Cytomegalovirus (CMV) Infection in Inflammatory Bowel Disease: Role, Prevalence, Predictive Features, and Outcomes
Andrea Jones, Thomas C Smyrk, David H Bruining. Mayo Clinic, Rochester, MN
Background: The role of CMV infection in inflammatory bowel disease (IBD) is unknown, particularly when deciding whether to initiate antiviral therapy. Clinical clues (systemic symptoms, exposure to immunosuppressive agents, lack or loss of response to therapy, and marked disease activity) often generate requests for ancillary studies, in the hope that antiviral therapy in positive cases will improve the patient's clinical condition and avoid colectomy. We aimed to assess prevalence of CMV infection in IBD patients deemed at risk, evaluate predictive parameters, histopathology, and clinical outcomes.
Design: We searched the electronic pathology data base for the years 2005-2011 for all patients with IBD in whom CMV was suspected. Slides from all positive cases were reviewed for disease activity, ulcer and number of viral inclusions. Medical records were reviewed for clinical presentation, endoscopy, treatment, and outcome.
Results: CMV was suspected in 1293 IBD patients; 70 (5.4%) had a positive result. 73% patients had received steroids within 2 weeks of the diagnosis. Systemic symptoms were uncommon; only 19% had fever. The endoscopic impression was “severe colitis” in 51% of cases and 77% had visible ulcers. Microscopically, ulcers were identified in 35% of cases. Inclusions were documented by IHC/ISH in 56 and by H&E only in 14 cases. In cases subjected to IHC/ISH, 19 had only a single inclusion, and 37 had more than one inclusion. There were no differences between these two groups in terms of clinical presentation, endoscopic findings or histopathologic activity. When CMV serology was performed, it was negative in 100% of patients with one inclusion, positive in 15% of those with 2-10 inclusions and positive in 70% of those with more than 10 inclusions. 50 patients received antiviral therapy; those with fewer inclusions were treated less often. Follow up endoscopy showed improvement especially in the treated group with higher number of inclusions. 47% of patients eventually had colectomy; treated patients were less likely to have colectomy in the first 60 days after positive CMV diagnosis.
Conclusions: Clinical parameters suggesting CMV infection in IBD are not specific; only 5.4% of biopsy sets had a positive result. Failure to respond to escalating therapy, severe colitis and visible ulcers seem to be the most sensitive clinical clues. The seemingly poor response to antiviral therapy for patients with fewer inclusions may indicate that CMV is an innocent bystander in those cases.
Tuesday, March 5, 2013 9:00 AM
Proffered Papers: Section D, Tuesday Morning