Surgical approach to neonatal intestinal perforation. An analysis on 85 cases (1991-2001)

AIM: Primary gastrointestinal perforations have an incidence of between 1% and 3% in NICU patients. The 3 Centers participating in this study cover nearly 40% of the NICU population of the Lazio Region--Italy. The aim of this study is to discuss factors affecting survival in patients affected by a primary intestinal perforation. METHODS: From 1991 to 2001, 67 cases of 85 with a neonatal gastrointestinal perforation, were related to primary bowel lesions. Necrotizing enterocolitis (NEC) was not always the cause of perforation and in many patients an isolated bowel lesion without signs of NEC was found. The aim of this study was to examine clinical and intraoperative findings of NEC and non NEC perforations and their impact on survival. A relevant number of these patients were extremely low-birth weight (ELBW). Controversies about treatment of this category of neonates are discussed. RESULTS: Patients were 37 males and 30 females (mean birth weight 1 274.8 g, mean gestational age 28.9 weeks, mean age at perforation 10 days). Overall survival was 56.8%. Patients were divided by intraoperative findings in 2 groups: NEC (n=48), or isolated intestinal perforation (IIP) without signs of NEC (n=19). Differences between these 2 groups with regard to birth weight, maturity, associated cardiac anomalies (patent ductus arteriosus, PDA) were significant. NEC and IIP behaved as 2 distinct entities, each with peculiar clinical (age at perforation, oral feeding, need of ventilatory support) and radiological aspects. At surgery, multiple lesion on necrotic bowel were typical of NEC versus single, isolated perforations on healthy bowel typical of IIP. Overall survival was almost identical in the 2 groups (59% vs 58%). ELBW patients (55% of the total neonatal intestinal perforations) were also studied. There were 21 patients with NEC and 16 with IIP. The 2 groups were different in age at perforation, previous oral feeding and associated cardiac anomalies (PDA). Overall survival was 62% for NEC and 50% for IIP. A laparotomy was always performed. Temporary peritoneal drainage was done in 4 cases only. Results were better when intestinal diversion was performed rather than resection and primary anastomosis. Almost all NEC patients had multiple perforations and extended bowel necrosis. CONCLUSION: NEC is the most frequent cause of neonatal intestinal perforation. This is a quite distinct entity from IIP, which must always be differentiated preoperatively and which is most frequently found among low birth weight newborns. As far as surgical treatment of perforation among ELBW neonates is concerned, peritoneal drainage might be reasonably performed when a single lesion on healthy bowel as in IIP is clearly diagnosed but it could be inadequate for NEC patients.