| Abstract|| |
Background: Neonatal pneumoperitoneum is attributed, in most of the reported cases, to necrotising enterocolitis (NEC). There are also other causes leading to free intraperitoneal air. The aim of this study is to describe the clinical spectrum, causes, management and outcome of neonates admitted with pneumoperitoneum in the paediatric surgery unit of a university hospital. Subjects and Methods: This retrospective study included neonates having radiographic evidences of pneumoperitoneum from 2012 to the end of 2014. Patients' files were analysed regarding age at admission, birth weight, history, clinical picture, management, operative findings and subsequent outcome. Results: Fifty-six out of 379 neonates (14.7%) were found to have pneumoperitoneum during the study period. There were 35 males (62.5%) and 21 females (37.5%). Cases diagnosed as NEC represented 27 neonates (48.2%). There were 29 cases (51.8%) with causes not related to NEC. Non-NEC causes were spontaneous intestinal perforation (8.9%), ano-rectal malformations (7.1%), Hirschsprung's disease (14.2%), ileal atresia (3.5%), incarcerated inguinal hernia (1.7%), gastric perforation (1.7%) and meconium ileus (1.7%). In seven patients (12.5%), those who were managed conservatively (8.9%) or whose laparotomy was negative (3.6%), no cause of pneumoperitoneum could be reached. Overall mortality was 25%, 78.5% of which was NEC-related. Conclusions: Neonatal pneumoperitoneum is an alerting finding for paediatric surgeons. Most cases imply serious causes with a significant morbidity and mortality, NEC being the most common cause. On the other hand, pneumoperitoneum is not an absolute indication for surgery. Careful assessment and tailored management can limit the morbidity of unnecessary laparotomies.
Keywords: Necrotising enterocolitis, neonate, pneumoperitoneum, spontaneous intestinal perforation
|How to cite this article:|
Morsi AH, Omar HR, Osama A, Khodary AR. Clinical spectrum of neonates presenting with pneumoperitoneum: A retrospective study. Afr J Paediatr Surg 2016;13:120-4
|How to cite this URL:|
Morsi AH, Omar HR, Osama A, Khodary AR. Clinical spectrum of neonates presenting with pneumoperitoneum: A retrospective study. Afr J Paediatr Surg [serial online] 2016 [cited 2020 Jan 22];13:120-4. Available from: http://www.afrjpaedsurg.org/text.asp?2016/13/3/120/187804
| Introduction|| |
Pneumoperitoneum is an alerting finding to physicians that is almost attributed to a perforation along the gastro-intestinal tract (GIT). In most of the published literatures, necrotising enterocolitis (NEC) accounts for the main proportion of neonates having a radiographic evidence of pneumoperitoneum. However, there are other causes implicated in neonatal pneumoperitoneum. All those diseases are associated with a significant morbidity and mortality. Surgical intervention based on the mere presence of pneumoperitoneum alone is still a matter of debate. On the other hand, pneumoperitoneum can be present without associated intestinal perforation at all.
| Subjects and Methods|| |
This is a retrospective study including all neonatal admissions having radiographic evidences of pneumoperitoneum from 2012 to the end of 2014. Patients' files during those 3 years were analysed regarding age at admission, birth weight, history, clinical picture, management, operative findings and subsequent outcome. All patients had complete blood count, C-reactive protein level, renal/liver function test and arterial blood gases. Resuscitation with intravenous fluids and parenteral antibiotics was done for all cases. Further management was tailored for each case according to the possible cause, general condition and individual circumstances.
| Results|| |
During the study period (2012-2014), the paediatric surgery unit at a university hospital received a total of 379 neonatal admissions. Out of the 379 total neonatal admissions, only 56 patients (14.7%) showed radiographic evidences of pneumoperitoneum.
There were 35 males (62.5%) and 21 females (37.5%). The mean age at presentation was 5.3 days. Most patients (87.5%) were pre-term. Very low birth weight patients accounted for 26.7% of the neonatal admissions with pneumoperitoneum. The mean birth weight was 2300 g in neonates having perforations not related to NEC. On the other hand, neonates with NEC-related perforations had a mean birth weight of 1200 g.
Neonates having perforations related to NEC were 27 (48.2%) in total. Bedside peritoneal drainage using a rubber drain was done in all the 27 patients as an initial management. In 14 patients, this was followed by noticeable improvement as indicated by improved general condition and tolerance to oral feeding.
The remaining 13 patients needed operative exploration because of deteriorating condition in spite of peritoneal drainage. Multiple perforations were found in nine patients, where intestinal diversion without resection was performed. Exteriorisation of an ileal perforation was done for each of the remaining four patients. Overall, 11 patients expired due to NEC-related perforations. Mortality rate for neonates with NEC-related perforations was 40.7%.
Only one male patient (1.7%), diagnosed as having respiratory distress, had a gastric perforation on the 2 nd day. He had nasal O 2 therapy, but no mechanical ventilation. There was a mild amount of turbid peritoneal fluid with a 2 cm long perforation on the anterior wall of the stomach close to the greater curvature. This was assumed to be caused by the nasogastric tube inserted on the 1 st day for feeding. The perforation was repaired in two layers with a transplyloric nasojejunal tube. By the 7 th day, this baby died of septicaemia.
There were four patients (7.1%) having neonatal pneumoperitoneum admitted with a diagnosis of ano-rectal malformations (ARMs). The mean age at presentation was 3 days. All the four patients had a high-type ARM with perforation at the sigmoid colon, though the thick meconium limited peritoneal contamination. Defunctioning sigmoid colostomy was performed for all. They had an uneventful post-operative course.
A 15-day-old patient (1.7%), with a history of forcible reduction of incarcerated inguinal hernia 2 days earlier, presented with bleeding per rectum. Abdominal radiogram showed pneumoperitoneum. Laparotomy revealed faecal peritonitis and an ileal perforation. Though bowel was looking viable, exteriorisation was done and the internal ring was closed.
Two neonates (3.5%) were diagnosed as ileal atresia with ruptured proximal pouch [Figure 1]. Mean age at admission was 3.5 days. Both of them had type II ileal atresia. Resection of the atretic segment with primary repair was done for both cases.
|Figure 1: Plain radiogram showing pneumoperitoneum due to a ruptured proximal ileal pouch in a neonate with ileal atresia|
Click here to view
Eight patients (14.2%) with intestinal perforations proved to have Hirschsprung's disease (HD) after exteriorisation and multiple biopsies. Total colonic agangliosis was proven in two (3.5%) neonates having multiple ileal perforations. Six (10.7%) patients with colonic perforations (four in transverse and two in sigmoid colon) proved to have HD.
Meconium ileus was found in one case (1.7%) with ascending colon perforation. A defunctioning stoma was done for this case.
As for the five neonates (8.9%) having intestinal perforations without features suggestive of NEC or other intestinal anomalies, spontaneous intestinal perforation (SIP) was the appropriate diagnosis. One out of five had an ileal perforation. The remaining four neonates had colonic perforations as follows; one patient with a SIP in transverse colon, one patient with a SIP of the transverse colon and two patients each with a SIP of the descending colon, one of which was already sealed by the time of laparotomy [Figure 2]. Intestinal diversion was done for all neonates with SIP except that with the sealed descending colon perforation where the site of perforation was secured with interrupted seromuscular suturing.
|Figure 2: Intra - operative view of a sealed perforation of the descending colon in a 22 - day - old neonate|
Click here to view
Conservative management was successful in five neonates (8.9%). The mean age at diagnosis was 16.5 days. Common features among those neonates were mean birth weight of 2500 g, sudden abdominal distension, good general condition, lack of signs of sepsis, passing bowel motions, tolerance to oral feeding and decreasing abdominal girth. Average hospital stay for this category was 7.3 days.
Laparotomy was negative for two patients (3.5%). No GIT perforation could be detected. Both neonates had mechanical ventilation. Both had peritoneal drainage prior to laparotomy with no improvement of the general condition. Careful review of radiograms showed pneumomediastinum as well. Both neonates expired.
Overall mortality rate was 25% (n = 14). NEC accounted for 78.5% (n = 11) of mortalities among neonates with pneumoperitoneum. [Table 1] shows a summary of the causes of neonatal pneumoperitoneum in this study.
| Discussion|| |
Neonatal pneumoperitoneum usually arises from a perforated abdominal visceral organ. In their retrospective analysis, Khan et al.  found that NEC remained the single most common cause of pneumoperitoneum in the new born. However, in around 50% of the patients, pneumoperitoneum was not related to NEC. Pneumoperitoneum can develop without intestinal perforation as well. , The present study showed nearly similar results where perforations related to NEC accounted for 27 cases (48.2%). Primary peritoneal drainage was done for the majority of these patients as an initial salvage. Surgical exploration was then offered to neonates who were not improving after the initial primary peritoneal drainage.
Meanwhile, 29 cases with pneumoperitoneum (51.8%) were not related to NEC. SIP, ARM, HD, ileal atresia, incarcerated inguinal hernia, gastric perforations and meconium ileus were the main causes of pneumoperitoneum which is not related to NEC in this study.
SIP without clinical or histological evidence of NEC is a distinct entity, especially in the low-birth weight neonates. SIP was laparotomy confirmed in five cases (8.9%) in this study, management of each differed according to the site and the condition of the perforation. Although the aetiology of SIP remains unknown, some authors argue that SIP and NEC represent different manifestations of the same pathogenic process. Factors related to increased susceptibility to SIP are pre-mature rupture of membranes, lower Apgar score and perinatal cardiovascular resuscitation. 
Bowel perforation occurs in 3.2-4.4% of patients with HD, and this happens most frequently during the neonatal period.  Perforations with HD are attributed to the increased intraluminal pressure rather than enterocolitis.  The colon was the site of perforation in 15 neonates in our study, six of which were histologically confirmed HD. Only two patients were found to have total colonic agangliosis with ileal perforations.
Isolated ileal atresia with perforation was found in two cases who presented with pneumoperitoneum and abdominal distension. Mean age at admission was 3.5 days. Both of them had type II ileal atresia with ruptured proximal pouch. Resection with primary anastomosis was done for both of them. Although perforations in the small bowel atresia have been reported by very few studies,  they can be seen as a result of antenatal accident or in certain cases due to delay in diagnosis. In this study, poor logistics contributed to delayed management.
Interestingly, a 15-day-old male neonate presented with bleeding per rectum after forcible reduction of incarcerated inguinal hernia by a non-medical personnel. He had pneumoperitoneum. Exploration confirmed an ileal perforation with faecal peritonitis. We searched the literature for similar case scenarios, but in vain.
Meconium ileus was a pre-disposing factor for a perforation of the ascending colon in one case (1.7%). This could be attributed to proximal bowel dilatation or enema during a trial of conservative management. Ein et al.  reported five cases of intestinal perforation after trials of non-operative treatment for meconium ileus.
Spontaneous neonatal gastric perforation is a rare neonatal event, occurring in 1:2900 live births, and its aetiology is controversial. Rosser et al.  and Leone and Krasna  reported a congenital muscular defect in the muscularis as a causative factor in neonatal spontaneous gastric perforation. Although the stomach is well vascularised, spontaneous perforations and necrosis do occur in neonates. The majority of defects are proximal linear tears along the greater curvature. We had only one case of gastric perforation in a 2-day-old male neonate with respiratory distress. This was assumed to be caused by the nasogastric tube inserted on the 1 st day for feeding rather than a spontaneous perforation. The perforation was repaired in two layers with a transpyloric nasojejunal tube.
Pneumoperitoneum may not always be an absolute indication for surgery in infants. Neonates with pneumoperitoneum require a proper clinical and the radiographic correlation to establish the aetiology of a perforation. A neonate with pneumoperitoneum, with otherwise normal abdominal examination, may deserve the trial of conservative management, thereby avoiding a laparotomy. There were five out of 56 patients with pneumoperitoneum managed conservatively. They showed improvement in general condition with subsequent tolerance to oral feeds and normal passage of stools. Later on, follow-up showed no abnormality at all.
Laparotomy confirmed SIP mortality is approximately half that of laparotomy confirmed NEC.  This relationship was observed across all birth weight categories. Overall mortality rate in this study was 25% whereas mortality rate for neonates with NEC-related perforations was 40.7%.
| Conclusions|| |
Neonatal pneumoperitoneum is an alerting finding for paediatric surgeons. Most cases imply serious causes with a significant morbidity and mortality, NEC being the most common cause. On the other hand, pneumoperitoneum is not an absolute indication for surgery. Careful assessment and tailored management can limit the morbidity of unnecessary laparotomies.
Financial support and sponsorship
We thank "Children action" for their support.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Khan TR, Rawat JD, Ahmed I, Rashid KA, Maletha M, Wakhlu A, et al.
Neonatal pneumoperitoneum: A critical appraisal of its causes and subsequent management from a developing country. Pediatr Surg Int 2009;25:1093-7.
Hill BJ, Saigal G, Patel S, Abdenour GE Jr. Transplacental passage of non-ionic contrast agents resulting in fetal bowel opacification: A mimic of pneumoperitoneum in the newborn. Pediatr Radiol 2007;37:396-8.
Goodman RA, Riley TR 3 rd
. Lactulose-induced pneumatosis intestinalis and pneumoperitoneum. Dig Dis Sci 2001;46:2549-53.
Khan RA, Mahajan JK, Rao KL. Spontaneous intestinal perforation in neonates: Is surgery always indicated? Afr J Paediatr Surg 2011;8:249-51.
Swenson O, Sherman JO, Fisher JH. Diagnosis of congenital megacolon: An analysis of 501 patients. J Pediatr Surg 1973;8:587-94.
Newman B, Nussbaum A, Kirkpatrick JA Jr. Bowel perforation in Hirschsprung's disease. AJR Am J Roentgenol 1987;148:1195-7.
Chiba T, Ohi R, Kamiyama T, Yoshida S. Ileal atresia with perforation in siblings. Eur J Pediatr Surg 1991;1:51-3.
Ein SH, Shandling B, Reilly BJ, Stephens CA. Bowel perforation with nonoperative treatment of meconium ileus. J Pediatr Surg 1987;22:146-7.
Rosser SB, Clark CH, Elechi EN. Spontaneous neonatal gastric perforation. J Pediatr Surg 1982;17:390-4.
Leone RJ Jr., Krasna IH. 'Spontaneous' neonatal gastric perforation: Is it really spontaneous? J Pediatr Surg 2000;35:1066-9.
Fisher JG, Jones BA, Gutierrez IM, Hull MA, Kang KH, Kenny M, et al.
Mortality associated with laparotomy-confirmed neonatal spontaneous intestinal perforation: A prospective 5-year multicenter analysis. J Pediatr Surg 2014;49:1215-9.
Ahmed Hosni Morsi
Department of Surgery, Paediatric Surgery Unit, Faculty of Medicine, Zagazig University Hospitals, Zagazig
Source of Support: None, Conflict of Interest: None
[Figure 1], [Figure 2]