Risk of Necrotizing Enterocolitis in Very-Low-Birth-Weight Infants with Isolated Atrial and Ventricular Septal Defects

Home , Cardiac shunt

ORIGINAL ARTICLE Risk of necrotizing enterocolitis in very-low-birth-weight infants with isolated atrial and ventricular septal defects

J Bain1,2, DK Benjamin Jr1,2, CP Hornik1,2, DK Benjamin3, R Clark4 and PB Smith1,2

OBJECTIVE: Necrotizing enterocolitis (NEC) is associated with a significant morbidity and mortality in premature infants. We sought to identify the frequency of NEC in very-low-birth-weight infants with isolated ventricular septal defects (VSDs) or atrial septal defects (ASDs) using a large multicenter database. STUDY DESIGN: We identified a cohort of infants with birth weight o1500 g cared for in 312 neonatal intensive care units (NICUs) managed by the Pediatrix Medical Group between 1997 and 2010. We examined the association between the presence of an ASD or a VSD with development of NEC using logistic regression to control for small-for-gestational age status, antenatal steroid use, antenatal antibiotic use, gestational age, sex, race, Apgar score at 5 min and method of delivery. RESULT: Of the 98 523 infants who met inclusion criteria, 1904 (1.9%) had an ASD, 1943 (2.0%) had a VSD and 146 (0.1%) had both. The incidence of NEC was 6.2% in infants without septal defects, 9.3% in those with an ASD, 7.8% in those with a VSD, and 10.3% in infants with both an ASD and a VSD. Compared with infants without septal defects, the adjusted odds ratios for developing NEC for each group—ASD alone, VSD alone and ASD with VSD—were 1.26 (95% confidence interval 1.07 to 1.49), 1.27 (1.07 to 1.51) and 1.79 (1.03 to 3.12), respectively. CONCLUSION: The presence of an ASD or a VSD was associated with NEC in this cohort of premature infants.

Journal of Perinatology (2014) 34, 319–321; doi:10.1038/jp.2013.174; published online 16 January 2014 Keywords: necrotizing enterocolitis; atrial septal defect; ventricular septal defect

INTRODUCTION METHODS Necrotizing enterocolitis (NEC) is a common and often fatal Study population complication of prematurity.1 Mortality associated with NEC Data were obtained from an administrative database that prospectively ranges from 15 to 30%.2–6 Survivors are at increased risk of captures information from daily progress notes generated by long-term morbidity, including short bowel syndrome, subsequent clinicians using a computer-assisted tool on all infants cared for intestinal strictures, neurodevelopmental impairment and growth by the Pediatrix Medical Group. Included in the data were diagnoses, delay.7 Physiologic states associated with the development of medications, gestational age, birth weight, method of delivery, Apgar score, sex, race and occurrence of NEC. The diagnosis of a septal defect NEC include: immature intestinal system, highly reactive inte- was made based on echocardiography results. We included all infants stinal immune system, altered microvascular environment admitted within the first 28 days of life with a birth weight of o1500 g including poor mesenteric blood flow and abnormal microbial who were discharged from 312 Pediatrix Medical Group neonatal intensive 8 colonization. care units (NICUs) in North America between 1997 and 2010. Exclusion Previous studies have identified term and near-term infants criteria were (1) death within the first 3 days of life, (2) treatment with with congenital heart disease as a risk factor for developing prostaglandins at any time before discharge, and (3) the presence of major NEC;9,10 however, they have not evaluated ventricular and atrial congenital anomalies (including infants with other structural heart septal defects (VSD and ASD) in isolation, but rather in association disease). with other congenital heart defects and in the surgical setting. VSD and ASD are the first and second most common congenital Definitions and statistical analysis heart defects, with a prevalence in the general population of For each infant, we identified the first episode of NEC occurring before 2.6/1000 and 1.6/1000, respectively.11 These defects may result in day of life (DOL) 120, including surgically or medically treated NEC. left-to-right shunting of blood, resulting in increased pulmonary We did not include those infants who had presumed or suspected NEC. blood flow, and a subsequent decrease in systemic, and thus We compared continuous and categorical variables between infants mesenteric blood flow. with a diagnosis of ASD, VSD, both ASD and VSD, or no cardiac septal Given the high prevalence of both VSD and ASD in infants, we defect. We used multivariable logistic regression models to evaluate the association between the presence of an ASD, a VSD, or both and the evaluated VSD and ASD as risk factors for NEC in premature infants diagnosis of medial or surgical NEC, controlling for the following using a large multicenter database. confounders present at birth: small-for-gestational age status, antenatal

1Duke University Medical Center, Duke Clinical Research Institute, Durham, NC, USA; 2Department of Pediatrics, Duke University Medical Center, Duke University, Durham, NC, USA; 3Clemson University, Clemson, SC, USA and 4Pediatrix-Obstetrix Center for Research and Education, Sunrise, FL, USA. Correspondence: Dr PB Smith, Department of Pediatrics, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715, USA. E-mail: [email protected] Received 1 July 2013; revised 31 October 2013; accepted 20 November 2013; published online 16 January 2014 Risk of NEC in infants with isolated ASD and VSD J Bain et al 320 steroid use, antenatal antibiotic use, gestational age, sex, race, Apgar RESULTS score at 5 min and method of delivery. We controlled for potential During the study period, 98 523 infants met our inclusion criteria site-specific effects using a fixed effect for site. Using a fixed effect model (Figure 1). Of these, 1904/98 523 (1.9%) had an ASD, 1943/98 523 that is conditioned on the NICU addresses the heterogeneity of baseline risk of outcomes in each individual NICU. We used Stata 12 (College (2.0%) had a VSD, and 146/98 523 (0.1%) had both an ASD and a Station, TX, USA) to analyze the data, and considered Po0.05 as VSD (Table 1). For infants with septal defect, the diagnosis was statistically significant. made on a median DOL 7 (interquartile range 3 to 26). The incidence of NEC in infants without cardiac septal defects was 5825/94 530 (6.2%); 3966/5825 (68.1%) of those episodes were medical NEC. Infants with an ASD had an incidence of 177/1904 (9.3%), 121/177 (68.4%) of which were medical NEC. Infants with a VSD had an incidence of 152/1943 (7.8%), and 107/152 (70.4%) of those episodes were medical NEC. Finally, infants with both an ASD and a VSD had an incidence of 15/146 (10.3%), 9/15 (60.0%) of which were medical NEC. The first episode of NEC occurred on DOL 21 (interquartile range 12 to 32), DOL 24 (14 to 37), DOL 22 (14 to 34) and DOL 18 (11 to 43) for infants without septal defects, infants with an ASD, infants with a VSD, and infants with an ASD and a VSD, respectively. Among infants who developed NEC, the incidence of surgical versus medical NEC did not differ between those with and without septal defects (P ¼ 0.75); it also did not differ across patients with different types of septal defects (P ¼ 0.69). Compared with infants without septal defects, the adjusted odds ratios for developing NEC for each group—ASD alone, VSD alone and ASD with VSD—were 1.26 (95% confidence interval 1.07 to 1.49), 1.27 (1.07 to 1.51) and 1.79 (1.03 to 3.12), respectively.

DISCUSSION In our multicenter study, we observed that infants o1500 g birth Figure 1. Study population. ASD, atrial septal defect; DOL, day of life; weight with an ASD or a VSD were at increased risk of developing NEC, necrotizing enterocolitis; VSD, ventricular septal defect. NEC as compared with infants without septal defects. To date, no

Table 1. Demographics

No cardiac septal defect, N ¼ 94 530 (%) ASD, N ¼ 1904 (%) VSD, N ¼ 1943 (%) ASD þ VSD, N ¼ 146 (%)

Gestational age, weeks p25 16 047 (17) 400 (21) 313 (16) 26 (18) 26–28 30 930 (33) 826 (43) 698 (36) 42 (29) 29–32 40 512 (43) 617 (32) 790 (41) 61 (42) 33–36 6721 (7) 52 (3) 135 (7) 17 (12) X37 291 (o1) 8 (o1) 7 (o1) 0 (0) Male 47 781 (51) 914 (48) 820 (42) 55 (38)

Birth weight, g o750 15 587 (16) 447 (23) 353 (18) 30 (21) 750–999 21 189 (22) 539 (28) 492 (25) 36 (25) 1000–1499 57 754 (61) 918 (48) 1098 (57) 80 (55) SGA 22 588 (24) 413 (22) 528 (27) 55 (38)

Apgar at 5 min 0–3 3327 (4) 84 (5) 80 (4) 9 (6) 4–6 14 031 (15) 330 (18) 329 (17) 26 (18) 7–10 75 368 (81) 1446 (78) 1493 (79) 107 (75)

Race/ethnicity White 43 994 (48) 904 (49) 997 (53) 79 (57) Black 24 243 (27) 437 (24) 394 (21) 25 (18) Hispanic 18 257 (20) 371 (20) 386 (21) 28 (21) Other 4841 (5) 134 (7) 98 (5.2) 6 (4) Cesarean section 67 356 (72) 1433 (76) 1453 (75) 112 (78) Antenatal steroid exposure 68 973 (73) 1412 (74) 1426 (73) 111 (76) Antenatal antibiotic exposure 38 304 (41) 776 (41) 810 (42) 54 (37) Abbreviations: ASD, atrial septal defect; SGA, small-for-gestational age; VSD, ventricular septal defect.

Journal of Perinatology (2014), 319 – 321 & 2014 Nature America, Inc. Risk of NEC in infants with isolated ASD and VSD J Bain et al 321 study has evaluated the incidence of infants with an isolated ASD CONCLUSION and/or VSD and the occurrence of NEC. In a previously conducted The association of isolated ASD and VSD lesions with NEC has not retrospective study using the National Inpatient Sample and the been previously studied. In this large multicenter analysis, we Kids’ Inpatient Database, infants with congenital heart disease found that infants with an ASD and/or a VSD had an increased were evaluated for the development of NEC after undergoing likelihood of developing NEC as compared with infants without 12 cardiac surgery. The investigators identified infants who deve- septal defects in our cohort group. loped NEC and who had a coexisting diagnosis of congenital heart disease. Of the 194 patients with congenital heart disease who developed NEC, 57 were diagnosed with secundum ASD, and 61 CONFLICT OF INTEREST were diagnosed with a VSD in isolation. This suggests that, of all Dr Benjamin receives support from the United States government for his work in infants who underwent surgery with heart disease and developed pediatric and neonatal clinical pharmacology (1R01HD057956-05, 1K24HD058735-05 12 NEC, B60% had either an ASD or a VSD. A study of infants with and NICHD contract HHSN275201000003I) and the non-profit organization Thrasher left-to-right intra-cardiac shunt, inclusive of infants with ASD, VSD, Research Fund for his work in neonatal candidiasis (www.thrasherresearch.org); he atrioventricular septal defect and persistent ductus arteriosus, also receives research support from industry for neonatal and pediatric drug demonstrated increased frequency of NEC.13 In our analysis, we development (www.dcri.duke.edu/research/coi.jsp). Dr Smith receives salary support excluded infants with a diagnosis of atrioventricular septal defect for research from the National Institutes of Health and the US Department of Health and those treated with prostaglandins. and Human Services (NICHD 1K23HD060040-01, DHHS-1R18AE000028-01 and The association of NEC with congenital heart disease in infants HHSN267200700051C); he also receives research support from industry for neonatal 14 and pediatric drug development (www.dcri.duke.edu/research/coi.jsp). The other has been well documented. The population most commonly authors declare no conflict of interest. studied consists of infants with heart lesions that would lead to lower systemic oxygen saturations including single ventricle physiology, transposition of the great arteries, truncus arteriosus REFERENCES and infants with a persistent ductus arteriosus.9 The development 1 Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011; 364: 255–264. of NEC in infants with congenital heart disease can be attributed 2 Guthrie SO, Gordon PV, Thomas V, Thorp JA, Peabody J, Clark RH. Necrotizing to many factors, including cardiac surgery and cardiopulmonary enterocolitis among neonates in the United States. J Perinatol 2003; 23: 278–285. bypass, a baseline elevation of circulating endotoxin, and pro- 3 Luig M, Lui K. Epidemiology of necrotizing enterocoliti—Part I: Changing regional 14 inflammatory cytokines. Additional pathophysiology is also trends in extremely preterm infants over 14 years. J Paediatr Child Health 2005; 41: thought to contribute to the development of NEC in these 169–173. infants. Previous studies have suggested that alteration in 4 Llanos AR, Moss ME, Pinzon MC, Dye T, Sinkin RA, Kendig JW. Epidemiology of mesenteric blood flow leads to intestinal ischemia, which results neonatal necrotising enterocolitis: a population-based study. Paediatr Perinat in intestinal injury, thereby creating an environment that Epidemiol 2002; 16: 342–349. predisposes the infant to NEC.15,16 Infants with congenital heart 5 Holman RC, Stoll BJ, Curns AT, Yorita KL, Steiner CA, Schonberger LB. Necrotising enterocolitis hospitalisations among neonates in the United States. Paediatr disease also remain at risk of NEC due to conventional risk factors Perinat Epidemiol 2006; 20: 498–506. of bacterial overgrowth with intestinal disruption. 6 Lin PW, Stoll BJ. Necrotising enterocolitis. Lancet 2006; 368: 1271–1283. Overall, in our population studied, the incidence of isolated ASD 7 Henry MC, Moss RL. Neonatal necrotizing enterocolitis. Semin Pediatr Surg 2008; was 1.9% and of isolated VSD was 2.0%. These values are higher 17: 98–109. than reported by previous investigators in a meta-analysis, in 8 Neu J, Mihatsch W. Recent developments in necrotizing enterocolitis. J Parenter which the median incidence of ASD and VSD was reported to be Enteral Nutr 2012; 36(1 Suppl): 30S–35S. 0.06 and 0.28%, respectively.17 The authors included literature 9 McElhinney DB, Hedrick HL, Bush DM, Pereira GR, Stafford PW, Gaynor JW et al. spanning as far back as 1955 and noted an increasing incidence of Necrotizing enterocolitis in neonates with congenital heart disease: risk factors ASD and VSD in the more recent publications.17 Technological and outcomes. Pediatrics 2000; 106: 1080–1087. 10 Bolisetty S, Lui K, Oei J, Wojtulewicz J. A regional study of underlying congenital advances, specifically the advancement of echocardiography, diseases in term neonates with necrotizing enterocolitis. Acta Paediatr 2000; 89: likely account for the higher incidence reported in more recent 1226–1230. publications. 11 van der Linde D, Konings EE, Slager MA, Witsenburg M, Helbing WA, Takkenberg Strengths of our study include its large multicenter cohort JJ et al. Birth prevalence of congenital heart disease worldwide: a systematic comprising a diverse NICU population of varying size, acuity and review and meta-analysis. J Am Coll Cardiol 2011; 58: 2241–2247. geography. Our analysis evaluated the risk of NEC based on 12 Mukherjee D, Zhang Y, Chang DC, Vricella LA, Brenner JI, Abdullah F. Outcomes variables available at birth. Unmeasured confounders that vary on analysis of necrotizing enterocolitis within 11 958 neonates undergoing cardiac daily basis limit our study, including markers of acuity of illness, surgical procedures. Arch Surg 2010; 145: 389–392. the presence and hemodynamic significance of a PDA, daily 13 Cheng W, Leung MP, Tam PK. Surgical intervention in necrotizing enterocolitis in neonates with symptomatic congenital heart disease. Pediatr Surg Int 1999; 15: amount and type of enteral nutrition, and drug exposure to 492–495. indocin or ibuprofen that may predispose infants to NEC. Many 14 Giannone PJ, Luce WA, Nankervis CA, Hoffman TM, Wold LE. Necrotizing infants classified as not having a septal defect did not have an enterocolitis in neonates with congenital heart disease. Life Sci 2008; 82(7-8): 341– echocardiogram; this may have led to misclassification of these 347. infants and ultimately biased our findings toward the null 15 Carlo WF, Kimball TR, Michelfelder EC, Border WL. Persistent diastolic flow reversal hypothesis. Ideally, to best characterize the associations between in abdominal aortic Doppler-flow profiles is associated with an increased risk of septal defects and NEC, all infants would be screened prospec- necrotizing enterocolitis in term infants with congenital heart disease. Pediatrics tively with an echocardiogram. In our analysis, the diagnosis of 2007; 119: 330–335. septal defect was a quantitative one; we did not have data on 16 Murdoch EM, Sinha AK, Shanmugalingam ST, Smith GC, Kempley ST. Doppler flow velocimetry in the superior mesenteric artery on the first day of life in preterm the quality of the septal defect to assess the patient-specific infants and the risk of neonatal necrotizing enterocolitis. Pediatrics 2006; 118: hemodynamics of each defect. In addition, we did not have access 1999–2003. to radiographs or surgical operative reports to confirm the 17 Hoffman JI, Kaplan S. The incidence of congenital heart disease. J Am Coll Cardiol diagnosis of NEC. 2002; 39: 1890–1900.

& 2014 Nature America, Inc. Journal of Perinatology (2014), 319 – 321