Mark Bishay

Hypercapnia and Acidosis During Open and Thoracoscopic Repair of Congenital Diaphragmatic Hernia and Esophageal Atresia Results of a Pilot Randomized Controlled Trial

Objective: We aimed to evaluate the effect of thoracoscopy in neonates on intraoperative arterial blood gases, compared with open surgery. Background: Congenital diaphragmatic hernia (CDH) and esophageal atresia with tracheoesophageal fistula (EA/TEF) can be repaired thoracoscopically, but this may cause hypercapnia and acidosis, which are potentially harmful. Methods: This was a pilot randomized controlled trial. The target number of 20 neonates (weight > 1.6 kg) were randomized to either open (5 CDH, 5 EA/TEF) or thoracoscopic (5 CDH, 5 EA/TEF) repair. Arterial blood gases were measured every 30 minutes intraoperatively, and compared by multilevel modeling, presented as mean and difference (95% confidence interval) from these predictions. Results: Overall, the intraoperative PaCO2 was 61 mm Hg in open and 83 mm Hg [difference 22 mm Hg (2 to 42); P = 0.036] in thoracoscopy and the pH was 7.24 in open and 7.13 [difference −0.11 (−0.20 to −0.01); P = 0.025] in thoracoscopy. The duration of hypercapnia and acidosis was longer in thoracoscopy compared with that in open. For patients with CDH, thoracoscopy was associated with a significant increase in intraoperative hypercapnia [open 68 mm Hg; thoracoscopy 96 mm Hg; difference 28 mm Hg (8 to 48); P = 0.008] and severe acidosis [open 7.21; thoracoscopy 7.08; difference −0.13 (−0.24 to −0.02); P = 0.018]. No significant difference in PaCO2, pH, or PaO2 was observed in patients undergoing thoracoscopic repair of EA/TEF. Conclusions: This pilot randomized controlled trial shows that thoracoscopic repair of CDH is associated with prolonged and severe intraoperative hypercapnia and acidosis, compared with open surgery. These findings do not support the use of thoracoscopy with CO2 insufflation and conventional ventilation for the repair of CDH, calling into question the safety of this practice. The effect of thoracoscopy on blood gases during repair of EA/TEF in neonates requires further evaluation. (ClinicalTrials.gov Identifier: NCT01467245)

Decreased cerebral oxygen saturation during thoracoscopic repair of congenital diaphragmatic hernia and esophageal atresia in infants

BACKGROUND/PURPOSE Congenital diaphragmatic hernia (CDH) and esophageal atresia with tracheoesophageal fistula (EA/TOF) can be repaired thoracoscopically, but this may cause hypercapnia, acidosis, and reduced cerebral oxygenation. We evaluated the effect of thoracoscopy in infants on cerebral oxygen saturation (cSO(2)), arterial blood gases, and carbon dioxide (CO(2)) absorption. METHODS Eight infants underwent thoracoscopy (6 CDH and 2 EA/TOF). Serial arterial blood gases were taken. Regional cSO(2) was measured using near-infrared spectroscopy. Absorption of insufflated CO(2) was calculated from exhaled (13)CO(2)/(12)CO(2) ratio measured by mass spectrometry. RESULTS CO(2) absorption increased during thoracoscopy with a maximum 29% ± 6% of exhaled CO(2) originating from the pneumothorax. Paco(2) increased from 9.4 ± 1.3 kPa at the start to 12.4 ± 1.0 intraoperatively and then decreased to 7.6 ± 1.2 kPa at end of operation. Arterial pH decreased from 7.19 ± 0.04 at the start to 7.05 ± 0.04 intraoperatively and then recovered to 7.28 ± 0.06 at end of operation. Cerebral hemoglobin oxygen saturation decreased from 87% ± 4% at the start to 75% ± 5% at end of operation. This had not recovered by 12 (74% ± 4%) or 24 hours (73% ± 3%) postoperatively. CONCLUSIONS This preliminary study suggests that thoracoscopic repair of CDH and EA/TOF may be associated with acidosis and decreased cSO(2). The effects of these phenomena on future brain development are unknown.

Morbidity after ganglioneuroma excision: is surgery necessary?

INTRODUCTION Ganglioneuroma (GN), the benign form of peripheral neuroblastic tumour, is often asymptomatic and the diagnosis can be incidental. Our aim was to evaluate the incidence of complications after surgical treatment following diagnosis of this tumour. MATERIAL AND METHODS 24 consecutive children were diagnosed with GN in our centre between January 1989 and December 2009. All patients had negative urinary catecholamines and/or biopsy confirming the diagnosis of GN. Data are reported as mean ± SD. RESULTS Age at diagnosis was 73 ± 43 months. The most common presentation was respiratory symptoms and/or abdominal pain; 9 (38%) patients were asymptomatic. Tumour location was in the chest (n=14), abdomen (n=7), or pelvis (n=3). 23 children (9 asymptomatic) were operated on; 1 child with a thoracic mass did not undergo surgery because of severe neurological impairments from birth unrelated to GN. 13 children (4 asymptomatic) had a thoracotomy, 8 children (4 asymptomatic) had laparotomy, and 2 (1 asymptomatic) underwent perineal resection. A macroscopically complete surgical excision was performed in 17 cases (74%) and a macroscopically near-complete excision in 6 (26%). At histological examination, resection margins contained tumour in 10 patients (43%) and were free of tumour in the remaining 13 (57%). 7 children (30%) had complications after surgery including 3 patients with Horners syndrome (which persisted in 2), 1 with chylothorax, 1 with pneumothorax, 1 with pain in the arm, and 1 who developed adhesive intestinal obstruction. 2 children received adjuvant chemotherapy. We re-evaluated the histology specimens according to the International Neuroblastoma Pathology Classification and found that the diagnosis of GN was confirmed in 20 cases (83%), while intermixed ganglioneuroblastoma (iGNB) was diagnosed in 4 patients (17%). At 33.5 ± 40 months (range 1-137) follow-up, all 24 patients, including the child not operated on and the children with incomplete resection or iGNB, are alive with no tumour progression or recurrence. CONCLUSIONS GN excision is associated with postoperative complications which can be persistent and may affect the quality of life of survivors. In our series we did not observe tumour progression in spite of incomplete excision. The rationale for GN excision should be revisited.

Intestinal failure-associated liver disease in surgical infants requiring long-term parenteral nutrition

PURPOSE Our aim was to determine incidence, severity, and outcome, as well as predisposing factors and underlying diagnoses, of intestinal failure-associated liver disease (IFALD) in surgical infants requiring long-term parenteral nutrition (PN). METHODS We retrospectively studied surgical infants receiving PN for at least 28 days for congenital or acquired intestinal anomalies over a 5-year period (January 2006 to December 2010). Intestinal failure-associated liver disease was defined as type 1 (early)--persistent elevation of alkaline phosphatase for 6 weeks or longer; type 2 (established)--additional elevated total bilirubin (≥ 50 μmol/L); and type 3 (late)--additional clinical signs of end-stage liver disease. RESULTS Eighty-seven infants required PN for at least 28 days. Intestinal failure-associated liver disease occurred in 29 infants (33%). Intestinal failure-associated liver disease was managed medically in all but 2 patients who underwent intestinal elongation. None were referred for intestinal or liver transplant. Intestinal failure-associated liver disease has been reversed in 17 (59%) of cases to date. Sixty-one children receiving long-term PN (70%) have achieved enteral autonomy, whereas 12 (14%) require home PN. Severity of IFALD was significantly associated with duration of PN and female sex. CONCLUSION Intestinal failure-associated liver disease remains a fairly common but rarely life-threatening complication of intestinal failure in surgical infants. Intestinal failure-associated liver disease can be reversed in more than half of these children, and enteral autonomy was achieved in more than two thirds, even with minimal use of intestinal elongation. This is the first study to demonstrate an association between the severity of IFALD in surgical infants and female sex.

Chlorhexidine antisepsis significantly reduces the incidence of sepsis and septicemia during parenteral nutrition in surgical infants.

BACKGROUND/PURPOSE After a change in national policy, central venous catheter (CVC) antisepsis with chlorhexidine was introduced in our hospital. Our aim was to evaluate whether this change reduced the rate of infection seen during parenteral nutrition (PN) in infants requiring gastrointestinal surgery. METHODS Two groups of consecutive infants were compared: control, 98 infants who had CVC antisepsis with 70% isopropanol alone, and chlorhexidine, 112 infants who had CVC antisepsis with 2% chlorhexidine in 70% isopropanol. Incidence rates of sepsis (blood cultures taken) and septicemia (blood cultures positive) were compared by Poisson regression. RESULTS Seventy-one percent of infants experienced clinically suspected sepsis. The incidence of septicemia was 32%. The incidence rate ratio for sepsis was 0.72 (95% confidence interval, 0.61-0.84) for the chlorhexidine group vs control (P < .0005). The incidence rate ratio for septicemia was 0.49 (95% confidence interval, 0.36-0.67; P < .0005); that is, over a given period of PN, patients had half the rate of positive blood cultures after the introduction of chlorhexidine antisepsis compared with before. CONCLUSION (1) The incidence of sepsis and septicemia among surgical infants on PN for gastrointestinal anomalies is high. (2) Chlorhexidine CVC antisepsis has significantly reduced this incidence, and we advocate its use in this group of patients.

Septicaemia due to enteric organisms is a later event in surgical infants requiring parenteral nutrition.

INTRODUCTION The purpose of this study was to determine whether, in surgical infants requiring parenteral nutrition (PN), septicaemia due to enterococci or Gram-negative bacilli occurs later than septicaemia due to coagulase-negative staphylococci (CNS). PATIENTS/MATERIAL AND METHODS We retrospectively studied 112 consecutive surgical infants (corrected gestational age up to 3 months) receiving PN for at least 5 days for congenital or acquired intestinal anomalies over a 2-year period (July 2007-June 2009). Data collected included diagnosis, duration of PN, episodes of septicaemia (defined as growth of bacteria from blood culture), and organisms cultured. We compared the time to first occurrence of septicaemia due to CNS with the times to first occurrence of septicaemia due to enterococci, Gram-negative bacilli, or other micro-organisms, using Kruskal-Wallis nonparametric ANOVA test and Dunns multiple comparisons test. Data are given as median (range). RESULTS 31 patients (28%) had a total of 65 episodes of septicaemia. Septicaemia due to CNS was most common, occurring in 22% of patients, after 17 days (1-239) of PN. Septicaemia due to enteric organisms was less common and occurred significantly later, at 59 (24-103) days for enterococci (p<0.01), and at 55 (30-106) days for Gram-negative bacilli (p<0.05). CONCLUSIONS Septicaemia due to enterococci or Gram-negative bacilli occurs later in the course of PN than septicaemia due to CNS, in surgical infants. This suggests that these infants become more vulnerable to the translocation of enteric micro-organisms after a longer period of parenteral nutrition.

Brain lipid–binding protein: a marker of differentiation in neuroblastic tumors

PURPOSE Neuroblastoma (NB), ganglioneuroblastoma (GNB), and ganglioneuroma (GN) are neuroblastic tumours (NT) of sympathetic nervous system origin. Brain lipid-binding protein (BLBP) has potential morphogenic activity during nervous system development but has not been studied in these tumours. We analyzed the expression of BLBP in NT according to histological subtypes and extent of differentiation. METHODS Thirty cases of NT (10 each of NB, intermixed GNB, and GN) were identified from the histopathology archive of a single center. Tissue sections were obtained from representative paraffin blocks and immunohistochemistry for BLBP performed. RESULTS Brain lipid-binding protein was not expressed in any NB case. In all cases of GN, BLBP was strongly expressed in the cytoplasm of mature ganglion cells but negative in Schwannian stroma. In the intermixed GNB, there was similar strong BLBP immunoreactivity in the cytoplasm of fully differentiated and differentiating ganglion cells but no BLBP expression in immature neuroblasts. CONCLUSION Brain lipid-binding protein is strongly expressed in mature and maturing ganglion cells in NT (GN and GNB), whereas it is absent in poorly differentiated neuroblasts of GNB and NB. Cytoplasmic expression of BLBP in NT increases as the cells undergo neural differentiation and is therefore associated with the extent of tumour differentiation and favorable histology.