Article Text

Download PDFPDF
Pneumatosis coli in preterm neonates: Can they be managed more conservatively to maintain the intestinal milieu?
  1. Elis Yuexian Lee1,
  2. Eric Ma1,
  3. Amudha Jayanthi Anand1,2,3,4 and
  4. Suresh Chandran1,2,3,4
  1. 1Department fo Neonatology, KK Women's and Children's Hospital, Singapore
  2. 2Paediatric Academic Clinical Programme, Duke NUS Medical School, Singapore
  3. 3Paediatric Academic Clinical Programme, Yong Loo Lin School of Medicine, Singapore
  4. 4Paediatric Academic Clinical Programme, Lee Kong Chian School of Medicine, Singapore
  1. Correspondence to Professor Suresh Chandran, Department of Neonatology, KK Women's and Children's Hospital, Singapore, Singapore; profschandran2019{at}gmail.com

Abstract

Necrotising enterocolitis (NEC) is a severe gastrointestinal disease mostly in premature infants due to intestinal necrosis. The aetiology of NEC is multifactorial and includes gut immaturity, intestinal dysbiosis and exaggerated intestinal mucosal reactivity to microbial ligands. Radiographic evidence of pneumatosis intestinalis has been a critical feature for diagnosing NEC Bell stage ≥IIA and recommended treatment includes prolonged antibiotics (7–14 days) while off enteral feeds. Pneumatosis coli (Pcoli), a mild or benign form of NEC, is characterised by pneumatosis limited to the colon in an infant having haematochezia, negative septic screening and no systemic signs. We report two healthy preterm infants with haematochezia and colonic pneumatosis while on breast milk feeds. The sepsis screen was negative. A brief period of antibiotics and gut rest led to the spontaneous resolution of haematochezia and colonic pneumatosis, facilitating early enteral feeds. This case report emphasises the need to differentiate NEC from benign Pcoli.

  • GI bleeding
  • Neonatal health
  • Ultrasonography

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Background

Necrotising enterocolitis (NEC) remains the most common gastrointestinal surgical emergency in preterm infants and still carries high mortality and long-term morbidity.1 Classic symptoms of early NEC in a preterm infant are feed intolerance, abdominal distension, bloody stool and a diagnostic confirmation with radiographic evidence of pneumatosis intestinalis (PI).2 The classical Bell staging of 1978 has been the most clinically used criteria to define NEC.3 Initially, Bell staging identified babies with NEC into three stages, and later Walsh and Kliegman revised the staging from 3 to 6 (Modified Bell staging) to guide therapeutic decisions.4 PI is a consistent radiological finding in NEC stage II and above. Even though NEC most commonly affects the terminal ileum, caecum and proximal colon, Bell staging does not specify the pneumatosis location for diagnosis.3 4 However, PI occurs in other neonatal conditions, including pneumatosis coli (Pcoli), intestinal obstruction, ischaemic intestinal necrosis and milk protein allergy, and these conditions to differentiate clinically from NEC are challenging.5

Pcoli is an insufficiently acknowledged benign medical condition and is often treated as NEC with prolonged antibiotics and gut rest, which can severely impact the intestinal microbiome. We present two growing preterm infants with visible blood in the stools (VBS) and colonic pneumatosis, which had a spontaneous recovery in less than a week, elucidating the benign nature of this condition.

Case presentation

Case 1

A female child was born as a second twin to a mother at 32+6 weeks gestation by caesarean section following incomplete antenatal steroid. Twin 2 was small-for-gestational-age with a birth weight of 1050 g (<10th centile), length of 42 cm (>50th centile) and head circumference of 29 cm (25–50th centile). She was born vigorous with Apgar scores of 8 and 9 at 1 min and 5 min of life, respectively. She required initially continuous positive airway pressure (CPAP) support for respiratory distress and was weaned off nasal cannula oxygen at 34 weeks. Her echocardiography confirmed a structurally normal heart at birth and normal pulmonary pressures at discharge.

Enteral feeding with breast milk was initiated on day 1 of life, with the gradual escalation of feeds. Probiotics were commenced with the first feed. Breast milk was fortified with human milk fortifier when on feeds >80 mL/kg/day. She was on full enteral feeds by day 7. However, on day 46 of life, she presented with three episodes of VBS. There was no feed intolerance, the abdominal examination was unremarkable, and the baby remained haemodynamically stable.

Case 2

A male infant was born at 29 weeks of gestation to a primipara mother with a birth weight of 1210 g (50th centile), head circumference of 25 cm (10–25th centile) and length of 42 cm (>90th centile). The antenatal screening was unremarkable, and the mother had two doses of antenatal steroids. Apgar scores were 9 at 1 min and 5 min of life. The baby required intubation, ventilation and surfactant following admission to the neonatal intensive care unit. His respiratory support was weaned to CPAP support at 7 hours and depronged to room air by 48 hours of life. Umbilical lines were inserted. In addition to parenteral nutrition, trophic feeding with pasteurised donor human milk was initiated on day 1 and was on full breast milk feeds by day 4 of life. He was on probiotics, and the breast milk was fortified. On day 22 of life, the infant presented with multiple episodes of bloody mucoid stools. The infant remained haemodynamically stable with a clinically benign abdomen. No anal fissure or perianal excoriation was found.

Investigations

Case1

Serial haematological investigations were unremarkable (table 1). Abdominal X-rays revealed a heterogeneous pattern of air lucencies over the right abdomen consistent with pneumatosis (figure 1A). An abdominal ultrasound scan (US) showed intramural gas in bilateral upper quadrant large bowel loops and portal venous gas (PVG) (video 1). There was no thickening of bowel walls, dilated fluid-filled loops or ascites. Doppler studies were unremarkable. Microbiological investigations of blood and stool were negative (table 1). A repeat abdominal X-ray 24 hours later showed significant resolution of pneumatosis.

Table 1

Demographic and clinical characteristics of infants with pneumatosis coli

Figure 1

(A) Abdominal X-ray shows the presence of bubbly lucencies in the right upper quadrant of the abdomen consistent with intramural gas (arrow). (B) A repeat abdominal X-ray done 4 days later shows resolution of the intramural gas. The bowel gas pattern is within normal limits.

Video 1 Video showing movement of an air bubble through the portal venous system in the liver.

Case 2

Haematological parameters were normal. Blood and stool cultures were sterile. Abdominal X-rays showed PI along the large colon (figure 2A). US abdomen confirmed colonic pneumatosis (figure 2B), with no PVG, gut wall thickening or ascites and normal Doppler studies (table 1).

Figure 2

(A) This abdominal radiograph shows the presence of multiple cystic lucencies lining the large bowel (arrows) consistent with pneumatosis intestinalis. (B) An US of the abdomen done on the same day shows the presence of gas within the wall of the large bowel in the right upper quadrant of the abdomen corresponding to the pneumatosis intestinalis seen on the abdominal radiograph. US, ultrasound scan.

Differential diagnosis

In the presence of VBS and PI, the most important differential diagnosis for Pcoli are NEC, sepsis and milk protein allergy. NEC and sepsis have significant haematological and microbiological markers to confirm the diagnosis.1 3 4 Milk protein allergy is characterised by VBS, abdominal distention and irritability with systemic eosinophilia and PI. Most growing preterm infants are exposed to cow’s milk protein in breast milk fortifiers or may have sensitised in-utero from cow’s milk ingested by the mother, which explains the onset of the symptoms early in life. There are no diagnostic tests available; symptoms resolve with gut rest and can remain well on elemental formula.6

Treatment

Case 1

The infant was kept nil-by-mouth (NBM) for bowel rest and initiated with cloxacillin, amikacin and metronidazole. Antibiotics were discontinued after 48 hours as both blood and stool cultures were sterile. The abdominal US done 48 hours later showed a marked reduction in intramural gas and resolution of PVG. Subsequent bowel output was normal, with no recurrence of haematochezia. Serial abdominal X-rays showed resolution of PI by day 4 (figure 1B). Maternal diet was eliminated from cows’ milk and milk products since the onset of VBG. Enteral feeds were restarted on day 5 with freshly expressed breast milk, with uneventful escalation to full feeds.

Case 2

He was kept NBM to give him gut rest. Antibiotics were initiated but discontinued in 48 hours with a negative blood culture report. Repeat abdominal X-ray, 17 hours later, showed resolution of PI (figure 3) and was confirmed by the abdominal US. Breast milk feeds were initiated on day 4 of illness with no further occurrence of haematochezia for the rest of the hospital stay. The infant was discharged home on full enteral feeds at term.

Figure 3

The pneumatosis intestinalis has resolved. The bowel gas pattern is unremarkable.

The clinical events of these two infants are depicted in figure 4.

Figure 4

Timeline depicting the events from birth to discharge from the hospital. PI, pneumatosis intestinalis; PVG, portal venous gas; US, ultrasound scan.

Outcome and follow-up

Both infants remained clinically well and in our hospital’s preterm follow-up programme.

Discussion

The preterm infants described here had VBS and PI/PVG in imaging, consistent with ‘definite NEC stage IIA’ as per modified Bell’s staging if abdominal signs/symptoms were present. Clinical improvement with normal stools in 48 hours of gut rest and spontaneous resolution of PI and PVG enabled successful initiation of feeds by day 5 and escalation of enteral nutrition. In the absence of abdominal distension, feed intolerance, lethargy, systemic derangement, negative blood and stool cultures and unremarkable haematological parameters, Pcoli was diagnosed.

Bright red blood from the rectum is included as an intestinal sign in Bell stage IIA for diagnosis of NEC.3 Sharma et al reported that the probability of having a bloody stool in <30-week gestational age infants with NEC is an uncommon finding compared with older infants.7 In the background of this observation, Gordon et al questioned the validity of bloody stools in Bell stage IIA as a diagnostic criterion in extreme preterm infants. Authors clarified that with the presence of ileus in modern preterm NEC, where the necrosis predominantly involves the distal ileum, passage of bloody stool through the colon is unlikely.8 This might explain why VBS is a consistent feature in Pcoli.

Several limitations have surfaced in recent times with the use of Bell staging for diagnosis of NEC. PI refers to the radiological finding of gas in the gastrointestinal wall.9 The radiolucencies seen in NEC are rounded when submucosal and linear when subserosal gas is located.10 PI has been incorporated as one of the radiographic signs in the modified Bell’s stage ≥IIA to facilitate diagnosis and management of NEC.4 The presence of PI in NEC varies from 19% to 98%.11 Bell staging confirms NEC in a symptomatic infant with abdominal symptoms/signs in the presence of VBS and PI.4 PVG, an extension of intramural gas into the venous system, is another radiological marker seen in stage ≥IIB but reported only in 30% of cases and is no more considered an ominous sign in NEC.10 12 Ascertainment of PI and PVG in abdominal X-rays is critical. However, substantial interobserver and intraobserver variability for individual radiologic signs in abdominal X-rays were observed by Rehan et al. The authors concluded that isolated radiologic signs should not be relied onto diagnose NEC.13 USs of the abdomen may be helpful to diagnose PI in the hands of an experienced sonologist,14 but systematic reviews reported low sensitivity.15

The incidence of Pcoli may be understated as it is not widely recognised. Pcoli is a benign gastrointestinal disease characterised by radiographic evidence of multiple gaseous cysts within the colonic wall in a clinically stable infant with VBS and negative laboratory markers of inflammation and infection. Clinical manifestations in neonates may sometimes resemble those seen in the milder spectrum of NEC.16 Keller et al classified PI according to its location in the intestine. They observed that more distal colonic involvement had less severe inflammatory changes than small intestine and caecum, explaining the milder systemic signs in Pcoli.17 This ensures that infants developing PI in the distal colon are more benign.

In 1975, Richmond and Mikity first reported seven infants having colonic pneumatosis, presenting with VBS, minimal or absent systemic signs and an uneventful clinical course as a benign form of NEC.18 In 1976, Leonidas and Hall described seven patients with the characteristic findings of Pcoli. The authors concluded that Pcoli should be recognised as a benign form of NEC and has a favourable prognosis.16 Since then, sporadic reports of Pcoli came out in the literature. Hoehn et al reported two cases of Pcoli, of which one even underwent a negative laparotomy, and the recovery was uneventful.19 Cordero et al reviewed the records of 65 newborn infants with VBS and PI. They found Pcoli in 32 of them, and the rest were NEC. Pcoli group had VBS and pneumatosis limited to colon and rectum and spontaneous recovery. In the NEC group, diffuse PI was noted in 50%, and 30% required laparotomy. Interestingly, the Pcoli group was mostly late preterm (63%) or term infants (28%) and needed less intensive care.20 Our reported cases were at 29 and 32 weeks at birth and developed symptoms at 32 and 39 weeks corrected age.

The pathophysiology of Pcoli and its benign clinical course remains enigmatic. Several theories have been proposed to explain the pathogenesis of Pcoli. In PI, the intramural air bubbles represent hydrogen gas produced by the bacteria in the colonic lumen from the fermentation of ingested carbohydrates.21 Excess production of hydrogen gas leads to local invasion of the intestinal wall, especially in the presence of colitis, increased intraluminal pressure due to obstruction or higher hydrogen tensions exceeding nitrogen tension in the blood, resulting in a hydrogen diffusion gradient towards the submucosal layer.22 A severe form of NEC in extreme preterm infants could be explained by the exaggerated immune reaction of immature human enterocytes to pro-inflammatory cytokines, causing marked local inflammation and progressive tissue damage.23 Cordero et al proposed that a higher incidence of Pcoli in late-preterm/term infants in their cohort might be due to an attenuated immune response in mature infants that limits local tissue damage and systemic involvement, allowing intestinal restitution, explaining the benign course of Pcoli.20

Decreased mesenteric blood flow has been described as a contributor to the pathogenesis of NEC. In piglets with normal birth weight on occlusion of mesenteric vessels, there was 40% less blood flow in the distal compared with the proximal ileum. In low-birth-weight piglets, the blood flow was less by 55% and could mount a compensatory collateral blood flow only in the proximal ileum, increasing the vulnerability of the distal ileum for NEC.24 Another theory is related to the susceptibility of the colon to ischaemia. Griffiths point (splenic flexure) and Sudek’s point (rectosigmoid junction) are highly prone to ischaemia as they fall in the watershed areas, the regions in the colon in between the two major arteries. The marginal artery of Drummond supplies the watershed areas, but in 50% of humans, this artery is poorly developed, and 70% of ischaemic colitis develops in this area.25 26 Normally, the oxygen extraction of the intestinal tissues from the mesenteric arteries is low. Intestinal injury follows low perfusion or reperfusion of the intestinal walls.27 Isolated colonic involvement in Pcoli may be due to autoregulation differences in the local areas.28

In healthy infants, the intestinal milieu nurtures the gut microbiome. Early introduction of breast milk in neonates promotes the establishment of gut microbiota, which improves intestinal barrier function, reduces proneness for gut mucosal inflammation and enhances the enteric immune system. Perturbation of the gut microbiome with an extended period of antibiotics and prolonged gut rest, depriving the intestine of breast milk immunomodulating components like immunoglobulin A and oligosaccharides (prebiotics), leads to gut dysbiosis, increasing pathogenic bacterial invasion.29 30 Treatment of infants with Pcoli involves primarily gut rest for 3 to 5 days. Antibiotics initiated on suspicion of NEC in these infants can be discontinued when haematological parameters are normal, and cultures are negative. Keeping antibiotics longer in preterm infants with Pcoli can derange the gut microbial balance resulting in an abundance of Enterococcus species.31

In conclusion, modified Bell staging remains a good tool for managing NEC. But if well infants beyond a few weeks of life develop VBS and colonic PI, initiate full workup for NEC, keep them NBM and start antibiotics. If workup is negative, the baby remains clinically well, and pneumatosis resolves, discontinue antibiotics and initiate breast milk feeds and upgrade under supervision. A timely and precise diagnosis of Pcoli can prevent unwarranted use of prolonged antibiotics and central line-related sepsis and facilitate an early resumption of enteral feeds, revitalising the gut microbiome.

Patient’s perspective

Case 1: We were very worried when we heard about necrotising enterocolitis and were praying. Our little girl was well all the time, even when there was blood in the stool. Thankfully our doctors gave excellent care and could feed in 5 days.

Case 2: Our son started passing blood in the stool and was worried about the outcome as we were told about necrotising enterocolitis in our counselling session on the day of delivery. We are grateful to our doctors for taking measures to settle her tummy so fast and feeding in 4 days.

Learning points

  • Pneumatosis coli is a rare condition and should be considered in the differential diagnosis of NEC in preterm infants with bloody stool, colonic pneumatosis and unremarkable systemic signs with no evidence of infection.

  • The pathogenesis of pneumatosis coli is likely multifactorial. Regional ischaemia in the ‘watershed zone’ of the colon is suspected to be the trigger for pneumatosis coli.

  • The prognosis of pneumatosis coli is good, with the spontaneous recovery of symptoms and resolution of radiographic features in less than a week following conservative management.

  • Timely diagnosis of pneumatosis coli is essential to prevent the use of prolonged antibiotics and fasting, both deleterious to the establishment of a healthy intestinal microbiome.

Ethics statements

Patient consent for publication

Acknowledgments

We would like to thank A/Prof Teo Eu-leong Harvey James, Department of Diagnostic Imaging, for providing the images/videos and legends for this case report. We appreciate the efforts of Ms Keerthi Karunakaran, Year 4, Medical Student, University College Cork, School of Medicine, Ireland, for preparing figure 4.

References

Footnotes

  • EYL and EM are joint first authors.

  • AJA and SC are joint senior authors.

  • Contributors EYL, EM: manuscript preparation, reviewed literature and added the references. AJA, SC: written the discussion and edited the final manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.