Necrotizing enterocolitis following spontaneous intestinal perforation in very low birth weight neonates

Necrotizing enterocolitis (NEC) and spontaneous intestinal perforation (SIP) are severe gastrointestinal complications of prematurity. The clinical presentation and treatment of NEC and SIP (peritoneal drain vs laparotomy) can overlap; however, the pathogenesis is distinct. Therefore, a patient initially treated for SIP can subsequently develop NEC. This phenomenon has only been described in case reports, and no risk factor evaluation exists. We evaluate clinical characteristics, risk factors, and outcomes of patients treated for a distinct episode of NEC after SIP.

We performed a retrospective review of very low birth weight (<1500 g) neonates who presented with pneumoperitoneum between 07/2004 and 09/2022. Data was obtained from two separate neonatal intensive care units that were part of the same institution. Patients with an initial preoperative, intraoperative, or pathological diagnosis of NEC were excluded. Patients with an intraoperative diagnosis of SIP or preoperative diagnosis of SIP successfully treated with a peritoneal drain (PD) were evaluated. Patients subsequently treated (medically or surgically) for NEC after SIP were then compared to SIP-alone patients. Clinical characteristics included demographics, gestational age (GA), birth weight (BW), perinatal risk factors (chorioamnionitis, steroids, indomethacin), postoperative feeding regimen, and length of stay (LOS) were compared.

Of the 278 patients included, 31 (11.2%) patients had NEC after SIP. There was no difference in GA (25 weeks vs 25 weeks, p = 0.933) or BW (760 g vs 735 g, p = 0.370) between NEC after SIP vs SIP alone cohorts, respectively. Twenty (64%) of NEC after-SIP patients were previously treated with LP. NEC after SIP occurred with a median onset of 56 days. Pneumatosis was the most frequent (81%) presenting symptom and 12 (39%) patients had hematochezia. Four (12.9%) patients required LP for NEC and all had NEC intraoperatively and on pathology. A majority (77.4%) of patients were on breast milk (BM) at time of NEC diagnosis. NEC after SIP patients had lower maternal age at delivery (29.0 vs 25.0, p = 0.055) and the incidence of NEC after LP (primary or failed drain) was higher than PD alone (16.7% vs 6.2%, p = 0.007). NEC after SIP patients had longer LOS (135 vs 81, p < 0.001).

We report an 11.2% incidence of NEC at a median of 56 days following successful treatment of SIP, resulting in increased LOS. SIP patients are a high-risk cohort and protocols to prevent this phenomenon should be investigated.