Efficacy and safety of mucous fistula refeeding in preterm infants: an exploratory randomized controlled trial

**Background:** Premature infants often require enterostomy creation for conditions like necrotizing enterocolitis and spontaneous intestinal perforation. High stoma output can lead to delayed growth, dehydration, electrolyte abnormalities, and prolonged parenteral nutrition (PN) with its associated complications. Mucous fistula refeeding (MFR), the practice of collecting proximal ostomy effluent and reinfusing it into the distal mucous fistula, has been proposed to prevent distal bowel atrophy and promote nutrient absorption. Prior evidence was limited to retrospective studies; no prospective randomized trial had been conducted.

**Methods:** This single-center exploratory RCT was conducted at Seoul National University Children's Hospital NICU from July 1, 2015 to November 11, 2019. Preterm infants born before 35 weeks' gestation with enterostomy were eligible. Exclusion criteria included congenital anomalies, blind pouches, stricture, and unstable vital signs. When full enteral feeding (>120 mL/kg/day) was achieved, infants with stoma output ≥40 mL/kg/day were assigned to the high-output MFR group (n=4 initially). Those with output <40 mL/kg/day were randomized 1:1 to normal-output MFR or control groups (n=8 each). Two control infants were later reassigned to high-output MFR when output exceeded 40 mL/kg/day. MFR was performed by collecting proximal stoma output every 8 hours and delivering 50% into the distal fistula initially, progressing to 100% every 4 hours via a 3- or 4-Fr Nelaton catheter. Primary outcomes were growth (z-scores of weight and length), colon diameter on loopogram, and serum citrulline levels. Secondary outcomes included PN duration, distal ileum histology at stoma closure, and safety. Statistical analysis used chi-squared tests, Mann-Whitney U tests, and ANCOVA adjusted for postmenstrual age (PMA) and body weight z-score.

**Key Results:** Twenty infants were finally analyzed (5 per group after exclusions and reassignments). Gestational age (25 vs. 26+6 vs. 29+0 weeks) and birthweight (540 vs. 880 vs. 1,190 g) were lower in the high-output MFR group (p<0.05), but comparable between normal-output MFR and control groups. Total PN duration from stoma formation to off-PN did not differ significantly between normal-output MFR and control groups (50.5 vs. 44 days, p=1.000). At reanastomosis, z-scores for bodyweight (−2.19 vs. −3.82) and length (−3.08 vs. −4.33) were numerically larger in the normal-output MFR group. At 3 months after reanastomosis, weight (−0.96 vs. −2.98) and length (−1.77 vs. −4.05) z-scores remained numerically larger in the MFR group, though not statistically significant. Colon diameter on loopogram before reanastomosis was significantly larger in the normal-output MFR group: transverse colon (10.15 vs. 6.05 mm, p=0.002), descending colon (10.22 vs. 6.06 mm, p=0.005), and sigmoid colon (12.48 vs. 7.07 mm, p=0.037). Serum citrulline levels did not differ significantly between groups, though median values were higher in the MFR group just before reanastomosis (31.19 vs. 26.29 µmol/L) and at 12 weeks after reanastomosis (35.45 vs. 27.23 µmol/L). The control group had more frequent severe chronic inflammation of the distal ileum (75% vs. 20%). In the combined MFR group (high + normal output), growth velocity increased significantly after MFR: weight (9.53 vs. 23.36 g/day, p<0.001) and length (0.09 vs. 0.15 cm/day, p=0.003), both remaining significant after PMA adjustment. Regarding safety, one bowel perforation occurred during manual reduction for stoma prolapse, requiring MFR termination. Two cases of culture-proven sepsis occurred during MFR (Klebsiella aerogenes and Staphylococcus epidermidis), though the direct relationship with MFR was unclear. Four additional sepsis cases occurred in the overall cohort without MFR interruption.

**Clinical Implications:** This first prospective RCT of MFR in preterm infants demonstrates that MFR benefits growth and intestinal adaptation, as evidenced by significantly increased growth velocity and colon diameter. MFR appears relatively safe when performed with a standardized protocol using appropriate catheter size and careful monitoring. The trend toward improved weight and length z-scores in the MFR group, despite lower baseline weight, suggests MFR may help overcome growth deficits. The higher rate of severe chronic inflammation in the control group's distal ileum supports the hypothesis that MFR maintains mucosal integrity. However, the occurrence of infectious complications warrants careful monitoring and further research. The small sample size limits definitive conclusions, but these findings provide a basis for larger RCTs. Clinicians should consider MFR for preterm infants with enterostomy, particularly those with high stoma output, while maintaining vigilance for infection.