Mistakes in acute severe ulcerative colitis and how to avoid them

Since the fifties, a short intravenous steroid course followed by early salvage colectomy in case of refractoriness has dramatically decreased ASUC mortality from one-third to 1% nowadays. However, some patients are still dying from an ASUC episode. In a British national audit conducted in 2008 and 2010⁵, ASUC mortality was directly related to age and comorbidities: a death rate that was nearly zero in patients less than 60 years old but increased to more than 10% in those aged 80 and more and was 0.4% in patients without comorbidities and 2.4-4.0% in those with.

In practice, the primary ASUC therapeutic goal should be personalized to age and comorbidities: saving life in elderly and/or fragile patients and saving the colon in others. Importantly, most of the deaths observed in cohorts of ASUC patients occurred after colectomy⁶, especially in the case of prolonged hospital stay before surgery (this point is developed below).^7,8

Patients admitted for an ASUC episode may be superinfected by an intestinal microorganism. Indeed, cytomegalovirus (CMV) or Clostridioides difficile are found in 10-30% of ASUC patients at admission.^9,10 These agents may trigger acute intestinal inflammation in UC patients with a loss of bacterial diversity of the gut microbiota. Importantly, specific antimicrobial therapies against CMV or Clostridioides difficile in ASUC patients do not improve the colectomy rate, suggesting that even when initiated by a microorganism infection, acute colonic inflammation in ASUC is mainly related to the disease.¹¹ 

In practice, intestinal superinfections should be ruled out at admission by appropriate investigating for pathogens and Clostridioides difficile toxins in the stools.³ In patients already receiving an immunomodulator at admission with an increased risk of infection, intestinal biopsies and systemic viral replication should be assessed for CMV colitis diagnosis. 

In patients with active IBD have an increased risk of VTE, estimated to be 2-3 fold higher than those without IBD.¹² When adjusted, VTE was associated with more significant mortality among IBD patients, with an odds ratio of 2.5 (95% confidence interval (CI): 1.83-3.43).¹³ Therefore, prophylactic subcutaneous low molecular weight heparin should be given to all inpatients with IBD, especially those admitted for ASUC, without an increased risk of intestinal bleeding exacerbation.

ASUC may be complicated by perforation, massive haemorrhage, or toxic megacolon. Toxic megacolon, described as a colonic dilatation larger than 6 cm, which is often associated with systemic symptoms, confers a high risk of perforation.¹⁴ Patients having complicated ASUC should be referred for emergent colectomy rather than treated medically. Patients with a high degree of suspicion for complications should be screened promptly, if possible, at admission before starting medical treatment, they must be screened promptly for complications by routine blood tests and abdominal imaging – CT-scan rather than abdominal X-ray.

Identification of ASUC is based on simple clinical and biological severity criteria. Original Truelove-Witts criteria that have been slightly modified to replace erythrocyte sedimentation rate by C-reactive protein (CRP) as a marker of systemic inflammation are still recommended by international guidelines for the diagnosis of ASUC3 (Table 1).

Therefore, endoscopic criteria are not part of ASUC diagnosis. Endoscopic assessment by flexible sigmoidoscopy with biopsies for CMV is required in patients admitted for ASUC. Deep ulcerations, corresponding to the Ulcerative Colitis Endoscopic Index of Severity (UCEIS) erosions/ulcerations subscore of 3, may be observed (Figure 1).¹⁵ The significance of these severe endoscopic lesions is still a matter of debate. In retrospective series of patients admitted for ASUC, identifying deep ulcerations was associated with higher risk of treatment failure and colectomy.^16-18 However, these results may be confounding if the decision concerning proceeding to colectomy or not was based on endoscopic findings.

In prospective cohorts, identifying deep ulcerations was predictive of steroid failure. UCEIS, with thresholds of 4 and 7, is one of the components of the Admission Model for Intensification of Therapy in Acute Severe Colitis (ADMIT-ASC) index developed and validated recently for predicting response to intravenous steroids at admission.² In steroid-refractory patients treated by second-line medical therapy, the prognostic value of severe endoscopic lesions was not confirmed in controlled trials. Response rates to ciclosporin or infliximab were not different in patients with or without severe endoscopic lesions.^19,20

Table 1 | ADMIT-ASC score.2 a | ADMIT-ASC score calculation at admission b | Rates of response to intravenous steroids in the whole cohort (discovery and validation) according to the ADMIT-ASC score.

Figure 1 | Pictures of patients admitted for acute severe ulcerative colitis: a | Abdominal X-ray, b | CT-scan c and d | Severe endoscopic lesions.

Intravenous steroids remain the mainstay of ASUC's first-line medical treatment.³ However, its efficacy is limited to 60-70% of patients.²¹ As postoperative morbidity and mortality in UC directly relate to the number of preoperative days elapsed from admission, steroid failure must be anticipated and rapidly assessed.⁷

In the landmark placebo-controlled trial from Lichtiger et al. evaluating intravenous ciclosporin in ASUC patients who had received at least seven days of intravenous steroids at baseline, none of those randomized in the placebo arm who continued steroids responded.²² Predictive indexes of steroid failure have been developed to identify steroid-refractory patients earlier than one week. Assessing the number of bowel movements with CRP levels after three days of intravenous steroids is recommended for predicting steroid failure in the last update of ECCO guidelines.^3,14 Overall, second-line medical treatment initiation or salvage colectomy should be started in the absence of response to intravenous steroids beyond three to a maximum of five days after admission.³

The Oxford group recently developed and validated the ADMIT-ASC index combining UCEIS, CRP and albumin levels at admission to anticipate steroid failure.² Patients admitted with thehighest ADMIT-ASC index, corresponding to those having CRP ≥ 100 mg/l, UCEIS ≥ 7, and albumin ≤ 25 g/l, will fail intravenous steroids in 100% of cases (Table 1). Studies are now awaited in the forthcoming years to determine if intravenous steroid course could be shortened or skipped in ASUC patients with the worst ADMIT-ASC score at admission.

Many molecules developed within the last decades have demonstrated their efficacy in phase 3 trials, including outpatients with moderate-to-severe UC. Notably, these studies excluded patients meeting ASUC criteria at baseline. Therefore, results cannot be extrapolated to inpatients admitted for ASUC who need a drug response within a few days of treatment. Accordingly, subcutaneous anti-TNF (infliximab, adalimumab and golimumab), anti-integrin or anti-interleukin 12/23 and 23 antibodies cannot be offered to ASUC patients due to their progressive efficacy related to their way of administration or mode of action. Intravenous infliximab remains the single monoclonal antibody that has proven its efficacy in this specific setting. ^19,20

Janus-kinase (JAK) inhibitors are small oral molecules that immediately affect UC symptoms, improving patients within the first week of treatment.²³ Due to this speed of onset and rapid plasma clearance, JAK inhibitors are attractive candidates for treating ASUC episodes in patients with prior anti-TNF failure. According to a recent review, cohort studies have shown the clinical benefits of tofacitinib in ASUC with a pooled colectomy avoidance rate of 77%.²⁴ Some studies explored high doses (30 mg per day) as an induction regimen, showing improved efficacy compared to regular doses.²⁵ This supposedly greater efficacy of high tofacitinib doses must be studied further, given the possible safety concerns in this critical population. Positioning tofacitinib or other JAK inhibitors and defining optimal dosing for ASUC patients are currently being explored.

IBD patients are commonly exposed to multiple immunomodulators such as steroids, conventional immunosuppressants (thiopurines, methotrexate, calcineurin inhibitors), biologics and JAK inhibitors. These agents are associated with an increased risk of infections, mainly related to opportunistic microorganisms.²⁶ This risk increases when drugs are combined, especially in case of associated malnutrition, older age, and comorbidities.²⁷ Therefore, ASUC appears as the clinical scenario with the highest risk for opportunistic infections in IBD patients. Cases of Pneumocystis jirovecii infection, some fatal cases, have been observed in ASUC patients receiving immunomodulators.²⁸ Accordingly, standard prophylaxis with trimethoprim-sulfamethoxazole should be strongly considered in patients exposed to three immunomodulators concomitantly, including steroids, immunosuppressants or biologics as well as for those including calcineurin inhibitor.²⁶ The last update of ECCO guidelines on infections in IBD also suggests prophylaxis for Pneumocystis jirovecii infection albeit with a lower evidence level in any case where there is use of high-dose corticosteroids and/or JAK inhibitors, and a low lymphocyte count is observed.

As mentioned before, managing a patient with ASUC is a ‘race against time’. The longer the preoperative medical management, the higher the risk of postoperative complications and death. The postoperative mortality rate is also strongly related to the centre experience.⁷

According to the US Nationwide Inpatient Sample database, hospital experience in colectomies was an essential predictor of in-hospital mortality.7 Mortality was significantly higher in low-volume centres (3 or fewer performed colectomies per year), with an adjusted OR for mortality was 2.42 (95%CI: 1.26-4.63) and in median-volume hospitals (4 –11 colectomies per year) with adjusted OR of 2.02 (95%CI: 1.02-4.01) as compared to high-volume (12 or more colectomies per year) (Figure 2). Centre experience also significantly impacts laparoscopic surgery rate and the type of restorative surgical procedures. When performed in an expert centre, two to three-fold more pouch surgeries were performed than others.^{7, 29} In practice, when ASUC is becoming refractory to first-line medical therapy or is complicated, it is recommended to refer patients to an experienced centre as soon as possible.

Figure 2 | Colectomy mortality according to centre volume.⁷