Characterization of esophageal biopsies from stem cell transplant patients with and without esophageal graft-versus-host disease.

Abstract

Graft-versus-host disease (GVHD) is a major complication of hematopoietic stem cell transplantation (HSCT). Histologic diagnostic criteria and several grading systems have been described for colonic GVHD; however, for esophageal GVHD (eGVHD) limited reports exist to date. In this study, a total of 130 HSCT patients with esophageal biopsies were included, with a median age of 44 years (2-77) and a male predominance (54.6%). Of these, 82 (63%) had a clinical diagnosis of eGVHD. Cases were divided into two groups: those without apoptotic bodies, dyskeratotic cells or ulceration (group 1, no histologic evidence of eGVHD) (42%) and those with at least one of those features (group 2) (58%). Group 2 cases were associated with extra gastrointestinal tract GVHD (p=0.024), a clinical diagnosis of eGVHD (p=0.001), older age (p<0.001), stem cells derived from peripheral blood (p<0.001), higher number of intraepithelial lymphocytes (p=0.002), presence of acute inflammation (p<0.001) and basal cell hyperplasia (p=0.016). Apoptotic bodies were seen in 65 (89%), dyskeratotic cells in 27 (37%) and an ulcer in 28 (37%) of the group 2 cases. The sensitivity (Sn), specificity (Sp) and accuracy (acc) of the group 2 cases for a clinical diagnosis of eGVHD was 68.3%, 60.4% and 65.4%, respectively. Apoptotic bodies (p=0.012) and dyskeratotic cells (p<0.001) but not ulceration (p=0.881), were associated with a clinical diagnosis of eGVHD. The Sn, Sp and acc for apoptotic bodies, dyskeratotic cells and ulcer were 59.3%, 63.8% and 60.9%, 30.9%, 95.7% and 54.7%, 21.9%, 79.2% and 43.1%, respectively. Cases with only apoptotic bodies or ulceration were considered as possible GVHD, and those with dyskeratotic cells as likely GVHD, which were associated with GVHD specific survival (p=0.030). This study provides a comprehensive characterization of the esophageal histologic findings in HSCT patients. Further studies are needed to corroborate these findings in other patient populations.