RESUMO
PURPOSE: Very low birth weight (VLBW) infants are at a risk of spontaneous focal intestinal perforation (FIP). Treatment includes supportive care, antibiotics, and drainage with/without surgery. Broad-spectrum antibiotic agents like carbapenems are applied frequently, although their use is not well-supported by the limited evidence of causal pathogens. We hypothesize that the use of carbapenems may not be necessary in VLBW infants with FIP. Our primary objective was to evaluate the antimicrobial use in VLBW infants with FIP in a cohort of the German Neonatal Network (GNN). The secondary objective was to characterize a subset in detail as a benchmark for future targets of stewardship. METHODS: Data on VLBW infants with FIP was collected prospectively within the GNN, a collaboration of 68 neonatal intensive care units (NICU). With regards to the primary objective, patient characteristics and antimicrobial treatment were extracted from the predefined GNN database. To address our secondary objective, an additional on-site assessment of laboratory and microbiological culture results were performed. RESULTS: In the GNN cohort, 613/21,646 enrolled infants (2.8%) developed FIP requiring surgery. They were frequently treated with carbapenems (500/613 (81.6%)) and vancomycin (497/613 (81.1%)). In a subset of 124 VLBW infants, 77 (72.6%) had proof of gram-positive bacteria in the abdominal cavity, coagulase-negative staphylococci (CoNS) predominantly. Despite the low prevalence of gram-negative bacteria (n = 6 (4.8%)), the combination of meropenem and vancomycin was prescribed most frequently (n = 96 (78.0%)). CONCLUSION: The use of carbapenems as broad-spectrum antimicrobials agents might not be justified in most VLBW infants with FIP. Knowledge on the development of the neonatal gut microbiota, local resistance patterns and individual microbiological findings should be taken into consideration when implementing antimicrobial stewardship programs (ASPs).
RESUMO
BACKGROUND: The pretreatment International Neuroblastoma Risk Group Staging System (INRGSS) discriminates localized tumors L1/L2 depending on the absence/presence of image-defined risk factors (IDRFs) at diagnosis. Referring to this new staging system, we assessed initial imaging of localized thoracic neuroblastoma (NB) and ganglioneuroma (GN) and the extent of initial tumor resection. METHODS: Patients with localized thoracic NB/GN from the German clinical trials NB97 and NB2004 were included. Imaging at diagnosis and operative reports were reviewed retrospectively. IDRFs were assessed centrally and correlated to International Neuroblastoma Staging System (INSS) stage and extent of tumor resection. Additionally, we analyzed data on surgery-related complications. RESULTS: Imaging series of 88 patients were available for central review. In 18 children, no IDRF was present, 28 exhibited one IDRF, 42 two or more IDRFs, resulting in 70 patients with L2 disease. The most frequently observed IDRF was encasement of any vessel (n = 38). Initial surgical resection was aimed for in 45 patients (L1: n = 11; L2: n = 34). Complete and gross total resection rates were higher children with L2 NB (n = 8/25 L1, n = 17/25 L2 vs. n = 2/15 L1, n = 13/15 L2, respectively). The proportion of surgical complications was very similar between INRGSS L1 and L2 (n = 4/11 vs. n = 17/34). All complications were manageable, and no surgery-related deaths were observed. CONCLUSION: In this retrospective cohort, the extent of resection and the rate of surgical complications did not differ substantially between patients classified as L1/L2, indicating that INRGSS L2 does not equate unresectability. It appeared that individual IDRFs differ in value. Larger studies are needed to assess the significance and therapeutic/prognostic impact of such findings.
