Review Article

The Practical Management of Acute Severe Ulcerative Colitis

D. Seah; P. De Cruz


Aliment Pharmacol Ther. 2016;43(4):482-513. 

In This Article

Predictors of Response to Salvage Therapy

Despite significant advances in medical therapy for ASUC, surgery remains a safe and viable option when medical alternatives have been exhausted. The decision to forego or abandon salvage therapy should be made as soon as an inadequate response to treatment has been identified, as delaying surgery can often worsen outcomes.[92–94] Timely cessation of medical therapy relies on recognition of patients likely to require colectomy and close monitoring for markers of treatment failure. Application of the Ho index for use in predicting response to ciclosporin has been proposed; however, novel prognostic tools have yet to be devised.[95] Currently, there is no established predictive index for assessment of response to infliximab. A number of methods to predict nonresponse to salvage therapy have been proposed and are explored below and outlined in Table 6.


Multivariate analysis of the CySIF trial identified age greater than 40 years as an independent predictor of treatment failure following salvage therapy (hazard ratio: 2.7; 95% CI: 1.2–6.1; P = 0.018).[10]

Previous Thiopurine Experience

Ongoing active disease despite treatment with an immunomodulator is predictive of nonresponse to salvage therapy with ciclosporin. In a retrospective single-centre study, nonresponsiveness to thiopurines was found to be predictive for nonresponsiveness to ciclosporin and colectomy.[96] A total of 19 patients with ASUC underwent colectomy following ciclosporin salvage therapy during the study, with 68% (13/19) having previously failed to respond to at least 3 months of azathioprine or mercaptopurine; in contrast, a significantly lower colectomy rate was reported among patients who were thiopurine-naïve – 31% (6/19) (P = 0.006). Prospective evaluation is warranted.

Clinical Parameters

Several clinical parameters have been shown to be predictive for salvage therapy failure. A retrospective review of 135 ASUC patients reported clinical outcomes during hospital admission, following treatment with intravenous ciclosporin.[97] Sixty-seven per cent (90/135) of patients responded to ciclosporin; however, 33% (45/135) underwent colectomy. Clinical factors that were predictive of nonresponse to ciclosporin included body temperature >37.5OC (hazard ratio: 1.94; 95% CI: 1.51–2.49) and heart rate >90 bpm (hazard ratio: 1.86; 95% CI: 1.45–2.38).

Laboratory Variables

High levels of CRP on admission and at the time of infliximab therapy initiation have been associated with colectomy.[68,69,98] A Danish retrospective study reviewed the records of 56 patients with acute steroid-refractory UC, who received infliximab as rescue therapy.[69] Patients were followed up for a median of 18 months. Patients who progressed to colectomy had significantly higher levels of CRP on admission (P = 0.0036) and at infliximab initiation (P = 0.012), when compared to nonsurgical patients. At first infusion, a CRP > 29 mg/L was predictive of colectomy, whereas CRP < 29 mg/L had a significantly reduced risk of colectomy (P = 0.035).

The recent Monterubbianesi study supported the association between high baseline CRP and progression to early colectomy, following infliximab salvage therapy (RR 2.15; 95% CI: 1.05–4.36; P = 0.003).[98] Kohn et al. followed 86 patients with acute ulcerative colitis, nonresponsive to steroids, who received infliximab, reporting similar findings. Elevated CRP levels on day 3 and 7 following the first infliximab infusion were predictive of colectomy within 2 months (P = 0.006 and 0.001 respectively).[68]

In their retrospective review of 50 ASUC patients treated with infliximab, Gibson et al. reported found that serum albumin >22 g/L at infliximab initiation was associated with a reduced colectomy rate (RR: 0.84, 95% CI: 0.75–0.95, P = 0.003).[74] On univariate analysis, investigators found that a CRP:albumin ratio ≤1.7 approached significance (P = 0.06) and was also predictive of colectomy-free survival.

Levels of CRP may also have a role in predicting responsiveness to ciclosporin rescue therapy. In addition to identifying clinical predictors of ciclosporin failure, Cacheux et al. found that CRP > 45 mg/L, at the time of ciclosporin initiation, were predictive for colectomy within 6 months (hazard ratio 1.70; 95% CI: 1.34–2.16).[97] Prospective evaluation of elevation in CRP as a predictor of nonresponse to salvage therapy is needed.

Raised FC levels have also demonstrated predictive value when assessing nonresponse to infliximab. This was investigated in a prospective observational study.[40] Analysis of 90 patients with ASUC was performed with outcomes defined as colectomy, corticosteroid response and infliximab response. Elevated calprotectin levels on admission were seen in cases that exhibited steroid and infliximab nonresponse as well as those that eventually progressed to colectomy within 1–8 days following infliximab initiation. Of 21 patients, who received infliximab, 11 (52%) did not respond and underwent emergency colectomy. Infliximab nonresponders had higher levels of FC than infliximab responders (1795.0 and 920.5 μg/g; P = 0.06); however, this result was not statistically significant. Prospective evaluation of FC monitoring in ASUC is awaited.

Endoscopic Severity

There is conflicting evidence regarding the relevance of endoscopic severity as a means to evaluate risk of salvage therapy failure. In their retrospective study, Cacheux et al. identified the presence of at least one severe endoscopic lesion at the time of ciclosporin initiation as an independent predictor of colectomy following intravenous ciclosporin salvage therapy (hazard ratio: 2.38; 95% CI: 1.80–3.14); however, Jarnerot et al. concluded that endoscopic severity did not predict for response to infliximab.[64,97] More recently, a randomised prospective study of 113 infliximab-treated patients with steroid-refractory UC, reported that the presence of severe endoscopic lesions at baseline, such as deep ulcerations or spontaneous bleeding on colonoscopy, was independently associated with colectomy following infliximab salvage therapy. On multivariate analysis, severe endoscopic lesions were associated with an increased risk of both early colectomy (RR = 5.13, 95% CI: 1.55–16.96; P = 0.007) and long-term risk of colectomy in the context of early colectomy avoidance (RR = 7.0; 95% CI: 1.09–44.7; P = 0.03).[98]

There is suggestion that 3-month endoscopic outcomes are predictive of longer term progression to colectomy. Three-year follow-up of the Jarnerot trial found that no patients (0/8) in endoscopic remission at 3 months had progressed to colectomy, whereas 50% (7/14) of those who were not in endoscopic remission had required surgery for a new attack of ASUC or steroid-dependent disease (P = 0.02).[6] However, the lack of standardisation of maintenance therapy in this trial, may have influenced colectomy rates.[65]

Therapeutic Drug Monitoring

Therapeutic drug monitoring may play an important role in predicting response to infliximab rescue therapy.

In a prospective cohort study of 115 UC patients given infliximab induction and maintenance therapy, a detectable trough serum infliximab concentration predicted for clinical remission (OR = 12.5, 95% CI: 4.6–33.9, P < 0.001) and endoscopic improvement (OR = 7.3, 95% CI: 2.9–18.4, P < 0.001).[99] Conversely, undetectable trough infliximab levels were associated with higher rates of colectomy (OR = 9.3, 95% CI: 2.9–29.9, P < 0.001). The presence of antibodies against infliximab did not impact on patient outcomes. Similar data was reported in a multicentre prospective study by Brandse et al., currently available only as an abstract. In this study, serum infliximab ≤7 μg/mL at day 42, following week 0, 2, 6 infliximab, was predictive of endoscopic nonresponse (OR = 36, 95% CI: 1.9–719, P = 0.03).[100] Furthermore, a recent prospective cohort study found that faecal infliximab concentrations were significantly higher in patients who experienced primary nonresponse to infliximab induction therapy, compared to patients who experienced clinical response at 2 weeks.[101] These findings support the theory that mucosal changes in severe UC lead to faecal loss of infliximab, which may dampen clinical efficacy. Evidence suggests that ASUC engenders accelerated infliximab clearance from the body – with accelerated clearance linked to hypoalbuminaemia, which has been associated with infliximab treatment failure.[102]

A retrospective analysis of the ACT1/2 studies has corroborated these prospective results.[103] Higher median infliximab concentrations, measured at weeks 8, 30 and 54 following three-dose induction, were associated with higher rates of clinical response, remission and mucosal healing. Multivariate regression analysis correlated low serum infliximab with reduced efficacy, regardless of antibody titre. Although these results have not been replicated in an ASUC setting, they represent a potential area of further research, with the aim of improving the use of biological therapies.

Genetic Variability

Genetic predictors of response in anti-TNF therapy have been identified within the paediatric IBD population. Six susceptibility loci were common amongst paediatric patients who exhibited primary nonresponse to infliximab (P < 0.05) – including the pANCA locus.[104] The presence of two or more loci was associated with a 15-fold relative risk of nonresponse. pANCA seropositivity has also been shown to be associated with early infliximab nonresponse.[105] Homozygous expression of the IL23R genotype was predictive of infliximab response at 16-month follow-up. Further investigation of genetic predictors of response to salvage therapy is required.

Post-salvage Therapy Monitoring

Patients who exhibit primary nonresponse to salvage therapy require a colectomy. If a clinical response is not achieved and symptoms fail to improve or deteriorate, represented by an increase in severity score (Truelove and Witts, Lichtiger, partial Mayo score), prompt liaison with a colorectal surgeon and stomal therapist is advisable to expedite surgery. Early involvement of a surgeon and stomal therapist will allow a patient and their family to make a considered decision about surgery. Sequential salvage therapy is an option only for a subgroup of patients with ASUC and its risks and benefits must be assessed for each patient individually, in consultation with a multidisciplinary team.

In patients who achieve clinical remission with salvage therapy, management may be continued in an out-patient setting. While patients may no longer require high acuity care, they continue to be at high risk of disease relapse. Following discharge from hospital, ASUC patients who have received salvage therapy must remain closely monitored with respect to clinical, biochemical and endoscopic disease activity. Current guidelines do not recommend any particular monitoring protocol, and current practice is largely at the discretion of the individual physician. A disease surveillance strategy that mirrors time points selected in published randomised controlled trials may allow comparison to published data on a case-by-case basis. Frequent re-evaluation of clinical and biochemical parameters, with emphasis on results at 1-, 3- and 12-month follow-up, as well as an endoscopic assessment at 3 months, may help guide decisions on maintenance therapy. If predictors of nonresponse to salvage therapy (Table 6) are present, an intensive maintenance strategy, including combination therapy with azathioprine and infliximab, may be warranted. Conversely, in the absence of these predictors, thiopurine monotherapy in thiopurine-naïve patients may be a sufficient maintenance strategy. Methotrexate may be an alternative bridging strategy among patients who are thiopurine-intolerant.[106]