Pneumatosis Intestinalis After Pediatric Thoracic Organ Transplantation

Abstract

Objective. To review and describe pneumatosis intestinalis (PI) in children who have undergone thoracic organ transplantation and evaluate potential risk factors. Methods. We retrospectively reviewed abdominal radiographs obtained from June 1992 through September 2000 in all pediatric (age <21 years) thoracic organ recipients who survived at least 1 week after transplantation. In this group, a case was defined as an episode of radiographically confirmed PI; those without PI were assigned as controls. Variables analyzed included demographic data, gastroenteritis history (stool cultures or symptoms of gastroenteritis), and transplant-related factors (ie, graft type, rejection history, immunosuppression regimen). Significance was defined as P < .05. Results. Over this 8-year period, PI occurred in 8 (7%) of 116 patients (0.86% annual risk). No child had >1 diagnosed episode of PI. Of these 8 cases, 7 presented with 1 or more abdominal symptoms. Three of these children had rotavirus antigen isolated in their stool, 2 others were noted to have stool positive for Clostridium difficile toxin, and in the other 3, no pathogen was identified. All cases were treated with a regimen of intravenous antibiotics and total parenteral nutrition. There were no deaths; however, 1 patient developed an Aspergillus pulmonary infection during his course of antibiotic therapy, and another underwent an exploratory laparotomy without bowel resection. Significant risk factors included black race (unadjusted odds ratio: 16), younger age at presentation (age <5 years; unadjusted odds ratio: 9), higher steroid dose (steroid dose >0.5 mg/kg/d; unadjusted odds ratio: 7), and a higher tacrolimus level at presentation (tacrolimus level >1; unadjusted odds ratio: 6). PI did not occur with a steroid dose <0.4 mg/kg/d. Variables not associated with increased risk for developing PI included gender, graft type, total white blood cell count, recent antibiotic use, concurrent use of an antimetabolite, cytomegaloviral infection, past use of extracorporeal membrane oxygenation, and graft rejection history. Conclusions. Significant risk factors for the development of PI in our pediatric thoracic organ transplantation population included black race, younger age, higher daily steroid dosing, and a high tacrolimus level at presentation. In the children diagnosed with PI, there were no related deaths, significant gastrointestinal sequelae, or complications. These findings suggest that in this population, PI will often have a benign course when treated aggressively, and that steroid dosing should be reduced to <0.5 mg/kg/d whenever possible.