Eosinophilic colitis: experience in a large tertiary hospital.

BACKGROUND Eosinophilic colitis (EC) is a rare and ill-defined entity with an unknown pathogenesis and an unsatisfactory treatment response. The standard histopathological criteria for EC diagnosis lack specificity and not all the cases fulfilling those criteria are considered clinically as EC and treated. The objective of this study is to refine diagnostic criteria for EC. METHODS Retrospective study of all the cases with a histopathological diagnosis of EC in Hospital Clinico San Carlos (Madrid, Spain) from 2006 to 2016. We have reviewed their clinical and pathological features and tried to define the features differentiating cases considered EC on clinical grounds. RESULTS We identified 106 EC cases. In 22 cases, a clinical EC diagnosis was established. Confirmed EC was associated with younger age, female gender, diarrhea, higher maximum number of eosinophils÷HPF (high-power field), intraepithelial eosinophils, architectural distortion and absence of acute inflammation. We chose a cut-off point of 40 for both mean and maximum number of eosinophils÷HPF. A mean number of eosinophils÷HPF higher than 40 was related to architectural distortion, mucosal atrophy, signs of eosinophil activation and submucosal infiltration. Cases with a maximum number of eosinophils÷HPF higher than 40 showed more architectural distortion, intraepithelial eosinophils, submucosal infiltration and lack of lymphoplasmacytic infiltration. CONCLUSIONS Histopathological diagnosis of EC is not well correlated with clinical EC. An increase in specificity can be achieved by raising the cut-off point to 40 eosinophils÷HPF and by combining mean and maximum number of eosinophils with other microscopic and clinical features suggestive of EC.