Acute severe colitis failing steroids on day 3: rescue or surgery?
A 29-year-old woman with left-sided ulcerative colitis diagnosed 3 years ago, maintained on mesalazine 3 g daily, is admitted with 10 bloody stools per day, urgency and fever for 5 days. Temperature is 38.3°C, pulse 110, blood pressure 100/62, haemoglobin 8.9 g/dL, CRP 58 mg/L, albumin 25 g/L. She has never taken steroids or biologics. Abdominal radiograph shows a transverse colon diameter of 5.8 cm with preserved haustra and no free air. Stool culture, C. difficile NAAT and ova and parasites are negative. She receives intravenous methylprednisolone 60 mg daily, fluids, venous thromboembolism prophylaxis, and transfusion if haemoglobin falls below 7 g/dL. On day 3 she has passed 9 stools, remains febrile at 38.1°C, and has mild lower abdominal tenderness without rebound or guarding. CRP is 61 mg/L.
1. Using the day-3 data, what is your decision and what criteria are you applying?
Read the worked answer
This is acute severe colitis by the Truelove and Witts criteria (six or more bloody stools daily plus systemic disturbance), and she has failed intravenous steroids. The objective day-3 assessment is the Travis index: more than eight stools in 24 hours, or a CRP above 45 mg/L in the presence of three to eight stools, predicts that about 85% of patients will need colectomy. She meets both criteria - nine stools and CRP 61 - so continuing steroids alone is not defensible.
The decision is to move to second-line therapy or surgery today, not to wait another 3 days. Before that, three things must be checked: an abdominal radiograph to exclude colonic dilatation, a stool test for C. difficile (done, negative) and CMV assessment, and a surgical review with stoma counselling. The absence of rebound tenderness and a transverse colon under 6 cm mean there is no immediate indication for emergency colectomy, so rescue medical therapy is reasonable.
Equally important is what not to do: do not give loperamide or other antimotility agents, do not delay surgery because the CRP is 'only' 61, and do not start infliximab before excluding active infection. If she develops fever with rigors, worsening tenderness, dilatation beyond 6 cm, perforation or uncontrolled haemorrhage, surgery becomes the priority and medical rescue is abandoned.
2. What is the role of CMV testing here, and how would a positive result change your management?
Read the worked answer
Cytomegalovirus reactivation is found in a substantial minority of patients with steroid-refractory severe colitis and is associated with worse outcomes, so it must be looked for before or alongside rescue therapy. Serum CMV DNA is convenient but insensitive and does not prove colonic disease; the decisive test is sigmoidoscopy with rectal biopsies for haematoxylin-eosin histology and immunohistochemistry, and viral culture or tissue PCR. This can be done at the bedside without full bowel preparation and does not delay treatment significantly.
If tissue immunohistochemistry or histology shows CMV inclusions, she should receive antiviral therapy, usually intravenous ganciclovir 5 mg/kg twice daily, for 2-3 weeks, with dose adjustment for renal function and monitoring of blood counts. A positive serum PCR alone, with negative tissue studies, is not an indication to treat but should prompt repeat tissue sampling if she deteriorates.
The presence of CMV does not automatically mean colectomy, but it makes steroid escalation unattractive. In practice, antiviral therapy is combined with either surgery or a rescue biologic, and the choice depends on how sick she is. The trap is to attribute all deterioration to CMV and persist with steroids, or conversely to ignore CMV and start a biologic with an uncontrolled opportunistic infection.
3. How would you choose between infliximab, ciclosporin and colectomy for rescue, and what safety steps precede it?
Read the worked answer
All three are legitimate, and the choice depends on patient preference, local expertise, infection risk and the ability to monitor. Infliximab 5 mg/kg is the usual first rescue: it is familiar, effective, and allows maintenance without calcineurin toxicity. In a patient with albumin 25 g/L, accelerated dosing (10 mg/kg or a second dose at 1 week) is often used because drug clearance is high; this is a judgement call, not a protocol certainty. Ciclosporin 2 mg/kg intravenously is equally effective for short-term rescue, but requires level monitoring, magnesium and cholesterol checks, blood pressure monitoring, and carries risks of nephrotoxicity, seizures and opportunistic infection; it is a poor choice if renal function is borderline or the patient cannot be monitored closely.
Colectomy is the right choice when there is toxic dilatation, perforation, uncontrolled bleeding, or when the patient prefers surgery to prolonged immunosuppression - and it must be offered, not presented as failure. Before either drug, screen for hepatitis B, hepatitis C, HIV and latent tuberculosis, check baseline blood counts and renal and liver function, and exclude abscess or perforation by imaging. Live vaccines are contraindicated after starting.
After rescue, response is judged at 24-72 hours by stool frequency, CRP and tenderness. If there is no response by day 7, surgery should be reconsidered rather than a third medical manoeuvre. Maintenance after infliximab rescue should include an immunomodulator or scheduled infliximab, and long-term colectomy risk should be discussed honestly.