SUMMARY Postchemotherapy residual masses in GCT can prove to be complicated cases to manage. Balancing survival with quality of life outcomes is important and requires a multidisciplinary team experienced in treating GCT.Immunohistochemical staining for Ki-67 is used to calculate a Ki-67 proliferation index (PI) that carries prognostic and predictive information in various cancers including breast carcinomas. Studies have documented challenges for observers to reproducibly estimate the Ki-67 PI. At present, no international consensus exists concerning scoring method (eg, hotspots vs. overall average, digital vs. manual counting) or even the definition of a Ki-67-positive cell. To clarify the approach to Ki-67 scoring and evaluation of the interobserver agreement among participants in the Nordic Immunohistochemical Quality Control (NordiQC) Breast Cancer Module, a study was set up on the basis of an online web module containing 15 digitized tissue microarray cores of breast carcinomas stained for Ki-67 in the NordiQC reference laboratory. All participants were invited to attend the study. In addition to Ki-67 scoring, they were asked to disclose their preferred method for Ki-67 estimation and their job title. For comparison, slides were analyzed using a Digital Image Analysis algorithm based on Virtual Double Staining. In total, 199 participants enrolled for the study. Overall, there was a good correlation in Ki-67 PIs among the participants, although results for some cores varied significantly. However, when applying a cutoff of 20%, a relatively low κ value of 0.52 was observed. Participants scoring in hotspots reported higher Ki-67 PIs than participants estimating an overall average, and, not surprisingly, participants who considered weak Ki-67 nuclear staining positive obtained higher Ki-67 PIs than those who did not. Ki-67 PI was not correlated to job title. The Virtual Double Staining algorithm obtained Ki-67 values close to the mean value of the human observers. Our study underlines the need for international standardization and guidelines in estimation of Ki-67 PI. Digital Image Analysis may be a useful tool in this process.BACKGROUND Simulation is invaluable for bronchoscopy training. Studies report improved procedure time, dexterity/technique, and trainee satisfaction supported by low-fidelity and high-fidelity simulators in structured-training programs. We sought to determine (1) Learning-gain in bronchoscopic dexterity after a single 45-minute unstructured exposure using a low-fidelity simulator. (2) Whether acquired skills are maintained 8 weeks later, during which trainees receive no interim exposure to simulation or clinical bronchoscopy. METHODS Using a low-fidelity model, medical students were assessed for bronchoscopicdexterity before and after an unstructured, self-directed 45-minute simulation. Bronchoscopic dexterity was assessed according to (1) Ability to enter a target-bronchus within a specified time. (2) The modified Bronchoscopy Skills and Tasks Assessment Tool (mBSTAT). Scores were compared at baseline, postsimulation, and 8 weeks postsimulation. Individual domains of the mBSTAT were compared with identify specific skills demonstrating more significant deterioration. RESULTS Twenty-eight medical students completed the initial-simulation session. Fifteen returned at 8 weeks. Statistically significant improvement in bronchoscopic-skills was observed immediately following the simulation session (mBSTAT scores 3.7±1.2 pretest vs. 7.0±0.9 posttest, P less then 0.001). mBSTAT scores had deteriorated significantly at 8 weeks (5.7±1.8, P=0.03) but remained superior to baseline scores (P=0.002). Of the 4 domains assessed, only Precision did not demonstrate any change between post-test and review assessments (P=0.14). All other domains demonstrated trends towards significant deterioration between posttest and review. CONCLUSION A single 45-minute unstructured bronchoscopy simulation session resulted in significant improvement in bronchoscopic dexterity. Significant decay in bronchoscopic dexterity is observed, suggesting repeat simulation may be valuable following periods without bronchoscopy exposure.Duodenal neoplasms of gastric phenotype are uncommon epithelial neoplasms. Pyloric gland adenomas should be recognized as neoplasms with risk for transformation into invasive adenocarcinoma (Choi et al. 2018). Here we report the case histories of two male patients, who presented with duodenal polypoid lesion. Endoscopic polypectomy and endoscopic submucosal dissection were carried out, respectively. Histopathologically, both polyps showed features of neoplasms of gastric phenotype. The clinical and endoscopic features, pathologic features, immunophenotype, molecular pathogenesis, clinical management and prognosis of the two cases will be discussed. We will also briefly review the latest literatures on duodenal neoplasms of gastric phenotype.OBJECTIVES Several predictive equations have been developed for estimation of resting energy expenditure, but no study has been done to compare predictive equations against indirect calorimetry among critically ill patients at different phases of critical illness. This study aimed to determine the degree of agreement and accuracy of predictive equations among ICU patients during acute phase (≤ 5 d), late phase (6-10 d), and chronic phase (≥ 11 d). DESIGN This was a single-center prospective observational study that compared resting energy expenditure estimated by 15 commonly used predictive equations against resting energy expenditure measured by indirect calorimetry at different phases. Degree of agreement between resting energy expenditure calculated by predictive equations and resting energy expenditure measured by indirect calorimetry was analyzed using intraclass correlation coefficient and Bland-Altman analyses. Resting energy expenditure values calculated from predictive equations differing by ± 10% framer (1990). None of the resting energy expenditure calculated from predictive equations showed very good agreement or accuracy. selleck CONCLUSIONS Predictive equations tend to either over- or underestimate resting energy expenditure at different phases. Predictive equations with “dynamic” variables and respiratory data had better agreement with resting energy expenditure measured by indirect calorimetry compared with predictive equations developed for healthy adults or predictive equations based on “static” variables. Although none of the resting energy expenditure calculated from predictive equations had very good agreement, Swinamer (1990) appears to provide relatively good agreement across three phases and could be used to predict resting energy expenditure when indirect calorimetry is not available.