Abstract
Gastrointestinal food allergies (GIA) in neonates and infants have several manifestations, including severe consequences like septic shock. Because oral milk (or breast milk) challenge after recovery is the most definitive method of confirming the diagnosis, early diagnosis of GIA is not possible. Gastrointestinal endoscopy (endoscopy) is attracting attention as one of the most reliable forms of examination for diagnosing GIA, but it is still controversial because endoscopic findings are nonspecific in many cases. In this study, we examined usefulness of endoscopy in GIA diagnosis and the correlation between clinical severity of GIA and endoscopic findings. We defined GIA endoscopic findings as “macroscopic findings are nonspecific” and “eosinophil infiltration (20 cells per high power field in two or more biopsy tissues).” The method provided sensitivity of 89% and specificity of 75%. It is especially useful for excluding differential diagnoses, such as inflammatory bowel diseases and primary immunodeficiency diseases. We investigated the correlation between clinical severity and three aspects of endoscopic findings, which are as follows : (1) location, (2) features of macroscopic findings, and (3) the degree of eosinophil infiltration in the colonic mucosa. No correlation was found between the severity of GIA and endoscopic findings. These results suggest that the correlation between the clinical severity of GIA and the number of infiltrated eosinophils or degranulation is not relevant for elucidation of the pathological mechanism of GIA.
