Introduction

Patients with inflammatory bowel disease (IBD) experiencing severe flares are at increased risk of opportunistic infections, particularly cytomegalovirus (CMV), due to immunosuppressive therapy.

Aims & Methods

This study aimed to assess the prevalence and risk factors associated with CMV infection during severe acute colitis.
We conducted a retrospective analytical study on hospitalized patients with severe acute colitis. CMV infection was confirmed by positive anti-CMV IgM serology, positive colonic biopsy PCR for CMV DNA, or histological evidence of CMV cytopathic effect. Patients were divided into two groups: Group 1 (CMV-positive) and Group 2 (CMV-negative). Clinical, biological, and therapeutic data were compared (SPSS, p<0.05).

Results

Among 41 patients (mean age 39.8 years, 14 men), 17 (41.5%) had confirmed CMV infection. CMV was diagnosed by PCR (n=13), serology (n=3), or histology (n=1). Most received antiviral therapy (Ganciclovir or Valganciclovir). One patient required emergency colectomy; three experienced relapse, with two receiving Foscarnet and one undergoing colectomy for bleeding. No significant association was found between CMV infection and age, sex, disease duration, corticosteroid or immunosuppressant use, anemia, hypoalbuminemia, or CRP >45 mg/L. However, CMV infection was significantly associated with steroid resistance (p=0.003).

Conclusion

CMV infection is frequent during severe IBD flares and is strongly associated with corticosteroid resistance. Early recognition of this risk factor is essential for timely diagnosis and management to avoid complications.

Introduction

Patients with ulcerative colitis (UC) and concurrent cytomegalovirus (CMV) colitis are at higher risk of poorer outcomes, including developing steroid refractoriness or toxic megacolon. Given its relevance, the European Crohn's and Colitis Organization recommends, in case of flare-ups, to search for CMV by endoscopic biopsy, preferably with tissue polymerase chain reaction (PCR) or alternatively with immunohistochemistry (IHC).

Aims & Methods

The aims of our study were to compare the diagnostic yield of tissue and plasma CMVDNA PCR to IHC and to explore the different level of tissue viral load in both steroid-refractory and steroid-sensitive UC patients. This prospective study included consecutive patients undergoing colonoscopy or proctosigmoidoscopy for UC, independent of disease activity, in a single tertiary center betwen July 2024 and January 2025. Patients underwent the same day endoscopy with multiple biopsies (3 from sigmoid colon and 4 from rectum) and plasma sampling for CMVDNA analysis using real-time quantitative PCR assay; histological presence of CMV was evaluated using IHC and hematoxylin-eosin staining. Concordance among CMV detection techniques was calculated using K cohen concordance index. Continuous variables were compared using Mann-Whithney test. ROC analysis allowed to identify an ideal cut-off of viral load value of CMV infection.

Results

Twenty-one consecutive patients were enrolled. Twelve (57,1%) patients were male and 9 (42,9%) female, (range 19-79 years). Nine (42,8%) patients had a diagnosis of pancolitis, three (14,3%) of left-sided UC, seven (33,4%) of procto-sigmoiditis and two (9,5%) of proctitis. Six (28,6%) patients were subclassified as steroid-refractory, the remaining fifteen (71,4%) as steroid-responsive.
IHC positivity was observed in tissue samples from 3 (14,3%) patients, while CMVDNA was detected in tissue samples from 5 (23,8%) patients and in 3 (14,3%) plasma samples. Comparing IHC to plasma CMVDNA, the k index was only fair consistent (0.277 CI -0.327 to 0.758), while IHC versus tissue DNA showed a higher consistency, with a k index of 0.500 (CI 0.048 to 0.952).
Interestingly, the mean value of tissue CMVDNA in the steroid-refractory group was higher, although not significant, than the steroid-responsive group (respectively 140.00 copies/mg ±260.00 vs 98.00 copies/mg ±360.00, p value= 0.14). In order to identify a possible tissue CMVDNA level able to distinguish between the two groups, a receiver operating characteristic (ROC) curve analysis was performed: a cut-off value of 102.56 was associated with a sensitivity of 92.86% and a specificity of 33.3%, resulting in a likelihood ratio of 1.39. The area under the ROC curve (AUC) was 0.6667 ± 0.1415, indicating a modest discriminative ability.

Conclusion

This study showed a good concordance between immunohistochemistry and tissue PCR for the detection of CMV in UC patients. Steroid-refractory UC patients showed a higher tissue viral load compared to steroid-responsive patients.

Disclosure

No conflict of interest to declare