Necrotizing enterocolitis following intensive phototherapy in full-term newborns – is there a possible association?

Introduction

The American Academy of Pediatrics (AAP) Subcommittee on Hyperbilirubinemia issued clinical practice guidelines on the management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation [1]. These guidelines that are widely accepted worldwide, include a recommendation to provide intensive phototherapy with high levels of irradiance in the 430- to 490-nm band (usually up to 30 μW/cm² per nm or even higher) delivered to as much of the infant’s surface area for the treatment of significant neonatal hyperbilirubinemia.

We present three cases of full-term newborns who were treated by intensive phototherapy, and developed necrotizing enterocolitis (NEC), that is currently not considered as a complication of phototherapy, and raise the possibility of an association between them. The hospital’s Ethics Committee waived the need for institutional review board (IRB) approval or parental consent, because the following descriptions do not include any identifying details of the infants discussed.

Presentation of the cases

Discussion

NEC is currently not considered a possible complication of phototherapy. The appearance of cases of newborns that were treated by intensive phototherapy and developed NEC that is not frequent in full-term babies, raised the question whether there may be an association between NEC and treatment with intensive phototherapy.

All three cases had isoimmunization with positive direct Coombs test causing significant jaundice with rapidly rising bilirubin levels mandating intensive phototherapy. The only two factors constant among these three cases are isoimmunization with rapid increase in bilirubin levels and the use of intensive phototherapy. In the following paragraph we will try to suggest an association between NEC and intensive phototherapy, while trying to address why the other etiologic possibilities related to isoimmune hyperbilirubinemia and other treatments given seem less likely, although not excluded.

NEC was previously described in association to exchange transfusion, but not phototherapy. Yet it was shown that postprandial increase in perfusion of the gut does not occur or is reduced under phototherapy [2]. The presumptive explanation is that intensive phototherapy causes marked vasodilataion in the skin, resulting in decreased perfusion of the intestine, ischemia, and NEC. Possibly, increased enteral feeds consumed by many newborns under intensive phototherapy, that by itself already predisposes the gut to relative ischemia, create imbalance between supply and demand of perfusion to the gut, which might lead to NEC. Another possible explanation is related to a study that showed that abdominal exposure to phototherapy had been associated with increased neonatal intestinal apoptosis in an animal model [3]. Neutrophil function is also altered by phototherapy [4], and it is well established that neutrophils play a major role in the pathogenesis of NEC. Another possible explanation may be the marked decrease in bilirubin, which is a known antioxidant, because of the intensive photoisomerisation that increases the risk for oxygen-radical-mediated diseases, such as NEC. The role of the hemolytic disease as a possible contributing factor to NEC in these babies cannot be excluded, although previously described only in Rh incompatibility [5], and without a good pathophysiological mechanism to explain this association. The infants in the cases presented did not have anemia secondary to the hemolysis before the appearance of NEC, thus a possible explanation relating anemia to decreased tissue oxygenation leading to NEC is less likely here. The use of IVIG in severe isoimmune hemolytic jaundice caused by Rh or ABO incompatibility in infants who did not sufficiently respond to phototherapy was also associated with increased incidence of NEC, although this might be applicable to case 1 only [6–8]. Although it may be argued that transfusion played a role in the etiology of NEC in case 1, this is highly unlikely, because he was not transfused with RBCs before he developed NEC. He also did not have bloody stools or eosinophilia that is related to NEC from this etiology [9]. It may be argued that case 3 had predisposition to NEC because of perinatal depression; however, he had no other signs consistent with such a possibility. Only case 3 also had bloody stools but the rest of his clinical presentation did not resemble atopic enteropathy. None of our infants had eosinophilia, which also disputes against this differential diagnosis [9].

The three cases hereby presented occurred over 6 years period. To our estimate this translates to a very low incidence of <0.1% of the term newborn infants undergoing intensive phototherapy for severe early hemolytic neonatal jaundice. Although rare, this occurrence of NEC in full-term newborn infants in relation to intensive phototherapy merits some caution and mainly clinical awareness to such possible complication of phototherapy.

Following the principles of Sir Bradford Hill on the inference of causality [10], we can say that although we have shown relationship in time (temporality), possible coherence of evidence and analogy, and have suggested an explanation with biological plausibility, the strength of the association between NEC and intensive phototherapy is relatively low, specificity is hard to demonstrate and we have no data or experimental proof to support consistency or biological gradient (dose-response relationship).

Although we cannot establish any causative relationship, the close temporal relationship between intensive phototherapy and the clinical presentation of NEC along with the suggested causative mechanism related to decreased perfusion to the gut under phototherapy makes this association very plausible. We hereby call for research studies on the possible effects of phototherapy on the vascular bed in the gut and other vital organs that may interfere with perfusion, thus having clinical implications.