Introduction Pneumatosis intestinalis (PI) is a rare condition in childhood and the aetiology is poorly understood. Whilst the majority of children makes a full recovery, the outcome is poor in some. Consensus how to manage these patients is lacking.¹

Aim Review of associated risk factors, clinical presentations, and outcome of PI in the paediatric age group.

Subjects and Methods All patients (> 1 month of age) with radiological evidence of PI identified from the radiology database (1991 to 2021) at tertiary children’s hospital were included. Patient records were reviewed retrospectively: Diagnosis, clinical presentation, feeding history, radiological & laboratory findings, management, and outcome were identified. Poor outcome was defined as loss of enteral autonomy, or death within one month of diagnosis of PI.

Results Thirty-one patients (21 male, 67.7%) were included, with median age of five years (3 months -15 years). The underlying diagnosis was heterogenous. Cerebral palsy and acute lymphocytic leukaemia (ALL) were most common (5/31 for each, 16.13%). Twelve patients (38.7%) developed PI 2–15 months post BMT. Most patients (n= 15, 48.4%) had no pre-existing gastroenterological disorder; 7/31 had gut GVHD (22.6%), 3/31 (9.7%) were receiving ECMO. In the majority of patients (11/31, 35%) PI was an incidental finding. Abdominal pain was the most common presentation in symptomatic patients (7/31, 22.6%).

Mode of feeding was variable: 11/31 (35.5%) oral diet, tube feeding 11/31 (35.5%), combination of oral/tube feeding 7/31 (22.4%) and no enteral intake 2/21 (6.4%). PI was diagnosed radiologically in all patients, 3/31 (9.7%) had additional portal venous gas, 2/31 (6.5%) pneumoperitoneum, and 2/31 (6.5%) pneumopericardium. Eighteen patients (58.1%) received steroids at time of diagnosis. All children (31/31, 100%) were managed conservatively with gut rest (median duration 9 days, range 3–21 days); 19/31 (61.3%) required parenteral nutrition. Twenty-five patients (80.6%) made a full recovery. Six (19.4%) patients (3/6 post BMT) had a poor outcome (1/31 permanent feeding intolerance, 5/31 died). When comparing patients who did well (group 1) to those with a poor outcome (group 2) worse prognosis was associated with a lower platelet count (median group 1: 237, group 2: 84.5 ;P value: 0.016), raised CRP (median group 1: 4.5, group 2: 53 ; P value: 0.008), higher creatinine (median group 1: 24, group 2:32; P value: 0.006) and urea (median group 1: 3.2 , group 2: 5.4 ; P value: 0.013).

Summary and Conclusion The overall prognosis of PI in childhood is good, but associated with significant morbidity and mortality in a small number of patients. Our data suggest that lower platelet count, higher urea, creatinine, and CRP levels might risk factors. Unlike previous reports the incidence amongst children undergoing BMT was low.

Reference

1. 1 Nellihela L, et al. Management of pneumatosis intestinalis in children over the age of 6 months: a conservative approach. Arch Dis Child 2018;103(4):352–355.