Is it possible to diagnose infectious oesophagitis without seeing the causative organism? A histopathological study.

Abstract

We investigated the utility of using histological changes to diagnose infectious oesophagitis when causative organisms cannot be seen. Sixty-seven endoscopic biopsy specimens (51 Candida, 9 herpes simplex virus, 4 tuberculosis, and 3 cytomegalovirus oesophagitis) collected from 2000-2010 that matched the investigative criteria were included in the study. Cases were re-evaluated for histological changes observed in oesophagitis, and the findings were statistically compared using nonparametric tests. Thirty-nine cases occurred in male patients, and 28 occurred in female patients; the mean age of the patients was 51±20.1 years (range, 5-94 years). All cases showed lymphocytic and neutrophilic infiltration; while 27 (40.3%) showed eosinophilic infiltration. The density of lymphocytes and eosinophils were 8.43±6 and 1.07±1.62 per high power field, respectively, and these rates were higher in tuberculosis oesophagitis cases. Lamina propria infiltration was present in herpes simplex virus and Candida oesophagitis. Dense neutrophilic infiltration (>50/high power field) was noted in herpes simplex virus oesophagitis. Candida colonization was observed in 82% of cases with eosinophilic infiltration, and 80% of cases with erosion. Ulceration was present in all tuberculosis oesophagitis cases (p<0.001). Basal cell hyperplasia, papillary elongation, and dilated intercellular spaces were seen in all cases except for 2 Candida oesophagitis cases. Lamina propria fibrosis was especially noted in cytomegalovirus oesophagitis cases. It is not possible to distinguish infectious oesophagitis from other subtypes, especially reflux oesophagitis, if the causative organism is not detected. Clinicopathological correlation and control with repeat targeted biopsies are essential for diagnosis.