Respiratory Syncytial Virus Bronchiolitis Complicated by Necrotizing Enterocolitis: A Case Series

Two patients with RSV bronchiolitis and NEC prompted us to query our quaternary care children’s electronic hospital record for patients of age <24 months who were admitted to the PICU with community acquired RSV bronchiolitis and diagnosed with NEC shortly after admission. A total of 821 term and late-preterm infants with RSV bronchiolitis, and an additional 652 with non-RSV bronchiolitis, were admitted from January 2010 to April 2020.

Demographics, clinical information, interventions, and outcomes were collected and tabulated using a standardized form. In each case, a suspicion of NEC prompted radiographic and laboratory investigations. Abdominal roentgenography (KUB) was acquired in all subjects initially, followed by abdominal ultrasound and/or computed tomography (CT) when patients demonstrated progressive abdominal distention. Diagnoses of NEC were confirmed dually by the interpretations of radiologists as well as consultation with pediatric surgery.

The 4 cases we identified have their clinical characteristics summarized in Table 1. Patients were born at 34 to 39 weeks’ gestation, and their ages ranged from 3 to 6 postnatal weeks. All were previously healthy and without medical problems; more specifically, they did not have cardiac disease, chronic lung disease, or immune dysfunction. Infants presented with symptoms of upper respiratory tract infection 1 to 5 days before hospital admission. RSV was detected by reverse transcriptase polymerase chain reaction of nasopharyngeal specimens in all patients; 2 specimens were collected at the referring hospitals where RSV was not subtyped. Complete viral panels were not performed; 3 patients did not have coinfection with influenza A or B. Three infants were intubated before PICU admission. One arrived on noninvasive ventilation but required intubation within 8 hours.

Continuous nasogastric feeds were initiated within 2 days of admission, and the maximum rates achieved ranged from 30 to 160 mL/kg per day. The sequence of symptoms is depicted in Fig 1. These infants did not experience periods of hypotension or protracted desaturation, and they did not require vasopressors or blood products before their NEC diagnoses. Infants 2 and 4 eventually required vasopressors and blood products for management of septic shock and multiorgan failure.

On the day of evaluation for NEC, all subjects had significant thrombocytopenia and hypoalbuminemia. C-reactive protein was markedly elevated in the 2 subjects in whom this laboratory was obtained. All patients had normal lactic acid levels at the time of NEC diagnosis.

Infant 1 had paucity of bowel gas on initial KUB, whereas infants 2, 3, and 4 demonstrated pneumatosis and overt portal venous air. Clinical concerns for ascites were confirmed by abdominal ultrasound in infants 1, 2, and 4.

Pediatric surgery consultations were obtained in all cases, and all patients were managed with broad-spectrum antibiotics, bowel rest and parental nutrition. The clinical course of each case was as follows:

• Infant 1 was initially managed with rectal irrigation for 5 days because of a nonconcerning KUB, poor stool output, and suspicion for Hirschsprung disease. Subsequently he developed complex ascites, followed by pneumatosis, free air, and loculated abdominal abscesses. His stage IIIB NEC required partial colectomy and diverting ileostomy, which was later reversed after recovery.

• Infant 2 was initially treated for 1 day with rectal irrigation for suspected septic ileus. Although her stage IIIA NEC was treated nonoperatively, her course was complicated by complex ascites, septic shock, and acute respiratory distress syndrome (ARDS) on hospital day 4 that required veno arterial extracorporeal membrane oxygenation for 4 days. A peritoneal drain placed at the time of cannulation recovered simple ascites. Later, on the general inpatient ward, she developed multiple strictures, which necessitated delayed small bowel resection with anastomosis.

• Infant 3 was treated nonoperatively with empirical antibiotics and bowel rest for stage IIB NEC, which resolved without complications.

• Infant 4’s stage IIIA NEC was managed with antibiotics and bowel rest alone. His clinical course was complicated by multiorgan failure and ARDS, requiring high frequency oscillatory ventilation; therefore, he was deemed too unstable for exploratory laparotomy. A peritoneal drain placed to improve respiratory mechanics evacuated simple ascites.

• Infants 1 to 3 were discharged from the hospital tolerating feeds. Infant 4 did not survive to PICU discharge.

To our knowledge, this is the first report in which the development of NEC in the context of RSV bronchiolitis and acute respiratory failure is described. NEC rarely occurs in infants born after 35 weeks’ gestation^6,8 ; when it does, most have associated risk factors such as intrauterine growth retardation, sepsis, blood transfusions, gestational diabetes, or maternal drug use.⁶  The infants in our case series lacked traditional risk factors for NEC. Their intubations were recorded as uneventful, and they had no early signs of systemic toxicity or abdominal pathology. Before developing clinical manifestations of NEC, no hypoxemic or hypotensive episodes were documented, and no pressors or blood products were administered.

The one infant managed medically was fed human milk, which has been associated with a decreased incidence and severity of NEC in premature infants.^6,9,10  Feeds were initiated early after intubation and in most cases were advanced to goal. Although there are no guidelines, such feeding advancement is not atypical for patients with RSV bronchiolitis and respiratory failure who lack signs of toxicity or multiorgan failure. Nevertheless, a contribution of early enteral feeding cannot be excluded.

These cases have other clinical commonalities that may suggest a pathophysiologic mechanism and highlight potential lessons for clinicians. Whereas most healthy infants tolerate RSV illness well, ours were admitted to the PICU in respiratory failure on hospital day 1. Furthermore, the peak levels of mechanical ventilatory support before NEC were high.

In a hypothetical framework for the pathogenesis of RSV-induced NEC, 4 features of severe RSV disease may be relevant: (1) a neonate’s immature immune system is less able to control viral replication when presented with a high viral load^11,12 ; (2) immune dysregulation favoring production of proinflammatory cytokines and chemokines are associated with severe lower respiratory tract infections^12,13 ; (3) RSV RNA has been identified in peripheral blood, and extrapulmonary manifestations may indicate systemic dissemination of RSV during severe disease⁵ ; (4) although our patients did not exhibit signs, overt sepsis due to RSV has been described.¹⁴ 

Gastroenteritis, diarrhea, hyperbilirubinemia, and GI hemorrhage have all been reported in association with RSV infection,³  whereas NEC has not. In healthy children, GI dysmotility may precede pneumatosis intestinalis, and portal venous gas is correlated with poor outcomes.¹⁵  NEC seems to evolve from interactions among intestinal hypoperfusion, hypoxia, GI dysmotility, and inflammation.^10,16,17  Respiratory failure and mechanical ventilation have been associated with all of these,¹⁸  and a gut–lung axis has repeatedly been demonstrated in humans.^19,20  Furthermore, RSV itself is known to increase the production of proinflammatory cytokines and chemokines.^12,13  As such, an analogy may be drawn with meconium aspiration as a predisposing factor for NEC, in which proinflammatory cytokines of pulmonary origin are thought to compromise protective barriers in the intestine by altering intestinal blood flow⁷  and decreasing expression of tight junctions, resulting in intestinal hyperpermeability and bacterial translocation.¹⁶  Our cases were neonates, raising the possibility that GI tract immaturity was also contributory.

A particularly virulent strain of RSV seems unlikely, given that our 4 cases occurred across 3 nonconsecutive RSV seasons. We asked whether these NEC events might have been “spontaneous” or related to other viral etiologies of bronchiolitis. To our surprise, we were unable to find cases of NEC among the 652 PICU admissions with non-RSV bronchiolitis over the same 10-year span. This makes it more likely that the NEC events described are phenomenologically associated with severe RSV disease. We cannot, however, exclude later diagnoses associated with neonatal NEC such as cystic fibrosis.⁸ 

Our cases appear to have followed a patterned progression: acute respiratory failure due to RSV infection, abdominal distention after initiation of enteral feeds, identification of diagnostic signs of NEC, and the development of simple or complex ascites. We do not believe this report provides reason for altering enteral feeding practices in infants with RSV. However, it may offer guidance for calibrating caution and clinical suspicion. For instance, neonatal age and severe RSV disease, even in the absence of sepsis, may represent a context in which enteral feeds are advanced with caution, and early signs of intolerance are considered a red flag. Sometimes there can be uncertainty about whether KUB findings indicate medical NEC or septic ileus, but with our small series, we suggest ascites may reflect the development of NEC.

Severe RSV disease in neonates may represent a risk factor for NEC, even in the absence of traditional risk factors such as prematurity, hypoxemic or hypotensive events, and blood transfusions. A high clinical suspicion for NEC is appropriate in such patients who developed acute abdominal distention after the initiation of enteral feeds. The development of ascites may favor the diagnosis of NEC.

We are indebted to our patients and their families for their support and consent. Our gratitude also extends to Kim Giles (Clinical Decision Support, Le Bonheur Children’s Hospital) and our VPS coordinators, Charlene Summerall and Karen Brinkley, for their assistance with queries.

ARDS

acute respiratory distress syndrome

CT

computed tomography

GI

gastrointestinal

KUB

abdominal roentgenography

NEC

necrotizing enterocolitis

RSV

respiratory syncytial virus