Journal of Pediatric Surgery 49 (2014) 508–513

Contents lists available at ScienceDirect

Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg

Prognostic value of abdominal sonography in necrotizing enterocolitis of premature infants born before 33 weeks gestational age☆ Aurélie Garbi-Goutel a,⁎, Véronique Brévaut-Malaty a, Michel Panuel b, Fabrice Michel c, Thierry Merrot d, Catherine Gire e a Service de médecine néonatale, unité de soins intensifs néonatale, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, Université de la Méditerranée, Chemin des Bourrellys, 13015 Marseille Cedex 20, France b Département d’imagerie médicale, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, Université de la Méditerranée, Chemin des Bourrellys, 13015 Marseille Cedex 20, France c Service d’anesthésie et de réanimation pédiatrique, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, Université de la Méditerranée, Marseille, France d Service de chirurgie pédiatrique, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, Université de la Méditerranée, Chemin des Bourrellys, 13015 Marseille Cedex 20, France e Service de médecine néonatale, unité de soins intensifs néonatale, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, Université de la Méditerranée, Chemin des Bourrellys, 13015 Marseille Cedex 20, France

a r t i c l e

i n f o

Article history: Received 15 April 2013 Received in revised form 1 October 2013 Accepted 12 November 2013 Key words: Necrotizing enterocolitis Sonography Preterm infant Prognosis

a b s t r a c t Objective: The purpose of this study was to assess the prognostic value of abdominal sonography in necrotizing enterocolitis (NEC) in preterm infants with a gestational age less than 33 weeks of gestation, using surgery and/or death as the primary outcome and stenosis as the secondary outcome. Methods: A retrospective study of 95 premature infants (mean gestational age: 28.6 weeks), presenting with NEC between January 2009 and November 2011 and who underwent plain abdominal radiography and sonography, was performed. In uni- and multivariate analyses, radiographic and sonographic findings were correlated with complications (‘surgery and/or death’ and ‘stenosis’). Results: Sonographic findings of free intraperitoneal air (odd ratio [OR] = 8.0; IC, 1.4–44.2), free abdominal fluid (OR 3.5; IC 1.3–9.4), portal venous gas (OR 3.9; IC, 1.2–12.9), and bowel wall thickening (OR 2.8; IC,1.1–7.2) were significantly associated with surgery and/or death. Intramural gas was significantly correlated (OR = 11.8; IC, 1.5–95.8) with intestinal stenosis following NEC. None of the radiographic findings were associated with complications. Conclusion: Abdominal sonography is a reliable tool for the prognostic assessment of NEC in preterm infants. © 2014 Elsevier Inc. All rights reserved.

Necrotizing enterocolitis (NEC) is a leading cause of morbidity in neonatal intensive care units (NICU). The rate of confirmed NEC varies between 5% and 10% in infants weighing less than 1500 g, with an associated mortality rate of 20%–30%, and up to 50% when surgery is necessary [1–4]. Diagnosis of NEC is based on the presence of nonspecific gastrointestinal clinical signs (feeding intolerance, abdominal distension, and bloody stools) in conjunction with at least one pathognomonic sign on plain abdominal radiography (intramural gas or portal venous gas) [2]. Bell’s classification of NEC, which was modified by Walsh and Kliegman in 1986, relates to the worsening of these clinical and radiological abdominal signs [5,6]. These staging criteria help pediatricians in the diagnosis, management, and prognostic assessment of NEC, but they are exclusively based on plain abdominal radiography, which remains the “gold standard” for the diagnostic assessment [2]. Nevertheless, the adverse prognostic value of certain radiological signs is currently debatable [7–11].

The usefulness of abdominal sonography in the diagnosis of NEC, known since 1984 [12], was only studied in small case series and heterogeneous gestational age populations [12–20]. This imaging modality allows for an earlier detection of typical NEC signs [17,19– 21], with more rapid disease management. However, few studies have investigated the correlation between NEC sonographic signs and patient outcomes [13]. The main objective of our study was to assess the prognostic value of abdominal ultrasound in premature infants with NEC, who were born before 33 weeks of gestational age. The primary outcome was surgery and/or death, and the secondary outcome occurrence of postNEC stenosis. 1. Patients and methods This was a single-center, retrospective study conducted in a Level 3 university maternity unit. 1.1. Inclusion and exclusion criteria

☆ Conflict of interest: We have no conflict of interest in relation to this study. ⁎ Corresponding author. Tel.: +33 4 91 96 87 50; fax: +33 4 91 96 46 75. E-mail address: [email protected] (A. Garbi-Goutel). 0022-3468/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2013.11.057

Inclusion criteria: premature infant of less than 33 weeks gestational age, born between January 1, 2009 and November 30, 2011, and

A. Garbi-Goutel et al. / Journal of Pediatric Surgery 49 (2014) 508–513

hospitalized in the NICU; confirmed gestational age using early ultrasonographic dating; absence of congenital malformation; clinical symptoms suggestive of NEC, as defined by the presence of abdominal distension, increased gastric residuals (N20% of enteral feeding volume), or blood in the stool (macroscopically or microscopically); confirmation of the NEC episode using abdominal imaging examinations, including one abdominal radiograph and ultrasonography, performed as soon as NEC was suspected. Exclusion criteria: absence of abdominal radiographic or ultrasonographic images for the NEC episode. Two groups were defined based on the outcome: a group for infants who have survived without surgery, and a group for infants who have needed surgery or have died. 1.2. Imaging examinations Plain abdominal radiographs were performed according to the following protocol: at the bedside, single anteroposterior view, patient in supine position, vertical X-ray beam, with a portable X-ray system Mobilett XP Siemens and Agfa CRHD 5.0 Directrix Microspot cassette, exposition data 60 kV, 1.25 mA. Abdominal ultrasound examinations were carried out at the bedside, by a radiologist, with mobile ultrasound units and high-frequency linear ultrasound transducers (Philips HD11, 5–12 MHz probe; Sonosite Mturbo, 6–13 MHz probe). Ultrasound examinations were performed according to the following protocol: quadrant by quadrant, swipescanning in transverse and sagittal planes, and spectral Doppler of the portal vein. All the operators were trained to the same protocol by a fellowship trained pediatric radiologist. First imaging examinations during the NEC episode were retrospectively and simultaneously read by a fellowship trained pediatric radiologist, with a 30-year experience in neonate ultrasounds, who had not performed the ultrasound examination, and a neonatologist. The physicians were blinded to the clinical and biological patient data. For each subject, they read first the radiograph, and then the ultrasound examination. Abdominal images were reviewed using the hospital’s Picture Archiving and Communication System (PACS, Centricity General Electric Medical System) on Barco Coronis 3MP displays. The following items were assessed on plain abdominal radiographs: abdominal distension, free peritoneal air, intramural gas, and portal venous gas. Abdominal sonography review was standardized, taking into account the following Epelman et al. features [22]: free peritoneal air, abdominal fluid, portal venous gas, portal vein Doppler examination (for detecting flow artifacts in relation with gas micro-bubbles), parietal pneumatosis, and parietal thickness. 1.3. Clinical data and management Antenatal, perinatal, and postnatal data were collected from the NICU’s computerized database. Collected antenatal data: antenatal corticosteroid therapy, intrauterine growth retardation (estimated fetal weight b 10th percentile for gestational age) [23], oligohydramnios, fetal heart rhythm abnormalities. Collected perinatal data: gestational age, gender, Apgar score, birth weight, and small for gestational age (birth weight b 5th percentile according to the Audipog curves [24]). Postnatal data: surfactant administration, respiratory distress syndrome, severe bronchopulmonary dysplasia (ventilatory support at 36 weeks of corrected gestational age), patent ductus arteriosus after 7 days of life, data regarding enteral feeding (minimal enteral feeding, type of milk, and volume of enteral feeding at the time of the NEC episode), and umbilical vein catheterization. The following data pertaining to the NEC episode were collected: clinical symptoms of NEC and signs of digestive intolerance; biological signs, including inflammatory syndrome (leucocytes N20 Giga/L or C-reactive protein [CRP] N10 mg/L), maximum CRP level, respiratory or metabolic acidosis (pH ≤ 7.1 and pCO2 greater than 48 mmHg or HCO3− less than 22 mmol/L); imaging data (radiological and ultrasonographic

509

signs); complications, such as death, surgery, and post-NEC stenosis (diagnosed via contrast study), occurring during the 6 first months of life. The management of children with NEC symptoms was standardized as follows: plain abdominal radiography, abdominal sonography, bowel rest with nasogastric tube, broad-spectrum antibiotic therapy, close clinical and ultrasonographic monitoring (abdominal radiography wasn’t repeated), minimal enteral feeding using breast milk (10 to 20 mL/kg/day during 10 days) after ultrasonographic normal control. Indications for surgery included lack of improvement or clinical deterioration despite optimal medical treatment, as well as pneumoperitoneum (with instability or severe alteration in clinical status) and post-NEC stenosis. Regarding stenosis during the first 6 months post NEC, the contrast study wasn’t systematic but only performed in front of occlusive syndrome in an old premature baby who presented with NEC during hospitalization. 1.4. Statistical analysis The primary outcome was death and/or need for surgery, and the secondary outcome intestinal stenosis. The groups “with complications” (surgery or death) and “without complications” (survival without surgery) were compared, as were the groups “with stenosis” and “without stenosis”. Two software versions were used to perform statistical analyses: SPSS 17.0 and Stata 8.0. Quantitative variables were presented as mean ± standard deviation (SD), and qualitative variables as number (n) and percentage (%) of subjects. The differences were analyzed using the Chi-squared test or Fisher’s exact test for variables below five, and Student’s t test for continuous variables. Odds ratios (OR) and their 95% confidence intervals (CI) were calculated using simple logistic regression. A multivariate analysis was conducted using multiple logistic regression including the following variables: acidosis, inflammatory syndrome, portal venous gas on sonography, bowel wall thickening on sonography, peritoneal fluid or free air on sonography, and gestational age for the outcome “death and/or surgery”; inflammatory syndrome, intramural gas on sonography, bowel wall thickening on sonography, and gestational age for the outcome “stenosis.” The threshold for statistical significance was set at p b 0.05. 2. Results 2.1. Description of the study population In total, 442 premature infants of less than 33 weeks gestational age were born between January 1, 2009 and November 30, 2011, of whom 123 (27.8%) presented with NEC symptoms (Fig. 1). For the 28 infants excluded from analysis, intrauterine growth retardation was more common (46.4% vs. 20.7%, p = 0.007) and birth weight lower (1037 ± 299 vs. 1175 ± 309, p = 0.038) than for the 95 infants included in the analysis. As to NEC stages, they were less than or equal to stage 1B of Bell’s classification. All of them survived without surgery. 2.2. Description of the included infants The antenatal, perinatal, and postnatal characteristics of the 95 patients included in the study and of the entire NEC population are detailed in Table 1. The mean age at the first NEC episode was 21.9 ± 15.5 days. Overall, 33 patients (34.8%) exhibited intramural gas on abdominal radiographs, with eight (8.4%) displaying no pneumatosis on sonography. Conversely, 62 neonates (65.2%) exhibited no intramural gas on abdominal radiographs, with 26 (27.3%) displaying pneumatosis on ultrasound. Fig. 2 illustrates the sonographic findings of one patient.

510

A. Garbi-Goutel et al. / Journal of Pediatric Surgery 49 (2014) 508–513

442 neonates

Prognostic value of abdominal sonography in necrotizing enterocolitis of premature infants born before 33 weeks gestational age.

The purpose of this study was to assess the prognostic value of abdominal sonography in necrotizing enterocolitis (NEC) in preterm infants with a gest...
599KB Sizes 0 Downloads 3 Views