Gastrointestinal graft-versus-host disease (GI-GVHD) is a potentially life-threatening complication after hematopoietic stem cell transplantation. Symptoms indicating GI-GVHD motivates endoscopy with biopsy sampling and histopathological confirmation. Optimal extent of endoscopy in children is, however, presently unknown. Therefore, we aimed to evaluate whether biopsies from the rectosigmoid area versus the rest of the colon/ileocolon with or without biopsies from simultaneous upper endoscopy, were equally reliable for detection of GI-GVHD and relevant differential diagnoses.Methods:
Retrospective multicenter study based on histopathological re-evaluation of biopsies and hospital record data, collected from children with suspected GI-GVHD.Results:
Forty-four children with 51 endoscopic occasions (81 procedures) were included. Thirty-nine of 51 (76.5%) were diagnosed as GI-GVHD, 14 (27.4%) received a differential diagnosis and 7 (13.7%) had normal histology findings. Comorbidity, that is, simultaneous detection of a differential diagnosis and GI-GVHD, was observed in 9 (23.1%) cases. Cytomegalovirus infection was the most frequent differential diagnosis, 6 of 7 were detected in biopsies from rectosigmoid and esophagogastroduodenal areas. Sensitivity for detection of GI-GVHD in biopsies collected from rectosigmoid-ileocolonic-, rectosigmoid-, or esophagogastroduodenal areas were 97.4%, 84.6%, 83.3%, respectively, and 97.4% when the latter 2 were merged. The difference, nondetected GI-GVHD in the rectosigmoid area versus detected elsewhere in the GI tract, was statistically significant (P = 0.03).Conclusions:
Biopsies collected from the rectosigmoid area solely were not optimal for detection of pediatric GI-GVHD. When biopsy sampling from rectosigmoid and upper GI tract areas was combined, the sensitivity for GI-GVHD was, however, equally high as for ileocolonoscopy or full upper and lower endoscopy.