Morphometric histology for infant gastroesophageal reflux disease: evaluation of reliability in 497 esophageal biopsies.

Objectives: We sought to determine the reliability of morphometric measurements on infant esophageal biopsies using a light microscope with eyepiece micrometer.

Methods: We measured epithelial thickness, basal layer thickness (B), papillary height (P) and epithelial lymphocyte and eosinophil numbers on approximately 500 existing esophageal suction biopsies from infants previously evaluated for reflux esophagitis. We tested these measurements for interobserver, test-retest and internal consistency reliability.

Results: Infants ages 0.25 to 23.75 (median, 6.25) months provided 497 biopsies. Both investigators scoring the biopsies independently judged 93% of them scorable. Of the biopsies scored by both, the 2 readings were within 0.15 of each other for P in 97% and for B in 81%. In addition to these correlative measures of consistency, categoric measures demonstrated that 373 (89%) of the 420 scorable biopsies with visible papillae produced agreement as to P being abnormal (317, 85%) or normal (56, 15%). Similarly, 360 (78%) of the 463 scorable biopsies produced agreement as to B being abnormal (339, 94%) or normal (21, 6%). P values were 0.17 to 0.94 (median, 0.67), and B values were 0.13 to 0.91 (median, 0.34). Lymphocytes numbered 0 to 40 (median 5) per high-power field. Only 12% had any eosinophils; none of those with completely normal morphometrics had any eosinophils; and only 2% had >5 eosinophils per high-power field.

Conclusions: Simple quantitative esophageal histological morphometric parameters are reliably measurable on suction biopsies from infants using a light microscope fitted with an ocular micrometer, even by nonpathologists.