No abstract

Of 37 samples, 51.4% were PCR+ only, 48.6% were IHC+/PCR+, 16.2% were HE+/IHC+/PCR +, and 32.4% were HE-/IHC+/PCR+. All HE+ were IHC+. Sensitivity for HE was 33%, NPV 61.3%. Kappa agreement between HE and IHC was 0.305 (0.075-0.536). Rates of surgery/death in IHC+/PCR+ patients were higher versus PCR+, with a trend towards significance (60% versus 30.8%, P ¼ 0.07). Death, colectomy, or bowel resection at any time after initial CMV PCR testing was significant in the IHC+/PCR+ group versus PCR-group (60% versus 26.6%, P ¼ 0.014). Rates of surgery were not significantly different between PCR+ and PCR-groups (41.4% versus 26.7%, P ¼ 0.12). Among PCR+ patients, differences in death or surgery with or without antiviral treatment were not significant (34.8% versus 80%, P ¼ 0.13) (Figure). CONCLUSIONS: 14% of moderate-to-severe IBD patients tested for CMV by intestinal tissue PCR were positive. Half of PCR+ samples were HE-/IHC-, reflecting high sensitivity and possible lack of PCR specificity. IHC detected CMV more frequently than HE, with HE sensitivity 33% (as supported in the literature). IHC/PCR positivity is associated with negative outcome compared to PCR positivity alone. However, among PCR+ patients, negative outcomes were not significantly affected by CMV treatment, raising questions of whether CMV acts as pathogen, marker of severe disease, or innocent bystander. Study limitations include referral bias, retrospective design, and small numbers. P-072 YI

No abstract