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

Predicting Disease Course and Requirement for Salvage Therapy

Optimal management of ASUC relies on timely and repeated evaluation of disease response to therapy, in order to improve clinical outcomes. One of the major challenges faced by clinicians is the absence of a widely accepted management and monitoring protocol.[11,12] In particular, three key stages of assessment in the acute setting have been investigated in the literature: (i) on admission, (ii) after initiation of steroid therapy and (iii) after initiation of salvage therapy. More recently, studies have also focussed on long-term outcomes; however, further prospective studies in this area are required.

Steroid Therapy

Preliminary risk stratification is generally performed according to Truelove and Witts' criteria (Table 1), with intravenous corticosteroids remaining the first-line therapy for severe flares. Management of severe flares is time-limited. Current guidelines recommend the use of hydrocortisone 100 mg intravenously q.d.s. or methylprednisolone 60 mg/day, administered via multiple intravenous injections or a continuous 24-h infusion.[13] There are few trials that have focussed on assessing steroid response at varying doses in ASUC. Oral prednisolone dosing has been evaluated in ambulatory patients with mild disease, with results suggesting that 60 mg/day was as effective as 40 mg/day; however, comparative studies have not been performed in the ASUC setting.[14] While studies have not yet established an optimal corticosteroid dosage, higher doses have not been shown to improve outcomes.[8,15] The use of intravenous corticosteroids has reduced mortality in ASUC from almost 30% to <1%.[16] Systemic steroid therapy was first established as a treatment for UC in a landmark trial by Truelove & Witts.[7] Outcomes following intravenous hydrocortisone were evaluated in a total of 210 patients with acute flares of varying severity. After 6 weeks of therapy, higher rates of clinical remission were seen in steroid-treated patients when compared with placebo controls [41% (45/109) and 16% (16/101)], and a greater proportion of patients in the cortisone group experienced endoscopic improvement. Longer term data indicated that rates of clinical remission remained higher in the cortisone group in those admitted with a first attack of UC; however, after 9 months, any clinical benefit associated with steroid therapy was lost in comparison to those who initially presented with a relapse of their disease. A subsequent study of 49 patients with severe exacerbations of UC reported a clinical remission rate of 73%, 5 days after receiving intravenous methylprednisolone.[17] There have been no head-to-head trials of intravenous hydrocortisone compared to methylprednisolone, and the optimal steroid dose for induction of remission in ASUC is unknown. Nonetheless, steroid efficacy in ASUC has been the focus of a large number of studies. A 2007 systematic review compiled the results of 26 studies between 1974 and 2006, assessing corticosteroid response in 1948 adult patients with moderate to severe UC flares. These data indicate that intravenous corticosteroids play a significant role in reduction morbidity and mortality associated with ASUC.[8]

Adjunctive Therapy

Discontinuation of nonsteroidal anti-inflammatory drugs, narcotics and anti-cholinergic agents is recommended to minimise the risk of further disease exacerbation. In addition, careful symptom monitoring is needed in patients started on an aminosalicylate agent during an acute flare, with prompt discontinuation of therapy if there is an association with worsening symptoms. The use of antibiotics has not been shown to improve patient outcomes in ASUC. Clinical trials have evaluated numerous antibiotics, including metronidazole, tobramycin and ciprofloxacin as potential adjuncts to corticosteroid therapy; however, there is currently no evidence to support their use in the absence of any infectious complications.[18–20] Similarly, despite early evidence suggesting otherwise, studies have not found parenteral nutrition to provide any clinical benefit in ASUC, when compared to a regular oral diet.[21–27] Instead, enteral feeding has yielded better outcomes in a head-to-head comparison with parenteral nutrition in this population.[24] In a randomised trial of 42 patients with moderate to severe flares of UC, enteral nutrition increased median serum albumin by 16.7% [−0.5% to 30.4%], whereas parenteral nutrition was associated with only a 4.6% gain [−12.0% to 13.7%] (P = 0.019). Enteral feeds were also associated with fewer complications (9% and 35%, P = 0.046); however, greater evidence is required before supplemental nutrition with enteral feeding (above and beyond oral nutrition supplements) can be routinely recommended.

Risk Stratification

Given that 31–35% of severe attacks of UC in the adult population are steroid-refractory, methods to predict disease course and the requirement for salvage therapy have become a priority.[8] The reason for this is twofold. In the context of minimal response to intravenous steroids, early steroid cessation firstly, prevents unnecessary immunosuppression, and secondly, facilitates earlier induction of salvage therapy – potentially improving outcomes.[28] Protracted steroid therapy beyond 7–10 days is not recommended and studies have proposed that 3–5 days is sufficient to determine likelihood of remission.[29,30]

According to European Crohn's and Colitis Organisation (ECCO) guidelines, published in 2008, steroid-refractory colitis is defined as active disease that remains present despite up to 0.75 mg/kg/day prednisolone over the course of 4 weeks.[31] However, as acknowledged by the authors, greater usage of newer biologic therapies may necessitate reworking of this definition. Steroid-refractoriness can be predicted with an evaluation of clinical, biochemical, endoscopic and radiologic markers.[32–35] The monitoring of clinical signs – predominantly, the number of bowel motions per day and presence of rectal bleeding, and acute phase reactants (CRP, ESR) has been advocated in multiple studies.

Prognostic stratification should begin on admission. Greater biological severity on admission, based on Truelove and Witts' criteria (Table 1), and more extensive underlying disease have been shown to correlate with increased rates of progression to surgery.[1,36] In particular, UC extending proximal to the splenic flexure has been associated with poorer prognosis and is predictive of steroid-refractory disease.[36] Retrospective data published by Dinesen et al. demonstrated that colectomy rates in ASUC increase in proportion to the number of Truelove and Witts' criteria satisfied. Patients admitted with one criterion in addition to ≥6 bloody stools per day had a colectomy rate of 8.5% (11/129), while patients with two additional criteria underwent colectomy at a rate of 31% (29/94). The highest colectomy rate, 48% (34/71), was observed in patients with three or more additional criteria on admission.[1]

An Italian prospective cohort study determined that, on admission, elevation in ESR above 75 mm/h and body temperature >38 °C were associated with a 4.6 and 8.8-fold increased risk of colectomy, respectively.[33] Persistence of diarrhoeal symptoms despite intravenous corticosteroids has also been associated with poorer outcomes. A less than 40% reduction in bowel motions after 5 days of steroids has been shown to be predictive of eventual treatment failure, whereas cessation of rectal bleeding was predictive of steroid responsiveness.[28,33]

Following 3 days of intravenous steroids, established prognostic models should guide further management. In a study by Travis et al., 51 cases of ASUC were treated with intensively scheduled 6-hourly 100 mg doses of intravenous hydrocortisone and rectal steroid enemas.[9] Low stool frequency and CRP on day 3 following steroid initiation were predictive of treatment response – with clinical remission defined as ≤3 stools/day with the absence of visible blood after 7 days. In contrast, patients with >8 stools/day, or 3–8 stools/day with CRP >45 mg/L had an 85% risk of steroid failure and progression to colectomy (PPV 85%) (Table 2).

The Ho risk index (Table 3) was derived from findings of a Scottish cohort study, which identified key predictors of steroid failure in 167 patients with ASUC.[37] A scoring system was established with regression analysis based on statistically significant variables – including day 3 stool frequency, serum albumin and the presence of colonic dilatation. The study showed that 6–9 daily bowel motions in conjunction with hypoalbuminaemia (<30 g/L) had a 45% positive predictive value for colectomy. The presence of colonic dilatation or >9 daily stools was predictive of nonresponse in 85% of patients. Hypoalbuminaemia also correlated with poorer outcomes in a Scottish retrospective study involving 39 patients with steroid-refractory ASUC.[38] Serum albumin levels of <34 g/L on day 3 of steroid therapy (seen in 31/39 patients) were found to be predictive of colectomy – 9/18 patients (50%) with albumin <34 g/L underwent colectomy compared to 1/13 patients (8%) with >34 g/L; (P = 0.02, OR = 12.0, 95% CI: 1.28–112.7). In this study, hypoalbuminaemia had a sensitivity of 57.1% (95% CI: 34.0–78.2%) and specificity of 90.0% (95% CI: 55.5–99.8) for colectomy.

Studies have also advocated the measurement of faecal markers, in particular faecal calprotectin (FC) and lactoferrin, and more recently S100A12 and M2-pyruvate kinase.[39,40] However, the utility of faecal markers in comparison to clinical monitoring is yet to be determined as the former are generic measures of colonic inflammation, lack specificity for the disease process and are limited by large inter-individual variation. While FC levels appear to be reliable markers of treatment response during steroid therapy, levels measured early in the disease course (days 0 and 4 post-steroid administration) have not been found to correlate with disease prognosis beyond 12 months.[41] In the paediatric setting, clinical monitoring with the Paediatric Ulcerative Colitis Activity Index (PUCAI) has been shown to predict for short-term colectomy more accurately than any faecal marker.[39]

Radiologic predictors of disease course include abdominal plain films to screen for colonic dilatation (>5.5 cm). Typically, the presence of colonic dilatation is indicative of the need for high acuity care and confers a 75% positive predictive value for progression to colectomy.[42] The presence of mucosal islands and small bowel ileus, evidenced by three or more small bowel loops of gas on X-ray, has also been shown to predict medical therapy failure, with the latter associated with colectomy in 73% of patients.[43] A potential role for colonic MRI has been explored in the literature. The total colonic inflammation score (TCIS) was devised to enable standardised MRI assessment of severity and prognosis in ASUC.[44] The score, ranging from 6 to 95, was derived from radiologist assessment of T2-weighted images, on the basis of segmental haustral loss, mural thickness, oedema and colonic dilatation. TCIS on admission was shown to correlate with levels of CRP (Kendall's tau = 0.45, 95% CI: 0.11–0.79, P = 0.006), and stool frequency (Kendall's tau = 0.39, 95% CI: 0.14–0.64, P = 0.02). Accordingly, the TCIS fell after therapy indicating a decrease in levels of inflammation. Unlike CRP and stool frequency, radiological scoring correlated with length of hospital stay (Kendall's tau 0.40, 95% CI: 0.11–0.69, P = 0.02). Among patients discharged within 1 week of admission, 14% (1/7) had an admission TCIS >20, whereas among patients who required more extensive admission for 1 week or longer, 64% (7/11) had an admission TCIS >20 [PPV 0.9, NPV 0.5, sensitivity 65% (95% CI: 36–86) and specificity 86% (95% CI: 42–99)]. Overall, these data suggest that the MRI-based TCIS may represent a radiological counterpart to established laboratory markers in UC. However, effective implementation of this approach is contingent on accessibility and availability of appropriate imaging modalities.

There is an emerging role for endoscopic risk stratification in identifying patients likely to require salvage therapy. A recent retrospective study of 89 patients with ASUC correlated greater endoscopic severity on admission with poorer disease outcomes.[45] Endoscopic severity was classified according to the Ulcerative Colitis Endoscopic Index of Severity (UCEIS), accounting for a range of endoscopic findings, including vascular pattern, active bleeding and the presence and depth of colonic erosions and ulcers.[46] Study outcomes were defined as colectomy, readmission, death and the requirement for rescue therapy, with a median follow-up duration of 13 months. In total, 36 patients (40%) required either infliximab or ciclosporin rescue therapy. Of patients with an UCEIS score ≥7/8, 11 of 14 (79%) required salvage therapy. Similarly, 27 of 54 (50%) patients with UCEIS score ≥5/8 needed salvage therapy; in contrast to only 9 of 33 (27%) patients with UCEIS score <5/8 (P = 0.037).

European Crohn's and Colitis Organisation guidelines define an adequate steroid response as a decrease in activity index of >30% with a reduction in rectal bleeding and endoscopic subscore.[31] At present there is no standardised activity index for ulcerative colitis and multiple activity indices are utilised in clinical practice.[47] While numerous studies have established the importance of monitoring clinical and laboratory variables alongside radiological and endoscopic parameters, the criteria for steroid nonresponse and predictors of nonresponse in ASUC remain relatively heterogeneous.