RESUMO
Background: Pre-treatment stratification and outcomes of neuroblastoma patients often depend on the assessment of image-defined risk factors (IDRFs) on MR Imaging, usually using Gadolinium-contrast materials which are cautioned in pediatrics. We aimed to address whether gadolinium contrast-enhanced sequences are necessary to identify the presence/absence of IDRFs. Methods: Patients with neuroblastoma with MR imaging were retrospectively identified from 2005 to 2021. Ninety confirmed IDRFs were evaluated in 23 patients. Corresponding MR studies were anonymized, randomized, and independently evaluated by 3 fellowship-trained pediatric radiologists. Each radiologist assessed the studies twice. At the first reading, all enhanced sequences were omitted, while in the second reading, the full study with enhanced sequences were included. Consensus reading was obtained among readers. Inter- and intra-rater agreements using Kappa statistics (κ) as well as the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of non-enhanced MR in assessing IDRFs with respect to enhanced MR were calculated. Results: There were substantial (ĸ: 0.64-0.73) intra-reader agreements, and moderate to substantial (ĸ: 0.57-0.62) inter-reader agreements among radiologists in identifying IDRFs using non-enhanced MR. Non-enhanced MR had a sensitivity of 87.8% (95% CI [79-94]), specificity of 93% (89-96), PPV of 82.3 (73-89), NPV of 95.4 (92-98), and accuracy of 91.6 (88-94) in identifying IDRFs. However, 5/23 patients (21.7%) had a change in staging with the inclusion of contrast sequences. Conclusion: Although contrast sequences have a role in IDRF assessment, the majority can be adequately assessed on MR without gadolinium-contrast enhancement. Validation in a larger cohort is an important next step.
RESUMO
PURPOSE: To analyze the experience using compressed gelfoam plugs (CGPs) in children during liver biopsies and other interventional procedures. MATERIALS AND METHODS: This was a single-center, retrospective, consecutive case series of 477 various pediatric interventional radiology procedures using CGPs (January 2012 to December 2016) performed on 397 children (209 males, 188 females; median age, 7 years [range, 9 days-17.8 years]; median weight, 18 kg [range, 2.7-141 kg]). Of 477 procedures, most (n = 371) were liver biopsies, 98 were biopsies of other organs, and 8 were non-biopsy procedures. Analysis focused on liver biopsies. RESULTS: Of liver biopsies, a median of 2 CGPs were used per procedure, and the mean drop in hemoglobin was -0.36 g/dl (-3.0% change). Thirty-seven mild, 8 moderate, 2 severe, and 1 life-threatening (an anaphylaxis to CGP) adverse events (AEs) occurred. Analysis of liver biopsies with AEs showed significant association between number of passes, cores, and focal-type lesions (unadjusted logistic regression: P = .007, P = .022, P = .028, respectively) and age, weight, and number of passes (adjusted multiple logistic regression: P = .006, P = .032, P = .046, respectively). Technical problems relating to CGP deployment were noted in 5 (1%), without any AEs. CONCLUSIONS: CGPs were used in a wide variety of procedures and organs in children. There was 1 life-threatening AE resulting from the rare risk of anaphylaxis caused by the gelfoam-containing plug. After liver biopsies, transfusion was required in 2/371 (0.5%) procedures, 1 related to pre-biopsy anemia (0.25%).