Cytomegalovirus and Inflammatory Bowel Disease
Cytomegalovirus and Inflammatory Bowel Disease
Clinical suspicion of CMV viraemia should be directed towards IBD patients presenting with prominent systemic symptoms, especially fever, lymphadenopathy, splenomegaly, leucopenia and mild hepatitis. However, CMV colitis need not have such features.
Iida et al. reported the endoscopic features of patients with UC and CMV. Patients were subdivided according to the presence of CMV antigen. There were no features that were diagnostic of CMV infection, but 'punched-out ulcers' and 'geographic ulcers' were more common in steroid-refractory patients' with CMV than those without CMV (P = 0.055 and 0.087, for punched-out and geographical ulcers respectively); however, these ulcers were substantially more common in steroid-refractory disease than in patients not receiving steroid therapy suggesting that ulceration was a manifestation of disease severity and not of CMV infection. Suzuki et al. found an association between CMV anti-genaemia and longitudinal ulceration; they claimed that such ulceration was 100% sensitive and 95% specific for CMV.
Omiya et al. took a different approach to a series of 20 in-patients being treated for UC: they treated 10 patients with deep ulceration with anti-viral therapy, but gave standard therapy to the others; all the patients on standard therapy responded, but the outcome for those with deep ulceration was less good: three underwent colectomy. They concluded that the absence of deep ulceration was predictive of latent CMV. This is partially true, 5/10 patients with deep ulceration had CMV DNA in mucosal biopsies, while only 2/10 patients without deep ulceration had CMV DNA. The risk of colectomy appeared greater in those with CMV, 2/7 underwent colectomy, compared with 1/13 without CMV. Arguably the CMV, like deep ulceration, was a marker of severe disease.
Clinical and Endoscopic Features
Clinical suspicion of CMV viraemia should be directed towards IBD patients presenting with prominent systemic symptoms, especially fever, lymphadenopathy, splenomegaly, leucopenia and mild hepatitis. However, CMV colitis need not have such features.
Iida et al. reported the endoscopic features of patients with UC and CMV. Patients were subdivided according to the presence of CMV antigen. There were no features that were diagnostic of CMV infection, but 'punched-out ulcers' and 'geographic ulcers' were more common in steroid-refractory patients' with CMV than those without CMV (P = 0.055 and 0.087, for punched-out and geographical ulcers respectively); however, these ulcers were substantially more common in steroid-refractory disease than in patients not receiving steroid therapy suggesting that ulceration was a manifestation of disease severity and not of CMV infection. Suzuki et al. found an association between CMV anti-genaemia and longitudinal ulceration; they claimed that such ulceration was 100% sensitive and 95% specific for CMV.
Omiya et al. took a different approach to a series of 20 in-patients being treated for UC: they treated 10 patients with deep ulceration with anti-viral therapy, but gave standard therapy to the others; all the patients on standard therapy responded, but the outcome for those with deep ulceration was less good: three underwent colectomy. They concluded that the absence of deep ulceration was predictive of latent CMV. This is partially true, 5/10 patients with deep ulceration had CMV DNA in mucosal biopsies, while only 2/10 patients without deep ulceration had CMV DNA. The risk of colectomy appeared greater in those with CMV, 2/7 underwent colectomy, compared with 1/13 without CMV. Arguably the CMV, like deep ulceration, was a marker of severe disease.
Source...