Spontaneous gastric perforation: is it really common?

Neonatal gastric perforation (NGP) is a rare, serious, and life-threatening condition which needs early recognition with prompt intervention for better prognosis. Its etiology is not yet well established, but multiple speculations have been put forth for its etiopathogenesis. Few most considered are traumatic, spontaneous, or secondary to inflammatory process like necrotizing enterocolitis. This study describes the etiological and prognostic factors for patients with NGP in our experience at a single center. A total of 46 neonates were included. Male gender predominated with M:F being 1.7:1. Most (n=36) neonates presented within 10 days of life whereas 8 neonates presented within 15 days. At presentation, gas under diaphragm was the most common radiologic finding in 25 (54.3%) neonates. On exploratory laparotomy, it was found that greater curvature was involved in maximum number of cases (n=27), followed by lesser curvature and anterior and posterior walls of the stomach. Most of perforations were within 1–5 cm in size. Gastrorrhaphy was done in all cases, and in two cases, feeding jejunostomy was done along with repair for feeding purpose. Finally, spontaneous NGP was diagnosed in 30 (60.8%), and NEC patches on other parts of the intestine were seen in 11 patients. Postoperatively, 28 neonates developed complications in the form of sepsis (n=13), wound infection (n=10), and burst abdomen (n=5). Regarding clinical outcome, 27 (58.7%) were discharged from the hospital whereas 19 (39.3%) patients died. Our results show that spontaneous NGP is most commonly associated with NEC in our population, usually affecting the greater curvature. We observed a high mortality rate; however, good ICU care may improve the survival.


Background
The first case of gastric perforation in neonates was reported by Siebold in 1825 whereas first survival of gastric perforation repair was reported in 1950 [1]. Among perforations of the intestine in neonates, the neonatal gastric perforation (NGP) has been reported in 7% with poor prognosis and significant mortality [2,3]. Prematurity, asphyxia neonatorum, birth stress, aggressive respiratory resuscitation at birth, anatomic pathologies causing gastric outlet obstruction, and few associated congenital anomalies have been reported to be the most important factors causing NGP in the literature [4][5][6][7][8].
The association of NGP with factors like male gender, metabolic acidosis, prematurity, or low birth weight claimed even worse prognosis [9,10].
Many recent studies tried to elucidate prognostic factors, but no definite correlations were concluded, so prognostic factors remain as yet controversial. The common identifiable sign is abdominal distension that is mostly associated with shock. Early detection of such cases is important as it was found to improve mortality and morbidity of these neonates. In this study of single tertiary care unit conducted over 3 year's period, we describe our experience of clinical findings, management, and outcome in neonates with NGP at a single center.

Methods
Formal institutional ethical approval was sought for this prospective study. During a 3-year period from June 2015 to June 2018, a total of 570 neonates (within 30 days of life) presented in neonatal emergency with diagnosis of pneumoperitoneum. We included 46 patients with NGP into this study which reflects 8% of total intestinal perforations. Demographic details like gender, age of presentation, gestational age, birth weight, mode of delivery, presence of neonatal asphyxia, main signs and symptoms, gastric tube insertion and mechanical ventilation before the onset of gastric perforation, associated perinatal risk factors, associated congenital anomalies involving both intestinal and extra intestinal regions like congenital heart disease, preoperative radiologic findings, surgical procedures performed, site of perforation, size of perforation, clinical outcome, and surgical complications were collected on a predesigned pro forma. Suspected patients were excluded who did not undergo exploration. All the recorded data was analyzed by SPSS 26 (version 233 South Wacker Drive, 11th Floor Chicago, IL 60606-6307 USA).
Qualitative variable was presented as mean SD, and for quantitative variables, we calculated as percentages and frequency.

Results
A total of 46 neonates were included. Male gender predominated with M:F being 1.7:1. Most of them were full term (69.6%). Twenty patients were having weight less than 2.5 kg, and 5 had < 2 kg weight. About 60.9% (n= 28) were delivered by caesarian section. The history of neonatal asphyxia prior to presentation was found in only 10 (21.7%) patients while none had history of mechanical ventilation, but 11 (23.9%) remained admitted in intensive care unit. Nasogastric tube was attempted in 14 (30.4%) neonates before admission. Three neonates (6.5%) had associated congenital heart disease, two had duodenal atresia, two with respiratory distress, and eight neonates had associated perinatal factors. Feeding was already started in 37 neonates, and 43.5% (n=20) were given formula feed; however, no feeding was started in 9 neonates. Most (n=38) neonates presented within 10 days of life whereas 8 neonates came within 15 days. At presentation, gas under diaphragm was the most common radiologic finding in 25 (54.3%) neonates, out of which 6 patients had typical food ball sign (Fig. 1). The demographic and clinical details are summarized in Table 1.
An early management for optimization of neonates includes either laparostomy tube placement followed by laparotomy or if already optimized then directly underwent for laparotomy. On exploratory laparotomy, it was found that greater curvature was involved in most cases (n=27), followed by lesser curvature and anterior and posterior walls of the stomach. Most perforations were 1-5 cm in size. Gastrorrhaphy was done in all cases, and in two cases, feeding jejunostomy was done along with repair for feeding purpose. Finally spontaneous NGP was diagnosed in 30 (60.8%) whereas no obvious cause elucidated on exploration, thus revived diagnosis label NGP. Due to associated NEC patches on other parts of the intestine, 11 neonates were labelled having NECassociated NGP. Three had findings similar to ischemic type and two cases to traumatic NGP. Details are given in Table 2.
Postoperatively, a total of 28 neonates developed complications in the form of sepsis (n=13), wound infection  Table 3.  They reported NGP being four times commoner in males than females [6,11]. We found that 69.6% neonates were full term and 30.4% were born between 33 and 36 weeks of gestation whereas a large number of neonates (n=22) had weight less than 2.5. Similar percentage was noted by Kara et al. in neonates with gastric perforation [6]. Byun et al. also reported that NGP is common in full-term neonate [12]. According to our results, most neonates with NGP were just presented within 10 days of life (78%), and only one patient presented late after 20 days. This finding was compared with Kara et al. and Byun et al. [6,12] Leone and Krasna reported that 4 out of 7 neonates had pre-existing cause while 3 had no identifiable cause of NGP [7]. In literature, the etiology of NGP is still debatable though many suggested causes like congenital weakness or absence of muscle [13], high gastric acidity leading to ulceration [14], trauma to gastric wall [15], neonatal asphyxia causing intestinal ischemia [16], raised intragastric pressure caused by swallowing of air or distal obstruction [17,18], and lack of intestinal pacemaker cells [19] or C-KIT + mast cells [20]. Usually, the mixture of causes is proposed as far as etiology of NGP is concerned [21], and the most common cause was found to be the raised intragastric pressure causing rupture of the stomach wall usually in a dilated stomach and uncoordinated vomiting due to poor neurologic control [8,18]. We reported spontaneous NGP in 30 neonates as no obvious and clear cause was found; associated NEC patches were seen in 11 neonates, three had findings suggestive of ischemic insult, and two cases were due traumatic injury. One hypothesis is that spontaneous perforations are due to the congenital defects in the muscular wall of the stomach [22]. In our study, 10 neonates had a history of neonatal asphyxia at birth and neonatal ICU admission which we suspect may be the cause of intestinal ischemia and perforation. In 14 cases, there was history of pre-admission NG placement, but only two cases of NGP found per-operatively to have features of traumatic type. Three of our cases had associated congenital heart diseases and had per-operative findings of ischemic variety of perforation. Two cases having duodenal atresia lead to raise intra-gastric. Regarding clinical outcome, 27 (58.7%) patients were discharged after recovery; 17 (39.3%) neonates died including five cases who left against medical advice. Despite all advances in neonatal pre-and post-operative care throughout literature, mortality rate associated with NGP is still very high. Leone and Krasna reported 70% mortality rate [7]. However, in a recent study by Yang et al. [19], mortality rate decreased with improvement in neonatal care; they divided study neonates into four groups, i.e., 1980-1989, 1990-1999, 2000-2009, and 2010-2016, and reported mortality for the four groups was (5/5) 100%, (12/24) 50%, (6/19)

Conclusion
Our results show that spontaneous NGP is most commonly associated with NEC in our population, usually affecting the greater curvature. We observed a high mortality rate; however, good ICU care may improve the survival.