Infliximab rescue therapy in a case of severe granulomatous colitis associated with rituximab use

Abstract

Colitis occurs in about 4% of individuals treated with rituximab. Optimal management of rituximab-induced colitis, which does not improve with cessation of the drug and supportive care alone, is poorly defined due to limited evidence. Severe refractory disease can lead to colectomy. We present a case of suspected rituximab-induced colitis occurring in a woman in her 70s suffering from rheumatoid arthritis. The patient achieved full clinical, endoscopic and histological remission of colitis with infliximab therapy. The use of biological therapy to treat rituximab-induced colitis can be a potentially organ-saving rescue therapy; however, it must be balanced against the increased risks of immunosuppression in patients already exposed to rituximab. While more evidence is required to fully understand the efficacy and risks of antitumour necrosis factor therapy in this scenario, our case provides an example of the successful use of infliximab for rituximab-induced colitis, which likely helped the patient avoid a colectomy.

Background

Rituximab is an anti-CD20 monoclonal antibody widely used in several haematological and rheumatological conditions. It binds to the surface protein CD20, expressed in B-cells, and induces B-cell death and depletion via a number of mechanisms including triggering direct apoptosis, initiating complement-induced cytotoxicity and recruiting natural killer cells against the B-cell population.¹ One of rituximab’s lesser known side effects is gastrointestinal inflammation, for which the mechanism is poorly understood.² The reported frequency of rituximab-induced colitis is estimated to be 4% in a large retrospective study in patients with cancer.³ Furthermore, patients receiving rituximab are six times more likely to develop, or subsequently be diagnosed with, inflammatory bowel disease (IBD) than the baseline population.⁴

There is very limited experience in treating rituximab-induced colitis with biological medication traditionally used in IBD. This is likely in part due to additional concerns of profound immunosuppression in patients already on rituximab. Therefore, while mild cases often resolve with discontinuation of rituximab,³ in documented cases of severe disease there is a low threshold for patients to undergo surgery.^5–7 In this report, we present a case of severe rituximab-induced colitis successfully treated with infliximab rescue therapy. Full clinical and histological remission was achieved, and the case represents an example to support the use of antitumour necrosis factor (anti-TNF) agents as rescue therapy in similar rituximab-induced colitis presentations to avoid colectomy.

Case presentation

An active woman in her early 70s presented to inpatient gastroenterology following an 8-week history of diarrhoea and left-sided abdominal pain. She had previously presented 2 weeks earlier to the surgical unit where she underwent abdominal CT imaging demonstrating a left-sided colitis from her rectum to the mid-descending colon. She had been initially treated with antibiotics for suspected infectious colitis, but stool samples were negative for a specific pathogen on PCR and therefore antibiotics were stopped. She represented with ongoing symptoms of bloody diarrhoea, a raised C reactive protein (CRP) of 243 mg/L (normal reference range: <10 mg/L), and malnutrition with 20% total body weight loss over the preceding 2 months.

The patient had a medical history of seropositive rheumatoid arthritis (rheumatoid factor and anticyclic citrullinated peptide positive), diagnosed in 2003. Her arthritis was initially treated with sulfasalazine and non-steroidal anti-inflammatory drugs (NSAIDs) before she was escalated to adalimumab therapy in 2011. She stopped adalimumab in 2013, due to secondary loss of response, and was commenced on rituximab therapy. She responded well to rituximab and continuously received this medication until her final dose 2 weeks prior to the onset of her diarrhoeal illness. Up to the point of her presentation, she was independent with a WHO performance status of 0.⁸

The patient was initially started on intravenous hydrocortisone by the admitting acute physician and then oral exclusive enteral nutrition was introduced by the IBD team. Following further dietetic input, a nasogastric tube was inserted due to the patient not meeting her nutritional requirements due to profound loss of appetite. Flexible sigmoidoscopy demonstrated severe patchy inflammation with deep ulceration from the anal canal to the maximal point of insertion in the sigmoid colon, with severe gastrointestinal inflammation corroborated by a faecal calprotectin at >6000 μg/g (normal reference range: < 50 μg/g). Biopsies were taken for histology, cytomegalovirus (CMV) PCR and immunohistochemistry. CMV PCR result was negative. Immunochemistry showed T-lymphocyte infiltration (CD3+), with a complete absence of B-cells (CD20) in keeping with rituximab-induced toxicity. Histology demonstrated features of a severe active chronic colitis with focal ulceration. The crypt architecture was relatively preserved. There were scattered neutrophil crypt abscesses with cryptitis, and non-caseating granulomata were present (see figure 1).

Figure 1.

Baseline endoscopy and histology photos. (A) Active colitis and ulceration. (B) Moderate and severe active chronic colitis with focal ulceration and non-caseating granulomata. (C) CD20 immunostaining demonstrating complete absence of B lymphocytes in the inflammatory infiltrate, compatible with rituximab therapy.

Histological findings were compatible with colonic Crohn’s disease or a drug-induced colitis. Given the history of rituximab exposure, the latter diagnosis was favoured. Due to the severity of inflammation, colorectal surgeons were asked to review the patient to consider colectomy, as supportive care with systemic steroids and exclusive enteral nutrition did not lead to an improvement in the patient’s condition after 3 days. Intravenous steroids were stopped due to lack of efficacy, hyperglycaemia⁹ and concerns of increasing perioperative complications in case of colectomy. This was an IBD multidisciplinary team (MDT) decision in this individual case and it should be noted that systemic steroids are an important therapy in severe ulcerative colitis management.¹⁰ Colonic-release budesonide 9 mg one time a day was commenced as an alternative steroid, given its low risk of systemic side effects. We note that this is off-licence, and that budesonide is generally not recommended in cases of severe colitis.

Despite 2 weeks of supportive care, the patient’s inflammatory markers remained high with ongoing severe symptoms of diarrhoea and malaise. A repeat sigmoidoscopy showed worsening ulceration in the sigmoid colon. The case was rediscussed at IBD MDT, where a decision in keeping with the patient preference to attempt avoiding colectomy was made to administer medical rescue therapy (similar to in acute severe ulcerative colitis). The patient consented to receive infliximab at 5 mg/kg. Within 1 week, her CRP fell from 253 mg/L to 9 mg/L and symptoms effectively resolved.

Differential diagnosis

We considered several differential diagnoses in this patient. Infectious colitis was excluded via stool pathogen PCR and Clostridium difficile testing. We also considered CMV opportunistic infection, given the background of immunosuppression on rituximab. CMV colitis was excluded by performing CMV PCR on colonic biopsies and, additionally, histology demonstrated no CMV inclusion bodies, which are the pathognomonic histological finding in colonic CMV. Ischaemic colitis was part of the considered differentials. The ‘watershed’ areas of the bowel most commonly affected are the splenic flexure and rectosigmoid junction. Patients with chronic inflammatory diseases can be at increased risk of ischaemic colitis. At endoscopy, petechial haemorrhages and longitudinal ulcerations may be seen.¹¹ Histology can be non-specific, although classically shows superficial mucosal necrosis, withered crypts and lamina propria hyalinisation ± haemorrhage. In our case, the microscopic and macroscopic features of the colitis did not fit with ischaemic colitis. Therefore, following investigations, the primary differentials left were new-onset IBD or a drug-induced phenomenon, given the rituximab usage. Both diagnoses were possible and compatible with the clinical, endoscopic and histological presentation. IBD can develop in later life and the granulomatous inflammation seen in this case would favour Crohn’s disease over ulcerative colitis. However, given the history, we felt that a drug-induced colitis related to rituximab was more likely. A careful drug history is needed to elicit potential causes of drug-induced colitis. Common culprits include NSAIDs, isotretinoin, contraceptives, antibiotics and immunosuppressants (such as rituximab itself).

Outcome and follow-up

The patient continued infliximab post discharge which also proved beneficial for her rheumatoid arthritis having responded to anti-TNF medication previously (adalimumab). Colonoscopy 3 months later demonstrated complete mucosal and histological healing (see figure 2). The patient remains in full clinical remission from both her colitis and arthritis.

Figure 2.

Follow-up endoscopy (A) and histology (B). Complete mucosal and histological healing.

Discussion

We have presented a case of severe rituximab-induced colitis that achieved full remission with infliximab rescue therapy. While many cases of rituximab-induced colitis have been published, previous papers have generally reported the use of mesalazine^12–15 and corticosteroids.5 13 15–17 Generally, severe cases have resulted in colectomy.^5–7 We are only aware of a single case series reporting outcomes of biological therapy for this indication, briefly summarising partial success of anti-TNF agents in two patients in that publication.¹⁸

The immunological mechanism by which rituximab can induce colitis is not clear. Indeed, rituximab had previously been studied in clinical trials as a treatment for IBD, although it was found to be of no benefit.¹⁹ The most studied drug-induced colitis syndrome is immune checkpoint inhibitor (ICI) colitis in cancer immunotherapy.^{20 21} These monoclonal antibodies block inhibitory immune pathways, primarily related to T-lymphocyte activity, to amplify the immune response against tumour-associated antigens.²¹ Autoimmune side effects are a well-recognised complication of ICI pharmacotherapy. However, the immunological pathway triggering colitis in rituximab is less clear.

Rituximab-induced colitis may relate to B-cell depletion in the intestinal wall causing immune-dysregulation and T-lymphocyte activation. This corresponds with research suggesting an anti-inflammatory function of B-cells in certain environments,^22–24 partly through interacting with regulatory T-cells via anti-inflammatory cytokines.^{6 25} However, there is histological evidence that B-cell depletion in colon samples does not correlate with gastrointestinal toxicity.²⁶ An alternative explanation may be that intestinal resident plasma cells, non-depleted by rituximab, can undergo autoreactive clonal activity following rituximab exposure.²⁷

Regardless of aetiology, it may be presumed that all drug-induced colitis syndromes could be treated the same. With this logic, the guidelines for ICI colitis would suggest anti-TNF therapy could be used following corticosteroid failure.²⁰ However, our initial concern in employing infliximab was that, unlike with ICIs, rituximab-treated patients have been potently immunosuppressed. Therefore, clinicians need to be aware of the risk of coadministering an anti-TNF agent and theoretically adding to the risk of opportunistic infection. This is a scenario with little clinical evidence. It should be noted, however, that high-dose corticosteroid therapy also confers this risk, and steroids have been associated with poorer postoperative outcomes in IBD surgery.²⁸

The alternative to biological immunosuppression was colectomy for this patient. Therefore, infliximab was given as an organ-saving treatment. We do note anti-TNF therapy has been reportedly given alongside rituximab in other case reports relating to non-gastrointestinal inflammatory diseases.^{29 30} A consideration in similar cases, however, could be finding a less systemically immunosuppressing therapy, such as vedolizumab.³¹ However, in our case, infliximab was chosen following consultation with our rheumatology team as it was felt this could also benefit the patient’s joint disease.

Learning points.

• Drug-induced colitis can mimic or induce inflammatory bowel disease (IBD) resulting in significant gastrointestinal inflammation. An awareness of this phenomenon, and culpable medications, is paramount to identifying cases quickly.

• Rituximab therapy is a known cause of drug-induced colitis, occurring in ~4% of treated patients. The exact immunological mechanism triggering this is unknown.

• The mainstay of therapy for rituximab-induced colitis is supportive care and cessation of the drug. However, surgery can be required in severe cases.

• Our report supports the use of anti-TNF therapy for severe refractory cases. This should be used cautiously in a well-consented patient, given theoretical risks of profound immunosuppression.

• More clinical outcome data are required to better understand the exact risks and efficacy of biological immunotherapy in rituximab-induced colitis.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s)