of expectations rather than a contraindication to appropriate surgery.

8.2.1. Acute and chronic pouchitis 

On the basis of symptoms and endoscopy, pouchitis can be divided into remission (normal pouch frequency) or active pouchitis (increased frequency with endoscopic appearances and histology consistent with pouchitis).15, 39 Active pouchitis may then be divided into acute or chronic, depending on the symptom duration. The threshold for chronicity is a symptom duration of >4 weeks. Up to 10% of patients develop chronic pouchitis requiring long-term treatment, and a small subgroup has pouchitis refractory to medical treatment.3

8.2.2. Scoring of pouchitis 

The Pouchitis Disease Activity Index (PDAI) has been developed to standardize diagnostic criteria and assess the severity of pouchitis.15, 39, 40 The PDAI is a composite score that evaluates symptoms, endoscopy and histology. Each component score has a maximum of 6 points. Patients with a total PDAI score ≥7 are classified as having pouchitis, so a patient has to have both symptoms and endoscopic or histological evidence of pouchitis and, ideally, all three. The problem is that about a quarter of patients with a high symptom score suggestive of pouchitis may not fulfil criteria for the diagnosis of pouchitis, as assessed by the PDAI, since endoscopic or histological criteria may be absent. Consequently a relatively large number of patients may be unnecessarily treated for puchitis when symptoms are due to other conditions. Other scoring systems have been devised, including that by Moskowitz21 and an index from Heidelberg. Comparisons with the PDAI41, 42 show that they are not interchangeable, but this affects clinical trials rather than clinical practice.

8.2.3. Recurrent pouchitis and complications 

Pouchitis recurs in more than 50%.3, 15, 39 Patients with recurrent pouchitis can broadly be grouped into three categories: infrequent episodes (<1/yr), a relapsing course (1–3 episodes/yr) or a continuous course. Pouchitis may further be termed treatment responsive or refractory, based on response to single-antibiotic therapy (see 8.3.2).7, 9 Although these distinctions are largely arbitrary, they help both patients and their physicians when considering management options to alter the pattern of pouchitis. Complications of pouchitis include abscesses, fistulae, stenosis of the pouch-anal anastomosis and adenocarcinoma of the pouch.7, 27, 39 This latter complication is exceptional and almost only occurs when there is pre-exiting dysplasia or carcinoma in the original colectomy specimen.

8.3.1. Acute pouchitis: antibiotics 

Treatment of pouchitis is largely empirical and only small placebo-controlled trials have been conducted. Antibiotics are the mainstay of treatment, and metronidazole and ciprofloxacin are the most common initial approaches, often with a rapid response. The odds ratio of inducing a response using oral metronidazole compared with placebo in active chronic pouchitis was 26.67 (95% CI 2.31–308.01, NNT=2).43 The randomised trials of both metronidazole and ciprofloxacin are, however, small.44, 45, 46 The two have been compared in another small randomised trial.47 Seven patients received ciprofloxacin 1 g/day and nine patients metronidazole 20 mg/kg/day for a period of 2 weeks. Ciprofloxacin lowered the PDAI score from 10.1±2.3 to 3.3±1.7 (p=0.0001), whereas metronidazole reduced the PDAI score from 9.7±2.3 to 5.8±1.7 (p=0.0002). There was a significantly greater reduction in the ciprofloxacin compared to metronidazole in terms of the total PDAI (p=0.002), symptom score (p=0.03) and endoscopic score (p=0.03), as well as fewer adverse events (33% of metronidazole-treated patients reported side-effects, but none on ciprofloxacin). Combination antibiotic therapy is an option for persistent symptoms (below).

8.3.2. Chronic pouchitis: combination antibiotic therapy or budesonide 

	ECCO Statement 8D
	In chronic pouchitis, combined antibiotic treatment is effective [EL1b, RG B]

For patients who have persistent symptoms, alternative diagnoses should be considered, including undiagnosed Crohn's disease, pouch-anal or ileal-pouch stricture, infection with CMV or Cl difficile, collagenous pouchitis, cuffitis, anatomical disorders, or irritable pouch syndrome. Approximately 10–15% of patients with acute pouchitis develop chronic pouchitis, which may be “treatment responsive” or “treatment refractory” to single-antibiotic therapy.39 Patients with chronic, refractory pouchitis do not respond to conventional therapy and often continue to suffer symptoms, which is a common cause of pouch failure. Combination antibiotic therapy or oral budesonide may work. 16 consecutive patients with chronic, refractory pouchitis (disease >4 weeks and failure to respond to >4 weeks of single-antibiotic therapy) were treated with ciprofloxacin 1 g/day and tinidazole 15 mg/kg/day for 4 weeks.47 A historic cohort of ten consecutive patients with chronic refractory pouchitis treated with 5–8 g oral and topical mesalazine daily was used as a comparator. These refractory patients had a significant reduction in the total PDAI score and a significant improvement in quality-of-life score (p<0.002) when taking ciprofloxacin and tinidazole, compared to baseline. The rate of clinical remission in the antibiotic group was 87.5% and for the mesalazine group was 50%.

In another study, 18 patients non-responders to metronidazole, ciprofloxacin or amoxycillin/clavulanic acid for 4 weeks were treated orally with rifaximin 2 g/day (a nonabsorbable, broad spectrum antibiotic) and ciprofloxacin 1 g/day for 15 days. Sixteen out of 18 patients (88.8%) either improved (n=10) or went into remission (n=6).48 Median PDAI scores before and after therapy were 11 (range 9–17) and 4 (range 0–16), respectively (p<0.002). A British group observed similar benefit in just 8 patients.49 In another combination study, 44 patients with refractory pouchitis received metronidazole 800 mg–1 g/day and ciprofloxacin 1 g/day for 28 days.50 36 patients (82%) went into remission and median PDAI scores before and after therapy were 12 and 3 respectively (p<0.0001). The alternative is oral budesonide CIR 9 mg daily for 8 weeks, which achieved remission in 15/20 (75%) patients not responding after 1 month of ciprofloxacin or metronidazole.51 The message is simple enough, even if the trials are underpowered: if ciprofloxacin does not work, then try it in combination with an imidazole antibiotic or rifaximin, or try oral budesonide.

8.3.3. Acute and chronic refractory pouchitis: other agents 

The variety of approaches illustrates the challenges of trying to find treatment that works for a new disorder. These are largely of historic interest. Budesonide enemas were as effective as metronidazole for acute pouchitis in a randomised controlled trial.52 Ciclosporin enemas were successful for chronic pouchitis in a pilot study53 and oral azathioprine may help if patients relapse become budesonide-dependent. Uncontrolled studies of short-chain fatty acid enemas54, 55 showed little value. Glutamine and butyrate suppositories have been compared for chronic pouchitis.56 Of more recent interest, infliximab has been tried in patients with chronic, (very) refractory pouchitis not responding either to metronidazole or ciprofloxacin 1 g/day for 4 weeks or oral budesonide CIR 9 mg/day for 8 weeks. 10/12 (83.3%) such patients treated with infliximab 5 mg/kg at 0, 2 and 6 weeks went into remission.57 The median PDAI score before therapy was 13 (range 8–18) and 2 (range 0–9) after infliximab (p<0.001) and the IBDQ also significantly improved from 96 (range 74–184) to 196 (92–230) (p<0.001).57 Infliximab has been used when the cause of pouch dysfunction is Crohn's disease, or fistulation.58 Benefit been also been reported from alicaforsen enemas (an inhibitor of intercellular adhesion molecule (ICAM)-1) in an open-label trial.59

8.3.4. Maintenance of remission: probiotics 

	ECCO Statement 8E
	VSL#3 (18×1011 of 8 bacterial strains for 9 or 12 months) has shown efficacy for maintaining antibiotic-induced remission [EL1b, RG B]. VSL#3 (9×1011 bacteria) has also shown efficacy for preventing pouchitis [EL2b, RG C]

In chronic pouchitis, once remission has been obtained, treatment with the highly concentrated probiotic mixture VSL#3 is able to maintain remission. Two double-blind, placebo-controlled studies have shown the efficacy of VSL#3 (450 billion bacteria of 8 different strains/g) to maintain remission in patients with chronic pouchitis. In the first study, 40 patients who achieved clinical and endoscopic remission after one month of combined antibiotic treatment (rifaximin 2 g/day+ciprofloxacin 1 g/day), were randomised to receive either VSL#3, 6 g/day (18×1011 bacteria/day), or placebo for 9 months.60 All 20 patients who received placebo relapsed, while 17 of the 20 patients (85%) treated with VSL#3 remained in clinical and endoscopic remission at the end of the study. Interestingly, all 17 patients relapsed within four months after stopping VSL#3.60 In the second study, 36 patients with chronic, refractory pouchitis who achieved remission (PDAI=0) after 1 month of combined antibiotic treatment (metronidazole+ciprofloxacin) received 6 g/once a day of VSL#3 or placebo for 1 year. Remission rates at one year were 85% in the VSL#3 group and 6% in the placebo group (p<0.001).61

8.3.5. Prevention of pouchitis: probiotics 

The same probiotic preparation (VSL#3) has been shown to prevent pouchitis within the first year after surgery, in a randomised, double-blind, placebo-controlled study. Forty consecutive patients undergoing IPAA for UC were randomised within a week after ileostomy closure, to VSL#3 3 g (9×1011) per day or placebo for 12 months. Patients were assessed clinically, endoscopically and histologically at 1, 3, 6, 9 and 12 months. Patients treated with VSL#3 had a significantly lower incidence of acute pouchitis (10%) compared with those treated with placebo (40%) (p<0.05), and experienced a significant improvement of quality of life.62 The mechanism by which therapy with probiotics works remains elusive, but has been investigated.63 Mucosa-associated pouch microbiota before and after therapy with VSL#3 shows that patients who develop pouchitis while treated with placebo have low bacterial and high fungal diversity. Bacterial diversity was increased and fungal diversity was reduced in patients in remission maintained with VSL#3 (p=0.001). Real time PCR experiments demonstrated that VSL#3 increased the total number of bacterial cells (p=0.002) and modified the spectrum of bacteria towards anaerobic species. Taxa-specific clone libraries showed that the spectrum of Lactobacillus sp. and Bifidobacter sp. was altered on probiotic therapy. The diversity of the fungal flora was repressed. Restoration of the integrity of a “protective” intestinal mucosa related microbiota could therefore be a potential mechanism of probiotic bacteria in inflammatory barrier diseases of the lower gastrointestinal tract.

9.1.1. Estimation of risk 

	ECCO Statement 9A
	Patients with longstanding ulcerative colitis appear to have an increased risk of colorectal cancer (CRC) as compared to the general population [EL2]

Patients with long standing UC have a higher risk of developing colorectal carcinoma (CRC) than the average population. The magnitude of this risk, however, is still the subject of a debate. Indeed, while older reports included in two meta-analyses65, 66 confirmed a rapid increase of the risk after ten years of disease, the magnitude of the risk in recent population-based studies appears much smaller.67, 68 In fact, although Eaden and colleagues computed a cumulative CRC risk of 18% in UC patients after 30 years of disease, risks of only 7.5% and 2.1% respectively were observed in two studies published since 2004.67, 68 Furthermore, in the largest report of surveillance colonoscopy in at-risk population of patients with extensive UC to date (600 patients over a 30 year period), the cumulative incidence of CRC by colitis duration was 2.5% at 20 years, 7.6% at 30 years, and 10.8% at 40 years.69 In this study from St Mark's, only 30/600 patients (5%) developed CRC. The reasons for such an improvement in the risk of CRC are still unclear but may include improved control of mucosal inflammation, more extensive use of 5-ASA compounds, the implementation of surveillance programmes and timely colectomy.70 Taken together these studies suggest that the CRC risk in UC patients should be kept under scrutiny. Nevertheless, the best evidence, as provided by concordant meta-analyses, indicates that the risk of CRC development is increased in UC [EL2]. The ECCO Consensus working party, through their answers to questionnaires, supported this evaluation of the data.

9.1.2. Risk factors for cancer development 

	ECCO Statement 9B
	Risk is highest in patients with extensive colitis, intermediate in patients with left-sided colitis, and not increased in proctitis [EL2].

Several independent factors affect the magnitude of the risk of malignant transformation. The duration of disease and extent of mucosal inflammation are the most prominent. There is no uniform definition of the duration of disease, although onset of symptoms has generally been used in the studies that have identified this parameter as a risk factor. In a review of 19 practice and population-based studies, Eaden confirmed that the CRC risk appears to increase 8–10 years after the onset of UC related symptoms65 and subsequently increases in later decades of the disease [EL2].

	ECCO Statement 9C
	Patients with early onset of disease (age<20 years at onset of disease) and patients with UC-associated primary sclerosing cholangitis (PSC) may have a particularly increased risk [EL2]. Persistent inflammation and family history of CRC may contribute to the risk of CRC in patients with UC [EL3]

The extent of mucosal inflammation (including backwash ileitis) has been correlated with the risk of CRC in several studies, as well as in a systematic review [EL2].66, 67, 71, 72, 73, 74, 75 Other factors have also been associated with a high CRC risk in all or part of these studies. These include young age at onset of the disease (less than 20 years of age at the time of diagnosis)65 and an association with primary sclerosing cholangitis (PSC) [EL2].76 However, there was no difference in median age at onset of colitis for those with or without CRC in the 600-patient study from St Mark's (p=0.8, Mann–Whitney) years.69 The persistence of mucosal inflammation72, 73 or a family history of CRC77 may also contribute to an increased risk, but the association has been less consistent across the studies [EL3]. In a nested case-control study from two well-defined, population-based IBD cohorts (Copenhagen County, Denmark and Olmsted County, Minnesota) 43 cases of IBD-associated CRC were matched on six criteria to 1–3 controls (n=102). Significant associations were found between PSC (OR 6.9, 95% CI 1.2–40.0), the percentage of time with clinically active disease (OR per 5% increase 1.2, 95% CI 1.0–1.4), and ≥12 months of continuous symptoms (OR 3.2 95% CI 1.2–8.6).78 The presence of pseudopolyps, which can be considered a marker of severity of inflammation, have been associated with double the risk of CRC (OR 2.5; 95% CI: 1.4–4.6),73, 79 which is a useful practice point for clinicians.

9.2.1. Screening and surveillance 

Since dysplastic changes in colonic mucosa are associated with an increased risk of CRC in UC, surveillance colonoscopy programmes have been developed with the aim of reducing morbidity and mortality due to CRC, while avoiding unnecessary prophylactic colectomy. Surveillance for CRC in patients with UC amounts to more than just performing repeated colonoscopies, but includes reviewing patient symptoms, medication use and laboratory values as well as updating personal and familial medical history. At the onset of these programmes, an initial screening colonoscopy is performed, with the goal of reassessing disease extent and confirming the absence of dysplastic lesions. Thereafter surveillance colonoscopies are regularly performed at defined intervals (below).

9.2.2. Effectiveness 

	ECCO Statement 9D
	Surveillance colonoscopy may permit earlier detection of CRC, with a corresponding improved prognosis [EL3, RG B]. Unequivocal evidence that surveillance colonoscopy prolongs survival in patients with UC is lacking [EL3, RG B]

The effectiveness of these programmes has been evaluated in some prospective studies, systematically reviewed by the Cochrane collaboration.66 An American consensus conference, held under the auspices of the Crohn's and Colitis Foundation of America, also reviewed the usefulness of screening and surveillance colonoscopies in 2005.80 The Cochrane collaboration used death related to CRC as the primary endpoint for the evaluation of surveillance programmes in UC, limiting their analysis to prospective randomised studies that included a control group. The authors were unable to demonstrate a benefit of surveillance programmes for preventing CRC-related death in UC by these strict parameters, but included only two studies in their final analysis.81, 82 An earlier meta-analysis included a third study yet to be published in full, but concluded that there was an improved 5-year survival in patients undergoing surveillance, compared to UC patients outside surveillance programmes.83 Furthermore, in the largest and most meticulous screening programme reported to date, involving 600 patients, 2627 colonoscopies, 5932 patient-years of follow-up and a caecal intubation rate of 98.7%, with no significant complications, 16/30 cancers were interval cancers.69

Unequivocal evidence for the benefit of these programmes is therefore lacking and the apparent benefit could still be linked to lead-time bias. Patients in surveillance programmes may have an earlier diagnosis of CRC even if CRC is detected independently of surveillance colonoscopy. Diagnosis of CRC in patients outside such programmes may arise from later, symptom-driven investigation [EL3]. These issues are best discussed with patients before entry into a surveillance programme.

The Consensus had divided opinions regarding the ability of surveillance colonoscopy programmes to improve survival in UC patients, in keeping with the contrasting results of the meta-analyses. Only one third of the voting experts considered that the procedure could achieve this goal, while two-thirds remained unconvinced or attributed any benefit to potential lead-time bias. Nevertheless, it should be noted that any benefit estimated in years of life saved may be much greater in UC patients than for general population screening. This is because UC-related CRC tends to occur earlier in life and with a higher frequency than in the general population. Mathematical models have evaluated that the duration of life saved per case screened ranges from 1.2 to 5 years in UC patients, compared to 1.2 to 4 months in general population screening,66, 84 depending on the parameters included in the calculation.

9.2.3. Initial screening colonoscopy 

	ECCO Statement 9E
	Screening colonoscopy should be offered 8–10 years after the onset of UC symptoms to all patients to reassess disease extent [EL5, RG D]

As duration of disease is a major risk factor for the development of CRC in UC patients, it is rational to propose a screening colonoscopy when the risk starts to increase, i.e. after 8–10 years from the onset of disease [EL2]. This initial colonoscopy also aims to reassess the extent of disease, since this parameter also impacts on the risk of CRC. Nevertheless, the appropriateness of screening colonoscopy as a way of reassessing disease extent and potential risk has not been formally established. It has been proposed in reviews and a prior consensus report,80 as well as being agreed during the present Consensus conference by the participating experts [EL5].

9.2.4. Surveillance schedules 

	ECCO Statement 9F
	In extensive colitis, surveillance should start after screening colonoscopy and be performed every other year up to year 20 of disease, then annually [EL2, RG B]. Surveillance should start 15 years after onset of disease in left-sided or distal UC. Proctitis does not require further surveillance [EL2, RG B]

The surveillance schedule is also arbitrary, but because CRC has been observed within 2 years of surveillance colonoscopy,85, 86 intervals between repeat investigations should not exceed this and should be shorter in patients with particularly high risks such as those with longstanding disease or PSC.

Furthermore, although disease extent is central to CRC risk assessment, this parameter may be difficult to define, implying that surveillance may be offered to large groups of patients. Considerable differences between extent assessed by colonoscopy and histology have been reported,87 as well as variations in extent over time.88 Neoplasia has been reported in areas of microscopic involvement without endoscopically visible inflammation. Thus, disease extent should be defined not only by the outcome of screening colonoscopy, but also by the results of previous procedures. In contrast, there is good evidence that the CRC risk is lower in patients with limited disease71, 75 as defined by colonoscopy or barium enema, so a reasonable compromise is to defer surveillance until later time points in patients with limited macroscopic disease [EL2]. This all assumes that a decision has been made with the patient that surveillance is appropriate. If the risk of CRC complicating colitis is thought to be no higher than the general population, surveillance may be considered unnecessary.

	ECCO Statement 9G
	If primary sclerosing cholangitis (PSC) is associated to UC, surveillance should be performed annually from the time of PSC diagnosis [EL3, RG B]

In other situations, such as patients with UC-associated PSC, the risk of developing a CRC is not only particularly high, but has been reported to occur early (median 2.9 years) in the course of the disease.89 These patients should enter in a more intensive surveillance programme once the diagnosis of PSC has been established.

The recommendations by ECCO (Statements 9E–9G) are contingent on a perceived increased risk of CRC in UC (Statements 9A–9C) and widespread acceptance in several European countries that screening for CRC in the general population is appropriate. If the latter applies, it is difficult to justify failure to screen a group of patients with higher risk of CRC more closely. The recommendation grades are appropriate to the strength of the evidence.

9.3.1. Number and site of biopsies 

Evidence for procedural techniques during surveillance colonoscopy is better documented than the benefit of the programme itself. At least 33 biopsies should be obtained from the various segments of the colon to achieve 90–95% sensitivity for the detection of dysplasia.90, 91, 92, 93 A reasonable approach would therefore to perform 4 random biopsies every 10 cm around the colon. Extra biopsies should be obtained from strictured or raised areas and from other abnormal areas in the colon. Full colonoscopy is necessary because about a third of UC-associated CRC develop in the proximal colon.85 This strategy is further supported by the observation that most dysplastic lesions are visible during careful colonoscopy. In a study performed on 525 patients who underwent 2204 surveillance colonoscopies, Rutter detected 110 neoplastic areas in 56 patients.94 Eighty-five (77.3%) were macroscopically visible at colonoscopy and 25 (22.7%) were macroscopically invisible. Fifty patients (89.3%) had macroscopically detectable neoplasia, while only 6 (10.7%) had macroscopically invisible lesions. The value of random biopsies, however, is limited compared to optical techniques that enhance detection of dysplastic epithelium.

9.3.2. Chromoendoscopy 

	ECCO Statement 9H
	Random biopsies (4 every 10 cm) and targeted biopsies of any visible lesion should be performed during surveillance colonoscopy [EL2b, RG B]. Methylene blue or indigo carmine chromoendoscopy is an alternative to random biopsies for appropriately trained endoscopists and is superior to random biopsies in the detection rate of neoplastic lesions [EL1b, RG B]

The yield of surveillance colonoscopy can be improved by spraying dyes (such as methylene blue or indigo carmine) that highlight subtle changes in the architecture of the colonic mucosa.95, 96, 97, 98, 99, 100, 101 All studies have confirmed an improved diagnostic yield for dysplasia detection using chromoendoscopy. With this method, random biopsies of apparently normal mucosa is of no additional value compared to targeted biopsies obtained after dye staining of the mucosa.101 Comparable diagnostic yields from chromoendoscopy have been obtained with both methylene blue and indigo carmine.97, 98 Despite these good results, a single from experienced investigators found that no dysplasia was missed even without dye spraying.94 However, trained endoscopists in chromoendoscopy may even further distinguish neoplastic from non-neoplastic changes, based on surface crypt architecture based on pit pattern recognition with a sensitivity of 93% and 97%, respectively. Colonoscopy with dye staining did not take significantly longer than conventional colonoscopy.98 This endoscopic approach may not only improve the yield of screening and surveillance colonoscopies, but also decrease the workload of pathologists because fewer biopsies are needed per procedure.

9.4.1. Risk of progression to cancer 

	ECCO Statement 9I
	A finding of dysplasia should be confirmed by an independent pathologist [EL2b, RG B]

The grade of dysplasia is important because it impacts on the sensitivity and specificity of the presence or future development of CRC. Dysplasia of any grade has been reported to have a 74% sensitivity and 74% sensitivity for CRC development, while in the same series from the Mayo Clinic, high grade dysplasia had lower sensitivity (34%) but 98% specificity for CRC detection.104 In the most recent meta-analysis, low-grade dysplasia was found to be associated with a 9-fold increased risk of developing CRC and a 12-fold risk of developing advanced neoplasia.105 Therefore, the finding of low-grade dysplasia carries a substantial risk: such a finding has important prognostic implications. For this reason, dysplasia should be confirmed by an experienced pathologist, because interobserver variation for the detection of dysplasia is high.106, 107, 108 Furthermore, individual studies that do not show an increased risk of malignant transformation in low-grade dysplasia109 need to be placed in the context of the meta-analysis.

9.4.2. Dysplasia and colectomy 

	ECCO Statement 9J
	High grade dysplasia in flat mucosa and adenocarcinoma are indications for proctocolectomy [EL2, RG B]. A patient with low-grade dysplasia in flat mucosa should be offered proctocolectomy or repeat surveillance biopsies within 3–6 months [EL2b, RG B]

Once dysplasia is found, the rationale of such a surveillance programme demands that colectomy is performed, because the risk of CRC is appreciably increased.105 If high grade dysplasia is present, the decision is easier, because the risk of concomitant CRC may be as high as 32%,106 assuming that the biopsies were indeed obtained from flat mucosa and not from an adenoma. If low-grade dysplasia is detected, the 9-fold increased risk of developing CRC reported in the most recent meta-analysis105 could reasonably be viewed as justification for colectomy as well, and this option should be discussed with the patient.110 However, because some follow-up studies of patients with \low-grade dysplasia have shown a low rate of CRC development,86, 111 it seems a reasonable compromise to continue surveillance with extensive biopsy sampling at shorter (perhaps 3–6 month) intervals in those who will adhere strictly to the surveillance program. This remains controversial in the literature and was discussed during the conference as well.66, 112

9.4.3. Dysplasia and raised lesions 

	ECCO Statement 9K
	A raised lesion with dysplasia should be completely resected. In the absence of dysplasia in the flat surrounding mucosa, meticulous endoscopic surveillance should be proposed [EL2b, RG B]. If endoscopic resection is not possible or if dysplasia is found in the surrounding flat mucosa, proctocolectomy should be recommended [EL2b, RG B]

Raised lesions on a background of UC have been traditionally referred to as “Dysplasia Associated Lesion or Mass” or DALM. Until recently this finding has been considered an absolute indication for colectomy. It is increasingly recognised, however, that some of these raised lesions may resemble sporadic adenomas and that they may be amenable to complete endoscopic resection.97, 113, 114, 115 If the polypectomy is confirmed complete by histology and if biopsies obtained from the flat mucosa immediately adjacent to the polypectomy site show no dysplasia and if, in addition, no dysplasia is found elsewhere in the colon, then colectomy may be safely deferred. Careful follow-up, preferably with surveillance colonoscopy at 3 months and then 6 monthly, is needed if this strategy is followed, because at least half of such patients in the four studies quoted developed further raised lesions. If the lesion does not resemble typical adenoma, is not respectable, or is associated with dysplasia in the adjacent mucosa, then colectomy is indicated due to the high risk of concomitant CRC.90, 113

10.2.1. Diagnostic threshold 

	ECCO Statement 10A
	Ulcerative colitis should be suspected in a child with chronic (≥4 weeks) or recurrent (≥2 episodes in 6 months) bloody diarrhoea, after exclusion of infective or other causes. This applies particularly when there is growth failure and/or pubertal delay, a family history of IBD, increased markers of inflammation, or if anaemia is present [EL2b, RG B]

In contrast to paediatric CD and its diverse symptomatology, the clinical manifestation of UC is almost uniformly bloody diarrhoea (84–94% of children), accompanied by tenesmus.132 Although an infective aetiology should be excluded, its presence does not exclude a diagnosis of UC or CD. The combination of rectal bleeding, anaemia and increased ESR identified 86% of patients with IBD prior to an endoscopic procedure.133 Other retrospective case series have confirmed the diagnostic value of increased inflammatory markers and anaemia for IBD.134, 135

A shorter interval from symptoms to diagnosis of UC probably explains why growth failure is half as common compared to CD. As with adults, the greatest risk factor for developing UC in childhood is to have a family member with ulcerative colitis (relative risk 7–17).137, 138 The risk for CD in a family member with Crohn's disease is a relative risk of 15–35. The stronger the family history, the earlier the onset of symptoms. For patients with early-onset UC (<5 years' age), 26%–44% have a family history of UC, compared to older patients or children with CD.139, 140 Genetic factors are likely to have a stronger influence in paediatric IBD, especially in early-onset acute severe UC, compared to older children or adults.141, 142 Nevertheless, most children with IBD have no family history and are considered sporadic.

10.2.2. Documentation 

	ECCO Statement 10B
	In all children with UC, the height, weight (and pre-diagnosis growth curve) and pubertal stage, should be recorded at diagnosis, and regularly during follow-up [EL3b, RG B]

Growth failure is a unique complication of paediatric IBD, caused by a combination of inadequate calorie intake, increased losses and active inflammation. Efficacy of medical treatment and concomitant mucosal remission is characterised by normal linear growth and pubertal development. In contrast, when catch-up growth does not occur after growth failure at diagnosis, or when height velocity decreases during maintenance treatment, it is highly likely that there is persistent disease activity, so therapy should be more aggressive and an adequate calorie intake ensured.136, 143, 144

10.2.3. Diagnostic procedures 

	ECCO Statement 10C
	Ileocolonoscopy and biopsies should be performed in all children or adolescents with a suspicion of inflammatory bowel disease (IBD). Upper gastrointestinal endoscopy is recommended when ileocolonoscopy does not confirm a diagnosis of ulcerative colitis [EL2a, RG B]

	ECCO Statement 10D
	In children and adolescents (up to 16–18 years of age), endoscopy should be performed by a specialist with experience in paediatric gastroenterology, preferably by a paediatric gastroenterologist [EL 5, RG D], in a setting that is suitable for diagnosing and treating children with IBD (paediatric hospital, with access to general anaesthesia)

The IBD working group of the European Society of Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN) has reached a consensus on the diagnosis of IBD in children, which has been summarised in the ‘Porto Criteria’.145 This group feels it essential to establish a diagnosis of the type of disease, as well as to determine severity, localisation, and extent of the disease, before treatment is started. Paediatric patients with UC have more extensive disease and rectal sparing in up to 30%,146 so a complete diagnostic work-up is warranted in children with bloody diarrhoea. Evidence supporting colonoscopy with ileal intubation and multiple biopsies, rather than sigmoidoscopy alone, is provided by retrospective cohort studies.146, 147, 148 In cases with extensive colitis that cannot be classified, gastroduodenoscopy may allow definitive diagnosis.149

The ECCO Consensus agrees that a paediatric gastroenterologist, rather than a specialist in adult endoscopy, should best perform endoscopy in children suspected of IBD. The most important argument is quality of care, particularly because endoscopy in children is preferably done under general anaesthesia: this is preferred for reasons of comfort and care and has been shown to be safe.150, 151, 152, 153, 154, 155 Moreover, the treatment and follow-up of children and adolescents with IBD should be in the hands of a paediatric gastroenterologist who is aware of age-related differences in disease presentation and treatment. Such specialists are experienced in handling problems such as linear growth retardation and pubertal delay.156

10.3.1. Distal colitis 

	ECCO Statement 10E
	Oral [EL2b,RGB] aminosalicylates and/or topical aminosalicylates (suppositories in proctitis, enemas in left-sided colitis) [EL5, RGD] are appropriate initial induction therapy for mild to moderate distal colitis in children or adolescents

No studies have been performed in children with distal colitis. A questionnaire sent to members of the IBD working group of ESPGHAN, however, revealed great variation of care in the treatment of distal colitis. The first choice was either oral treatment alone (mesalazine 21%, sulfasalazine 36%), or in combination with topical treatment (mesalazine 36%, corticosteroids 7%). Considering the rarity of proctitis in children, no standard treatment protocols exist.

10.3.2. Extensive colitis 

	ECCO Statement 10F
	For mild to moderate pancolitis in children, oral mesalazine/sulfasalazine is recommended as first line therapy [EL2b, RG B]. Oral steroids are appropriate if the response is insufficient [EL4, RG D]

Only one prospective study has confirmed the efficacy of oral aminosalicylates in children with active ulcerative colitis.157 In this trial, a clinical response at 8 weeks was seen in 79% of patients receiving sulfasalazine (60 mg/kg/day) and 50% of patients on olsalazine (30 mg/kg/day). Retrospective studies have also shown that oral aminosalicylates effectively induce clinical remission.158, 159, 160, 161, 162

Although sulfasalazine may cause more gastrointestinal side-effects, it is the preferred aminosalicylate treatment in young children who cannot swallow tablets, because it is available as a suspension. Alternatively, mesalazine can be given as an enteric-coated granule formulation. Based on expert opinion and extrapolation from pharmacokinetic studies,159, 161, 163 the advised dose (oral and rectal mesalazine combined) in children aged 12 years or older of mesalazine, is 50–75 mg/kg/day with a maximum of 4 g/day. For sulfasalazine it is 100 mg/kg/day with a maximum of 6 g/day.

Concerning oral corticosteroids, no studies have been performed in children with UC. Nevertheless, prospectively collected data from the US Pediatric Inflammatory Bowel Disease Collaborative Research Group Registry database provides a useful insight.164 In 97 children (age <16 yr) with newly diagnosed UC between 2002 and 2005, with a minimum of 1 year of follow-up, 79% received corticosteroids, most (62/77) within 30 days of diagnosis. For those treated within 30 days, disease activity at 3 months was inactive in 60%, mild in 27%, and moderate or severe in 11%. At 12 months, 31 of 62 (50%) of the early corticosteroid-treated patients were considered corticosteroid responsive and 28 (45%) were corticosteroid-dependent. A total of 4 patients receiving corticosteroids required colectomy in the first year. Immunomodulators were used in 61% of all corticosteroid-treated patients. This is similar to adults: early response is excellent, but dependence is common, even with immunomodulators. Evaluation among the IBD working group of ESPGHAN demonstrated that 46% favoured addition of corticosteroids if response to aminosalicylates was found to be insufficient. Oral prednisolone is given as a once daily dose of 1–2 mg/kg/day, with a maximum dose of 40 mg, for 2–4 weeks (until clinical remission), then tapered to zero in 6–8 weeks. Although not supported by clinical evidence from randomised clinical trials, calcium and vitamin D are usually supplemented during a course of steroid treatment.

10.3.3. Severe colitis 

	ECCO Statement 10G
	For severe pancolitis in children, corticosteroids are first line therapy [EL4, RG D]. If the response is insufficient, intravenous ciclosporin [EL4, RG C] or infliximab [EL4, RG C], or colectomy are appropriate options

Although no randomised clinical studies have been performed in children with acute severe UC, all respondents to the ESPGHAN questionnaire agreed that corticosteroids are the first line therapy in severe pancolitis. In a meta-regression of response to steroids in 32 studies involving 1991 patients (1974–2006), only 43 children were included.165 To evaluate the outcome in children, a retrospective study of 74 admissions in 63 children (57% males, age at diagnosis 10.9±4 yr, 79% extensive colitis) treated at Toronto SickKids Hospital 1995–99 was performed.166 41% failed intravenous steroids by discharge and 23 (37%) came to colectomy on that admission. By one year, 54% and by 5 yr 59% had come to colectomy. There was no clear consensus from ESPGHAN as to whether corticosteroids should be given as the only treatment (25% of respondents), or in combination with oral mesalazine (25%), or intravenously with adjunctive parenteral nutrition (50%). Given the similarity in the response of children to steroids compared to adults (Section 5.2.4, preceding paper same issue), it seems unlikely that mesalazine is necessary. Although nutritional support is particularly appropriate in children, TPN in adults has not been shown to offer any advantage when managing acute severe colitis (Section 5.2.4, preceding paper same issue).

When 3–5 days of intravenous corticosteroids are ineffective, rescue therapy with ciclosporin, tacrolimus, or infliximab are the only two options to avoid or postpone colectomy. An objective assessment of the response to steroids facilitates management as it does in adults (see Section 5.2.5, preceding paper same issue). A paediatric index of severity as been developed166 and when calculated on day 3, strongly predicts failure of intravenous steroids.167 As with adults, stool frequency (p=0.001) and CRP (p=0.045) on day 3 (but not day 1) predict failure, along with temperature (p=0.001). Case studies with intravenous ciclosporin in children with severe, steroid-refractory colitis who are candidates for surgery, have shown remission of the disease in up to 80% of cases.168, 169, 170, 171, 172 In many children, however, tapering of oral ciclosporin resulted in a relapse and was followed by colectomy within a year of cessation of treatment. In occasional patients, short term ciclosporin treatment can effectively induce remission while waiting for azathioprine maintenance treatment to take effect.170

Infliximab has not been studied prospectively, but small retrospective studies in new-onset steroid-refractory patients show complete remission in 75–88% of patients.173, 174, 175 With the small numbers of patients studied and limited follow-up, it is currently unknown whether infliximab therapy is effective in avoiding colectomy, or whether it simply defers it. The Consensus view is that rescue therapy with either ciclosporin, tacrolimus, or infliximab should only be initiated in a specialist centre where a paediatric and/or colorectal surgeon are available and involved in the treatment of these severely sick children.

10.4.1. Aminosalicylates 

The efficacy of mesalazine or sulfasalazine maintenance treatment has not been studied in children with UC. From the IBD working group of ESPGHAN questionnaire respondents, 57% advised continuing the same mesalazine dose as used for induction, while 43% advised a lower dose. The Consensus view is based on results from adult studies that indicate that high dose 5-ASA is effective maintenance treatment. Long-term corticosteroids are absolutely contraindicated, because they do not maintain remission and have a negative effect on linear growth and bone mineralisation. Ciclosporin maintenance treatment is ineffective and inappropriate, because serious, sometimes irreversible, side-effects may occur.

10.4.2. Thiopurines 

Retrospective cohort studies have demonstrated that maintenance with azathioprine/mercaptopurine is effective, while achieving a steroid-sparing effect.176, 177, 178, 179 This steroid-sparing effect is more evident when azathioprine treatment is started early in the course of disease, within 2 years after diagnosis.179 The advised dose for azathioprine in children is 2–3 mg/kg/day and that for mercaptopurine is 1–1.5 mg/kg/day.

10.5.1. Indications 

In acute severe colitis, the decision to perform colectomy should be evaluated on a day-to-day basis by both the medical and surgical team. If the disease is not responding to 7–10 days of either calcineurin inhibitors (ciclosporin, tacrolimus) or infliximab, colectomy is indicated.

Colectomy is also indicated for persistently active disease, when corticosteroid dependency exists despite concomitant therapy with azathioprine/mercaptopurine, or when immunosuppressive treatment has side-effects. Growth failure despite apparently adequate maintenance therapy is also an indication for colectomy, even when clinical symptoms appear mild.180, 181, 182, 183, 184 Preliminary (and anecdotal) experience with infliximab in children suggests that it may be more effective in acutely ill patients, compared to patients with chronic refractory disease.173, 175, 185 It rarely achieves steroid-free remission. Therefore, infliximab cannot be recommended for chronic steroid-dependent disease in children.

10.5.2. Procedures 

	ECCO Statement 10J
	Colectomy should be performed by an experienced paediatric surgeon, ideally with the assistance of a colorectal surgeon with paediatric experience; ileo-pouch-anal anastomosis (IPAA) should only be performed in a highly specialised centre [EL 4, RG C]

Depending on the local circumstances, a child needing colectomy should be referred for expert care at a specialist centre. Case series of IPAA in children show good results in terms of quality of life, continence and incidence of pouchitis.181, 182, 186, 187, 188, 189 However, in very young children (<10 years), pouchitis is reported in 75% of the patients.190 Because IPAA decreases female fecundity,191, 192 colectomy with ileorectal anastomosis until later IPAA may be a better option in girls.193 Expert advice should be sought.

12.2.1. Aetiology 

	ECCO Statement 12A
	A speculated association between psychological factors and the aetiology of ulcerative colitis cannot be proven. There is, however, an influence of psychological distress and mood disorders on the course of the disease [EL1b, RG B]

Studies about the influence of psychological factors on the development of ulcerative colitis are very limited. There are a few studies with hypothetical interpretations about the influence of psychosocial factors on the aetiology of the disease.209, 210, 211, 212 Many studies on psychosocial factors relate to inflammatory bowel disease (IBD) rather than ulcerative colitis or Crohn's disease in particular.

12.2.2. Pattern of disease 

	ECCO Statement 12B
	There is evidence of an interaction between psychological factors and IBD activity. Depression and perceived chronic distress represent risk factors for relapse of the disease. It remains controversial whether acute life events trigger relapses [EL 1b, RG B]. Most patients consider stress to have an influence on their illness [EL 4, RG C]

Psychological factors are considered to have an influence on the course of the disease, which is consistent with evidence in the recent literature about the influence of subjective perceived psychological distress on disease activity of ulcerative colitis.213, 214, 215, 216, 217 Studies about the influence of major life events on the biological disease activity have yielded contradictory results.218, 219, 220 Patients themselves and the majority of European experts at the Consensus conference consider psychosocial distress as influencing the risk of relapse.221, 222 One study shows a heightened response to stressors and the greater epithelial damage in IBD patients, which suggests that stress-induced activation of the brain-gut axis and of mucosal mast cells may be important in the initiation and/or flare up of IBD.223

12.4.1. Communication with patients 

	ECCO Statement 12E
	The psychosocial consequences and health-related quality of life of patients should be taken into account in clinical practice at regular visits. Individual information and explanation about the disease should be provided through a personal interview. The course of the disease can be improved by combining self-management and patient-centred consultations [EL 1b and 3b, RG B]

Health perceptions impact on the experience of the illness.226 Increasing physician awareness of the fact that psychologically distressed patients have difficulty in processing clinically relevant information 246 may lead to improved doctor–patient communication.247 It is important that patients are informed about their condition through an individual interview, in conjunction with emotional support. This is because a lower level of information is associated with greater concern,248 despite the impression of some doctors that more information increases the level of anxiety. Psychosocial factors are strongly correlated with health care utilization.249 Self-management guidebooks together with patient-centred consultations improve patients' disease control.250, 251 It should however be recognised that educational booklets on their own do not seem to be helpful and may even worsen the health-related quality of life of patients attending tertiary centres.252 In addition, patient education programmes seem to have very limited or even no influence on the course of their illness, their health-related quality of life, or their psychological affect.253, 254, 255 Almost all experts at the Consensus (91%) are convinced that a good doctor–patient relationship is helpful psychologically and take psychosocial factors into account for both diagnosis and therapy. Most experts at tertiary centres have the opportunity for integrated somatic and psychological care of patients. However, patients describe deficiencies in the care of family members, insufficient information and inadequate access to healthcare resources.256

12.4.2. Psychological support 

	ECCO Statement 12F
	Physicians should assess the patient's psychosocial status, quality of life and demand for additional psychological care and recommend psychotherapy if indicated. Integrated psychosomatic and gastroenterology care should be available [EL 2b, RG B]. Patients should be informed of the existence of patient associations [EL 5, RG D]

For assessment of quality of life, two validated IBD-specific questionnaires have been shown to be sensitive, reproducible and responsive for use in clinical trials: the Inflammatory Bowel Disease Questionnaire (IBDQ)257 and the Rating Form of Inflammatory Bowel Disease Patient Concerns (RFIPC).258 Detection and treatment of psychological distress has the potential to improve health-related quality of life.259 The presence of psychological disorders contributes to poor quality of life and the number of doctor visits, regardless of the severity of the condition.249 This is the common experience of doctors caring for patients with IBD, even if the potential or need to treat this aspect of the illness is perceived.

To assess the demand for psychological care in chronic diseases, a validated questionnaire is available, developed and based on inflammatory bowel disease.260 Most experts (80%) feel themselves able to recommend psychotherapy during a discussion with patients. There is no study on their competence at doing this, but this is consistent with the experience of participants in the European Consensus on the management of Crohn's disease,206 the German Consensus on Crohn's disease,56 and that on ulcerative colitis.261, 262 Since strategies aimed at improving social support can have a favourable impact on psychological distress,263 training of gastroenterologists to integrate psychosocial factors in clinical practice should be taken into consideration.

12.4.3. Therapeutic intervention 

	ECCO Statement 12G
	Psychotherapeutic interventions are indicated for psychological disorders associated with ulcerative colitis [EL 1b, RG B]

Psychotherapy and relaxation methods have a positive influence on IBD, mainly affecting the psychological dimensions of the illness such as psychological well-being, coping strategies and psychological distress,264, 265, 266, 267, 268 as well as perception of pain.269 This underpins the recommendation (Statement 12G). The diagnosis of ulcerative colitis is insufficient alone to recommend psychotherapy, since studies of psychotherapy on patients without psychological disturbance show little or no benefit.270 One study that combined patients with Crohn's disease and ulcerative colitis reported an influence of psychotherapy on disease activity, but there was inhomogeneity in randomisation of the treatment and control groups, so the results are not included in the evidence-based recommendation.

12.4.4. Therapeutic choice 

	ECCO Statement 12G
	The choice of psychotherapeutic method depends on the psychological disturbance and should best be made by specialists (Psychotherapist, Specialist for Psychosomatic Medicine, Psychiatrist). Psychopharmaceuticals should be prescribed for defined indications [EL 5, RG D]

There is no evidence that one psychotherapeutic method should be preferred over another. Relaxation exercises are useful, since they are easy to learn and perform. Expert opinion believes that there is an advantage if the psychotherapist has experience in the treatment of patients with chronic inflammatory bowel diseases and works closely with the patient's gastroenterologist. There are no specific studies on the use of individual psychopharmaceuticals in ulcerative colitis.271 In spite of this, almost all experts believe that there are clinical situations in which antidepressants should be recommended for treatment of psychological distress associated with ulcerative colitis.

13.2.1. Pauciarticular peripheral arthropathy 

Type I arthropathy277 predominantly affects weight-bearing joints, including the ankles, knees, hips, wrists, elbows and shoulders. Pauciarticular means that fewer than five joints are affected. The arthritis is usually acute, self-limiting, resolves within weeks as disease activity decreases, and leaves no permanent joint damage. Clinical examination reveals painful, tender, swollen joints. Aspiration is unnecessary unless an alternative diagnosis is suspected. The differential diagnosis includes osteoarthritis, septic arthritis, pyrophosphate arthropathy, coincidental rheumatoid arthritis, or occasionally, gout. If just one hip joint is affected then steroid-induced osteonecrosis should be considered.279

13.2.2. Polyarticular peripheral arthropathy 

Type II arthritis predominantly affects the small joints of both hands as a symmetrical arthropathy. Pain is commonly disproportionate to the signs of arthritis. It usually persists for months or years and follows a course independent of IBD activity. It may persist after colectomy or start after ileo-pouch-anal anastomosis. The differential diagnosis includes osteoarthritis, but also includes treatment side-effects such as steroid-induced pseudorheumatism (which is common after withdrawal of long-term steroids) and mesalazine- or azathioprine-induced arthropathy.278

13.2.3. Axial arthropathy 

Asymptomatic sacroiliitis is common, with up to 50% of colitis patients having abnormal radiography.279 Symptomatic sacroiliitis is characterised by pain in the buttocks after rest, which then improves with movement. The clinical sign of discomfort in the sacroiliac joints during bilateral pressure on the pelvic brim is indicative. The principal symptom of ankylosing spondylitis is persistent low back pain, usually beginning before the age of 30. Clinical examination reveals loss of the lumbar lordosis and limited spinal flexion. Conventional radiographs of the back are usually normal in the early stages of disease. Spinal CT scans and radionuclide bone scans are more sensitive than plain radiographs, but the gold standard is now magnetic resonance imaging (MRI).280, 281 There is however an impression that minor abnormalities of little or no clinical consequence may be seen on MRI; this remains to be determined by longer-term follow-up. In advanced cases there may be squaring of the vertebral bodies, marginal syndesmophytes and bony proliferation, with ankylosis producing the classical “bamboo spine”. HLA B-27 associations is found in a majority (up to 75%) of patients with axial arthritis, but is less common than in patients with ankylosing spondylitis not associated with IBD. It is unrelated to sacroiliitis and HLA typing has no role in the management of individual patient.282, 283

13.2.4. Therapy of arthropathies 

	ECCO Statement 13B
	Treatment of arthritis and arthropathy associated with UC is based almost entirely on extrapolation from that for other forms of arthritis. There is some support for use of sulfasalazine, simple analgesics, non-steroidal anti-inflammatory agents, local steroid injections, and physiotherapy [EL4, RG D]. In Type I peripheral arthritis the emphasis should be on that of the underlying colitis [EL2c, RG C]. In axial arthritis the arguments in favour of intensive physiotherapy [EL2a, RG B], sulfasalazine [EL2a, RG C], methotrexate [EL3b, RG C], and infliximab [EL2a, RG C], are somewhat stronger

Treatment of arthritis and arthropathy associated with IBD is largely empirical. This includes the use of simple analgesics, sulfasalazine, local steroid injections and physiotherapy, but whether or not to use non-steroidal anti-inflammatory agents is a continuing dilemma, even though short term use appears not to exacerbate colitis.284

For Type I peripheral arthritis the emphasis should be on the treatment of active disease, including steroids, immunomodulation, and anti-TNF therapy as appropriate. Resolution of the arthropathy can be expected. The joint-specific drug of first choice for all forms of IBD-related arthritis appears to be sulfasalazine, but convincing evidence to support this is lacking. Nevertheless, it was favoured by the greatest minority of panel members (41%). Symptomatic relief may be obtained from simple analgesics, rest and physiotherapy.279, 284, 285, 286 Although there is concern that non-steroidal anti-inflammatory agents (conventional and COX II inhibitors) may aggravate the underlying colitis,287, 288 they have been used by many gastroenterologists to good effect with limited risk of exacerbating colitis. A previous history of flare related to NSAID intake seems to be the best indicator of individual risk. A randomised study of the safety of celecoxib in colitis283 indicated that short-term use (<2 weeks) did not exacerbate colitis. Local steroid injection into the worst-affected joints often provides rapid, but temporary relief. Type II arthritis generally resolves with effective treatment of the colitis.289

Treatment of axial arthritis should include intensive physiotherapy, together with disease modifying drugs such as sulfasalazine, and methotrexate.279, 285, 289 The safety and efficacy of infliximab in ankylosing spondylitis is established, but is best reserved for intractable or severely debilitating symptoms.290, 291 This is because of the 15% prevalence of immunogenicity and the small, but definable risk of notable adverse events such as sepsis, tuberculosis, or demyelination.

13.3.1. Erythema nodosum 

Erythema nodosum is usually readily recognised. It is characterised by raised, tender, red or violet subcutaneous nodules of 1 to 5 cm in diameter. It commonly affects the extensor surfaces of the extremities, particularly the anterior tibial area, and usually occurs at times of activity of the colitis. A firm clinical diagnosis can normally be made, and biopsy is not normally appropriate. If performed, the histology reveals a non-specific focal panniculitis.292, 293

Because erythema nodosum is closely related to disease activity, despite a genetic link,294 treatment is based on that of the underlying colitis. Systemic steroids are usually required (76% Consensus view). In resistant cases or when there are frequent relapses, immunomodulation with azathioprine and/or infliximab may be added, but it is exceptional to need such measures just because of erythema nodosum. Oral potassium iodide has been used successfully in refractory cases.295

13.3.2. Pyoderma gangrenosum (PG) 

Lesions are often preceded by trauma at the site (which may have been many years earlier) through a phenomenon known as pathergy. PG can occur anywhere on the body, including the genitalia, but the commonest sites are the shins and adjacent to stomas. Initially they take the form of single or multiple erythematous papules or pustules, but subsequent necrosis of the dermis leads to the development of deep excavating ulcerations that contain purulent material that is sterile on culture unless secondary infection has occurred.

Treatment of pyoderma gangrenosum has relied on topical and systemic steroids. Steroids were considered the most effective treatment for pyoderma gangrenosum (54% Consensus view), with intravenous ciclosporin or tacrolimus reserved for refractory cases.296, 297, 298 There are, however, no reliable trials to support the use of high dose steroids or calcineurin inhibitors and these drugs have appreciable potential side-effects. Infliximab has changed the management of PG. In the first controlled trial in pyoderma (which also included patients without IBD) infliximab 5 mg/kg or placebo was given at week 0.299 At week 2 (the primary end point), significantly more patients in the infliximab group had improved (46% (6/13)) compared with the placebo group (6% (1/17), p=0.025). Overall, 29 patients received infliximab with 69% (20/29) demonstrating a beneficial clinical response. Remission at week 6 was 21% (6/29). There was no response in 31% (9/29) of patients. Infliximab is still reserved for more troublesome cases, but is highly effective.

13.3.3. Sweet's syndrome 

Sweet's syndrome is characterised by tender, red inflammatory nodules or papules, usually affecting the upper limbs, face or neck. It has only been recognised as an extraintestinal manifestation of IBD relatively recently.300, 301 It is part of the group of acute neutrophilic dermatoses that includes pyoderma gangrenosum, but can be distinguished by its appearance, distribution and histological features. There is a strong predilection for women (87%), patients with colonic disease (100%) and those with other extraintestinal features (77%). The rash is associated with active disease in 67–80%, but may precede the onset of intestinal symptoms in 21% and has been reported 3 months after proctocolectomy for ulcerative colitis.

13.4.1. Episcleritis 

Episcleritis may be painless, presenting simply with hyperaemic sclera and conjunctiva, but itching and burning sensations may also occur.303 Diagnosis of simple episcleritis depends on the exclusion of the more sinister features of uveitis. When this is not possible referral to an ophthalmologist for an expert opinion and slit-lamp examination is essential. Episcleritis usually does not require specific treatment other than for underlying disease activity. It will respond to topical steroids if symptoms are troublesome — but take care that infection (including herpetic), ulceration, and uveitis are not overlooked.

13.4.2. Uveitis 

Uveitis is less common, but has potentially severe consequences. When related to ulcerative colitis it is frequently bilateral, insidious in onset and long-lasting.304 Patients complain of eye pain, blurred vision, photophobia and headaches. Potential progression to loss of vision should prompt urgent referral to an ophthalmologist. Slit-lamp examination will confirm the diagnosis and differentiates between anterior and posterior uveitis. Uveitis is treated with steroids, and it may be necessary to use both topical and systemic routes. Infliximab is rapidly effective,304 but treatment should be guided by specialists.

13.4.3. Cataracts 

Chronic corticosteroid use for treatment of UC is associated with numerous complications and may result in posterior subcapsular cataracts develop in a significant proportion (25%) of patients receiving 15 mg or more of prednisone for 1 year.305 Although steroids do not prevent relapse and have no place in the long-term management of UC, any patient on long-term steroids should undergo routine (probably annual) slit lamp examination.

13.5.1. Primary sclerosing cholangitis 

For patients with cholestasis the probability of PSC is appreciably increased if the ultrasound scan is normal, if drug side-effects are thought unlikely, and if serological tests for primary liver disease are negative. Magnetic resonance cholangiography (MRC) is now established as the first-line diagnostic test for PSC.307 The characteristic pattern shows irregular bile ducts, with zones of both narrowing and dilatation. If MRC is normal and PSC still suspected (such as otherwise unexplained cholestasis in a patient with IBD), then it is safer and probably more effective to do a liver biopsy rather than diagnostic ERCP, since this will detect predominant small duct disease. PSC substantially increases the risk of both cholangiocarcinoma and colorectal carcinoma (9.1.2, 9.5).

PSC appears to respond to ursodeoxycholic acid (ursodiol), which improves abnormal liver function tests308 and may, at 20 mg/kg, possibly improve prognosis. It is possible that it also reduces the risk of colonic cancer in these patients118 (Section 9.5). Tacrolimus has yielded a rapid decrease in liver enzymes, but no histological improvement.309 ERCP may be used to treat dominant strictures by dilatation and/or stenting. Advanced liver disease may necessitate transplantation, but recurrence of PSC in the transplanted liver occurs in approximately 20% of patients.308, 310 Specialist advice is appropriate when managing a patient with PSC and IBD. Because of the higher risk of colorectal cancer, it is generally considered appropriate to perform annual screening colonoscopy from the time of diagnosis.

13.6.1. Diagnosis and fracture risk 

Osteoporosis and osteopenia are common in patients with IBD (20–50%), but the actual number of fractures in IBD is only slightly increased to the general population.311, 312 In a study using the general practice research database, the relative risk of hip fracture was 1.62 (95% CI 1.24–2.11) for all IBD, 1.49 (1.04–2.15) for ulcerative colitis and 2.08 (1.36–3.18) for Crohn's disease.312 Contributing factors include age, steroid treatment, smoking, low physical activity (including that from hospitalisation), inflammatory cytokines, and probably also resection with pouch formation.

Diagnosis is conventionally based on bone densitometry (DEXA scanning), osteoporosis being defined as a T score of less than −2.5. Ultrasound has been suggested as method of screening, but is not yet reliable. The presence of osteoporosis increases the risk of fracture of long bones and of the spine, although probably a great deal less in young patients than was once thought. It is conventional to use a radiological diagnosis of osteoporosis as an indication for specific therapy.

Osteopenia (T score less than −1.0) is thought by some to be an important risk factor for fracture in its own right, but this is increasingly questioned.313 It is, however, probable that it is a marker of increased risk of later osteoporosis even if the risk is not absolute. Therapeutic intervention is probably not justified on present knowledge, but continued surveillance for bone loss is appropriate. It is important to put risks into perspective when discussing with patients.

The risks of osteoporosis (and the potential risks from osteopenia) should be explained. The recommended dietary calcium intake should be 1000–1500 mg/day, which often means supplementation even in patients not taking corticosteroids. It should be noted that recommendations for the treatment of osteoporosis apply only to adults over the age of 25 years, and that evidence for treating oestopenia is circumstantial. The diagnosis of osteoporosis in children and long-term consequences of treatment with bisphosphonates are unknown.

13.6.2. Management 

	ECCO Statement 13G
	Weight-bearing exercise [EL2b, RG B], stopping smoking [EL3b, RG C], avoiding alcohol excess [EL4, RG D], and maintaining adequate dietary calcium (>1 g/day) [EL2b, RG B] are beneficial. In post-menopausal women with osteoporosis, regular use of bisphosphonates, calcitonin and its derivatives, and raloxifene reduce or prevent further bone loss [EL2b, RG C]. Data in males with osteoporosis are less secure but bisphosphonates are probably of value [EL3b, RG C]. Newer data also support the use of strontium salts [EL2a, RG B]

The risks of osteoporosis (and the potential risks from osteopenia) should be explained. The treatment of osteoporosis is based on studies that are not specific to IBD.314 Weight-bearing, isometric exercise, stopping smoking, avoiding alcohol excess and maintaining adequate dietary calcium (>1 g/day) are beneficial, but such advice is often overlooked. Hormone replacement treatment is no longer generally advised in post-menopausal women with osteoporosis, because studies have demonstrated a slightly increased risk of breast cancer and of cardiovascular events.315 Regular use of bisphosphonates, calcitonin and its derivatives, and raloxifene (a selective oestrogen receptor modulator) may reduce or prevent further bone loss. Data in males with osteoporosis are few, but bisphosphonates are probably of value and an important practice point is that testosterone should be measured. Those with low testosterone may benefit from supplementation. Routine administration of vitamin D is not warranted. Patients on corticosteroids for short periods do not merit routine use of bone protection with bisphosphonates, assuming a normal calcium intake and all other risk factors being equal.315