30: Predicting Enteral Autonomy in Short Bowel Syndrome in a Large Multicenter Multinational Cohort

Introduction: Previous studies that evaluated predictors of Enteral Autonomy (EA) in pediatric Short Bowel Syndrome (SBS) have conflicting results. Many were frequently limited by small sample size, single-center design, and the use of historical cohorts. The aim of this study was to assess predictors of EA in pediatric SBS in a large cohort of patients from six pediatric intestinal rehabilitation programs. Methods: A multi-institutional retrospective cohort study of infants <12 months with SBS (N=367). The cohort was stratified according to the achievement of EA vs. parenteral nutrition (PN) dependency. Statistical testing included a two-sample t-test, Chi-Square, log-rank for univariate analysis, and Cox Proportional Hazards regression for multivariable analysis. To explore the impact of residual small bowel (SB) and large bowel (LB) length on EA, a scatterplot of percent residual bowel (adjusted for age) divided patients into four anatomical quadrants – ≥ 50% vs. <50% of LB and SB. A Kaplan–Meier curve was created based on these anatomical strata. A time-to-event analysis was also performed comparing EA according to etiology (NEC vs. other etiologies). Results: EA was achieved in 229 patients (62.3%). The mean time to achieve EA was 420 days, while the mean time of follow-up in the PN-dependent group was 1591 days. In the univariate analysis, patients who remained PN dependent were more likely to have shorter percent residual small bowel (30.3 vs 54.4%; p<0.0001) and residual colon length (66.8 vs 87.9%; p<0.001), and less likely to have an ileocecal valve (ICV, 26 vs 68%; p<0.001). In the multivariable analysis, percentage of the residual colon (HR=1.02; 95% CI 1.01–1.02) and small bowel (HR= 1.01; 95% CI 1.01–1.02) length, and presence of ICV (HR= 2.02; 95% CI 1.41–2.88) were positively associated with EA, while the presence of a stoma at the time of shortest bowel length measurement was a negative predictor of EA (HR= 0.72; 95% CI 0.52–1.00). The time-to-event analysis of the 4 anatomical strata revealed that: 80.4% of infants with ≥ 50% SB and LB weaned from PN, after a median time of 209 days; 62.5% with ≥ 50% SB and<50% LB weaned off PN, after a median time of 397 days; 58.3% of infants with < 50% SB and ≥ 50% LB weaned off PN after a longer median time of 1192 days, while only 25.9% of children with < 50% SB and LB weaned from PN (Log-rank p<0.001) (Figure 1). Diagnosis of NEC was not associated with a better achievement of EA (NEC vs. other SBS etiologies (Log-rank p = 0.33) (Figure 2).Conclusions: Overall 62% of infants with IF secondary to SBS achieved EA. Residual small and large bowel length plays a major role in the achievement and duration of time to EA. A colon length of >50% can compensate for the loss of small bowel <50% and account for a similar EA rate as in children with residual SB>50%. Surgical preservation of maximal small and large bowel length early in life is critical for a favorable prognosis in pediatric SBS.