Defining and assessing acute severe colitis

The improved outcome for patients with acute severe ulcerative colitis, a potentially lethal condition in the era before the use of corticosteroids, is arguably one of the major achievements in the care of patients with inflammatory bowel disease (IBD). Still, severe colitis not responding to steroids poses an important challenge to the surgeon and physician alike. Hospitalisation should be considered in all patients who have six or more stools with visible blood per day, associated with fever, dehydration, tachycardia, malaise and/or increased C-reactive protein (CRP).1–3 Recent retrospective data indicate that the presence of several additional criteria in addition to the ≥6 bloody stools a day, increased the likelihood for colectomy.4 Upon admission infectious causes of colitis should be excluded. Clostridium difficile infection complicating ulcerative colitis is increasingly recognised as a cause of severe disease and it infers an additional risk of colectomy and mortality.3 Also, other enteric pathogens should be excluded. Un-prepped sigmoidoscopy with minimal inflation performed by an experienced endoscopist provides additional information on disease severity and provides mucosal biopsies. Biopsies are particularly useful to exclude cytomegalovirus (CMV) colitis in patients with prolonged exposure to steroids and/or azathioprine, even if the precise role of active CMV replication in acute severe colitis is still debated. The severity of the endoscopic lesions may predict clinical outcomes as discussed below. Response to intravenous (IV) corticosteroids should be assessed early, since in patients not responding after 3 days the immediate risk of colectomy rises to 85%.5 Hence, a reassessment by the medical and surgical team is vital and any of the following options should be discussed with the patient: intravenous cyclosporine (2 mg/kg for 7 days, serum level controlled), infliximab (5 mg/kg IV) or total colectomy.6–10 The choice between these options is a medical–surgical decision mostly based on clinical signs, radiological findings and blood analysis (CRP, serum albumin). Complications such as toxic megacolon or uncontrolled bleeding necessitate emergency colectomy. Megacolon, as defined by the dilation of the (right) colon, is not an absolute indication for surgery, and can be managed by treating the underlying colitis. However, the combination of this complication with clinical deterioration warrants surgical intervention.

Effectiveness of medical rescue therapy

In two controlled trials intravenous cyclosporine, a fungal calcineurin inhibitor, has been shown to be an effective rescue therapy for severe attacks of ulcerative colitis (table 1).7 8 In the trial by Lichtiger et al, nine out of 11 patients treated with 4 mg/kg cyclosporine IV avoided colectomy versus none of the nine placebo-treated patients.7 Data from a single centre controlled trial in 73 patients indicate that 2 mg/kg/day IV cyclosporine is as effective for severe attacks of ulcerative colitis, although not all of the patients in the trial were failing intravenous corticosteroids.9 Cyclosporine is also effective in patients with severe colitis before failure of IV corticosteroids.8 When results from controlled and non-controlled trials are pooled 76% to 85% of patients will respond to IV cyclosporine and avoid colectomy in the short term.7–9 11 12 The proposed doses of cyclosporine only apply to the first 48 h and dose adjustments should be made based on the cyclosporine blood levels. Cyclosporine carries an immediate risk of seizures and, although exceedingly rare, of anaphylactic reactions in patients with an allergy to the castor oil additive in the IV preparations.13 The response to cyclosporine is rapid (median of 4 days8), which allows colectomy to be carried out in non-responders within 1 week. Tacrolimus and oral cyclosporine can also be considered to treat severe attacks of ulcerative colitis but only retrospective uncontrolled data are available.14–16

The efficacy of infliximab as rescue therapy for severe to moderately severe ulcerative colitis not responding to therapy with intravenous steroids has been demonstrated in a single placebo controlled trial. Significantly more patients treated with placebo (14/21) required colectomy by 3 months as compared to those treated with a single dose of infliximab 5 mg/kg IV (7/24).17 Infliximab has a long half-life and dose monitoring is not required. Anaphylactic infusion reactions to infliximab can occur but are exceedingly rare at the first dose.

Following initial response to IV cyclosporine about 50% of patients avoid colectomy at 3 years.11 12 18 Higher colectomy rates have been reported with follow-up extending to 7 years.12 The patient population already failing adequate courses of azathioprine or 6-MP is most prone to colectomy following initial response to cyclosporine.11 12 Infliximab has been introduced more recently in the management of severe UC and colectomy data from long-term cohorts are not available. Data from the Scandinavian controlled trial, however, indicate that even after a single dose infliximab protects against colectomy at least 2 years after the initial event.19 Even if medical rescue therapy shifts the surgical procedure from an emergency to an elective colectomy months or years down the road in a substantial proportion of patients, this should not be considered an argument against an attempt for medical therapy since urgent rescue colectomy is associated with increased morbidity and mortality.20

Choosing between cyclosporine or infliximab

At present no comparative trial data are available to guide the choice between infliximab and cyclosporine in patients with acute severe colitis. Predictors of early colectomy in patients treated with cyclosporine have been retrospectively assessed in a cohort treated in Paris, France.21 A combination of clinical (mild fever, tachycardia), biological (CRP >45 mg/l), and endoscopic criteria allowed the classification of the patients into two different groups (80% vs 20% colectomy at 6 months). In one controlled trial only active smoking was inversely associated with the need for colectomy (OR 0.06, 95% CI 0.008 to 0.407).9 In the long term, patients already exposed to azathioprine prior to their acute severe colitis episode are at increased risk of colectomy12 and therefore cyclosporine is particularly used as a bridging strategy to the effect of azathioprine.

Predictors of colectomy with infliximab can be inferred from several sources although not always controlled nor all in acute severe colitis. In the Järnerot placebo controlled trial, infliximab did not significantly reduce the colectomy rate in patients included based on a high fulminant colitis index at the day of hospital admission (69% placebo, 47% infliximab, p=0.3).17 In the Acute Colitis Trials 1 and 2, colectomy rates long term were reduced in patients treated with 5 and 10 mg/kg infliximab but the difference with placebo reached significance only in the 10 mg group. An increased baseline CRP (>20 mg/l), concomitant steroid use, duration of ulcerative colitis of less than 3 years and baseline Mayo score of at least 10 points predicted colectomy.22 The retrospective cohort studies reported by Ferrante et al and Seow et al confirmed increased CRP (>5 mg/l) as a predictor of colectomy, although most of the patients were not treated initially for acute severe colitis.23 24 Interestingly, in the recently reported cohort from Mount Sinai, Toronto, the presence of antibodies to infliximab and undetectable drug levels early after the first infusions of infliximab (IFX) were associated with a higher colectomy rate. Also patients with a baseline Mayo score >10 (severe colitis) had an increased risk of colectomy (OR 3.42, 95% CI 1.56 to 7.51).24 Although mostly uncontrolled and often retrospective the cumulative data available in the literature suggest that infliximab may be preferred for patients with less severe colitis and in those already exposed to purine analogues, whereas cyclosporine would be selected in patients with acute severe colitis naïve to purine analogues. However, the data from ongoing controlled comparative trials should be awaited and the data from St Antoine hospital in Paris also suggest that cyclosporine-treated patients with more severe endoscopic lesions and a higher inflammatory burden are also at increased of colectomy.21

Both cyclosporine and infliximab are immunosuppressive agents associated with an increased risk of serious infections including lethal opportunistic infections.24–26 Inevitably, previous steroid exposure adds to the burden of toxicity.27 The concomitant use of steroids is a proven risk factor for postoperative complications in patients who eventually come to colectomy.28 Cyclosporine, as such, is not associated with an increased risk of perioperative complications.29 Initial reports on an increased risk of postoperative septic complications associated with the use of infliximab in patients with active colitis30 31 have not been confirmed in other cohorts31 32 and also not by the data from the Scandinavian controlled trial in severe colitis.18 Therefore, the impact of infliximab use on complications after colectomy is still debated. In the cohort reported by Ferrante et al of outcomes after colectomy in patients with infliximab therapy a surgical procedure without temporary diverting ileostomy was independently associated with a higher risk of complications (OR 6.45, 95% CI 2.12 to 19.64, p=0.001).33

Successive use of both medical rescue therapies

Even if timely colectomy should always be considered in patients with severe colitis not responding to medical rescue, consecutive use of cyclosporine and infliximab has been reported. Results from a cohort of patients treated at the Mount Sinai Hospital in New York suggested that patients receiving infliximab followed by cyclosporine or vice versa have a substantial risk of serious adverse events including mortality,34 but the data from collaborative groups in France35 and Spain36 suggest that successive use of both agents is a therapeutic option. Of the 86 patients in the French cohort 65 used cyclosporine first and 62% of all patients who received both agents consecutively, avoided colectomy short term although remission rates were below 30% and 3 year colectomy rates 63%. Also, in this study serious infections were observed in 14% of patients and one patient died. In the smaller Spanish cohort (n=16) of patients treated with infliximab after failure of cyclosporine reported by Manosa et al the short term colectomy rate was 30% with a low incidence of side effects.36 Clinicians should carefully consider if the number of colectomies avoided by the consecutive use of cyclosporine and infliximab outweighs the cumulative immunosuppressive risk. Even if the serum half life of cyclosporine is less than 24 h, the impact on the risk of opportunistic infections may last much longer.