Citations

Abstract BACKGROUND Clinicoradiologic differentiation between benign and malignant peripheral nerve sheath tumors (PNSTs) has important management implications. OBJECTIVE To develop and evaluate machine-learning approaches to differentiate benign from malignant PNSTs. METHODS We identified PNSTs treated at 3 institutions and extracted high-dimensional radiomics features from gadolinium-enhanced, T1-weighted magnetic resonance imaging (MRI) sequences. Training and test sets were selected randomly in a 70:30 ratio. A total of 900 image features were automatically extracted using the PyRadiomics package from Quantitative Imaging Feature Pipeline. Clinical data including age, sex, neurogenetic syndrome presence, spontaneous pain, and motor deficit were also incorporated. Features were selected using sparse regression analysis and retained features were further refined by gradient boost modeling to optimize the area under the curve (AUC) for diagnosis. We evaluated the performance of radiomics-based classifiers with and without clinical features and compared performance against human readers. RESULTS A total of 95 malignant and 171 benign PNSTs were included. The final classifier model included 21 imaging and clinical features. Sensitivity, specificity, and AUC of 0.676, 0.882, and 0.845, respectively, were achieved on the test set. Using imaging and clinical features, human experts collectively achieved sensitivity, specificity, and AUC of 0.786, 0.431, and 0.624, respectively. The AUC of the classifier was statistically better than expert humans ( P = .002). Expert humans were not statistically better than the no-information rate, whereas the classifier was ( P = .001). CONCLUSION Radiomics-based machine learning using routine MRI sequences and clinical features can aid in evaluation of PNSTs. Further improvement may be achieved by incorporating additional imaging sequences and clinical variables into future models.

Introduction: Cardiovascular implantable electronic devices (CIED) infections are frequently complicated with bloodstream infection (BSI). Complete device removal and prolonged antimicrobial therapy is essential for cure. The frequency of persistent BSI after source control (device extraction) is not well described. This study aims to assess the utility of repeat blood cultures (BC) after device extraction as recommended by the American Heart Association (AHA) guidelines. Methods: We selected patients who presented with BSI in the setting of CIED infection and underwent device removal at Mayo Clinic Rochester between 2012 and 2017. Cases where BC were not drawn prior to extraction, or repeated within 72 hours of device extraction, or met criteria for contamination were excluded. Results: Of 656 patients who underwent CIED extraction for device infection, 190 with post-extraction BC met study criteria. Among the final study cohort (Table 1), 159 patients had negative BC, while 31 had positive BC following CIED extraction. Pre-extraction, the most common causative organism was Staphylococcus aureus (91/190, 48%) (Figure 1a). Post-extraction, 22 of the 31 (71%) cases of persistent BSI were due to S. aureus (Fig 1b). The median duration of BSI was not significantly different between negative and positive BC groups (3 versus 4 days, p=0.92). Conclusions: Majority of patients had resolution of BSI after CIED extraction. S. aureus was the most common cause of persistent BSI post-extraction. Our findings support the AHA guidelines of repeating BC after device extraction, especially for S aureus .

Introduction: Heart failure (HF) is a leading cause of hospitalization, morbidity and mortality. Deep learning (DL) techniques appear to show promising results in risk stratification and prognosis in several conditions in medicine. However, few methods using DL exist to help quantitatively estimate prognosis of HF. We hypothesized that deep learning (DL) techniques could prognosis of HF using simple variables. We propose application of a custom-built deep-neural-network model to identify mortality in HF patients. Methods: Custom-built deep-neural-networks were assessed using survey data from 42,147 participants from the National Health and Nutrition Examination Survey 1999-2016 (NHANES). Variables were selected using clinical judgment and stepwise backward regressions to develop prediction models. We partitioned the data into training and testing sets and repetitive experiments. We then evaluated model performance based on discrimination and calibration including the area under the receiver-operator characteristics curve (C-statistics), balanced accuracy, probability calibration with sigmoid, and the Brier score, respectively. As sensitivity analyses, we examined results limited to cases with complete clinical information available. We validated models’ performance using Mount Sinai database. Results: Of 42,147 participants with 4,060 variables, 1,491 (3.5%) had HF and HF mortality was 51.8%. In validation cohort, of 26,333 HF patients, the mortality in HF patients was 405 (1.5%). Final model using only 20 variables (age, race, gender, BMI, smoking, alcohol consumption, HTN, COPD, SBP, DBP, HR, HDL, LDL, CRP, A1C, BUN, creatinine, hemoglobin, sodium level, on statin) was tested. A state-of-the-art deep learning models achieved high accuracy for predicting mortality in HF patients with an AUC of 0.96 (95% CI: 0.95-0.99) in the first cohort and AUC of 0.93 (95% CI: 0.91-0.96) in validation cohort. Conclusions: A deep neural network model has shown to have high predictive accuracy and discriminative and calibrative power for prediction of HF mortality. Further research can delineate the clinical implications of DL in predicting HF mortality.

Abstract Background Reductions in the use of broad-spectrum antibiotics is a cornerstone of antimicrobial stewardship. We aim to demonstrate use of the Desirability of Outcome Ranking Approach for the Management of Antimicrobial Therapy (DOOR MAT) to evaluate the treatment of Escherichia coli and Klebsiella pneumoniae bloodstream infections in patients from the Veterans Health Administration (VHA) across a decade. Methods Using electronic records, we determined empiric and definitive antibiotic treatments, clinical characteristics, and 30-day mortality of patients with monomicrobial E. coli and K. pneumoniae bloodstream infections hospitalized in VHA medical centers from 2009 to 2018. Focusing on patients treated with parenteral β-lactams and with available antibiotic susceptibility testing results, we applied a range of DOOR MAT scores that reflect the desirability of antibiotic choices according to spectrum and activity against individual isolates. We report trends in resistance and desirability of empiric and definitive antibiotic treatments. Results During the 10-year period analyzed, resistance to expanded-spectrum cephalosporins and fluoroquinolones increased in E. coli but not in K. pneumoniae, while resistance to carbapenems and piperacillin-tazobactam remained unchanged. In 6451 cases analyzed, we observed improvements in DOOR MAT scores consistent with deescalation. Improvement in desirability of definitive treatment compared with empiric treatment occurred in 26% of cases, increasing from 16% in 2009 to 34% in 2018. Reductions in overtreatment were sustained and without negative impact on survival. Conclusions DOOR MAT provides a framework to assess antibiotic treatment of E. coli and K. pneumoniae bloodstream infections and can be a useful metric in antimicrobial stewardship.

BackgroundPhysicians use fixed C‐arm fluoroscopy equipment with many interventional radiological and cardiological procedures. The associated effective dose to a patient is generally considered low risk, as the benefit‐risk ratio is almost certainly highly favorable. However, X‐ray‐induced skin injuries may occur due to high absorbed patient skin doses from complex fluoroscopically guided interventions (FGI). Suitable action levels for patient‐specific follow‐up could improve the clinical practice.There is a need for a refined metric regarding follow‐up of X‐ray‐induced patient injuries and the knowledge gap regarding skin dose‐related patient information from fluoroscopy devices must be filled. The most useful metric to indicate a risk of erythema, epilation or greater skin injury that also includes actionable information is the peak skin dose, that is, the largest dose to a region of skin.Materials and MethodsThe report is based on a comprehensive review of best practices and methods to estimate peak skin dose found in the scientific literature and situates the importance of the Digital Imaging and Communication in Medicine (DICOM) standard detailing pertinent information contained in the Radiation Dose Structured Report (RDSR) and DICOM image headers for FGI devices. Furthermore, the expertise of the task group members and consultants have been used to bridge and discuss different methods and associated available DICOM information for peak skin dose estimation.ResultsThe report contributes an extensive summary and discussion of the current state of the art in estimating peak skin dose with FGI procedures with regard to methodology and DICOM information. Improvements in skin dose estimation efforts with more refined DICOM information are suggested and discussed.ConclusionsThe endeavor of skin dose estimation is greatly aided by the continuing efforts of the scientific medical physics community, the numerous technology enhancements, the dose‐controlling features provided by the FGI device manufacturers, and the emergence and greater availability of the DICOM RDSR. Refined and new dosimetry systems continue to evolve and form the infrastructure for further improvements in accuracy. Dose‐related content and information systems capable of handling big data are emerging for patient dose monitoring and quality assurance tools for large‐scale multihospital enterprises.

Abstract Comparative data guiding initial therapy for Waldenström macroglobulinemia (WM), an infrequently encountered non‐Hodgkin lymphoma, are sparse. We evaluated three commonly used rituximab‐based frontline regimens: rituximab‐bendamustine (R‐Benda); dexamethasone, rituximab, cyclophosphamide (DRC); and bortezomib, dexamethasone, rituximab (BDR) in 220 treatment‐naïve patients with WM, seen at Mayo Clinic between November 1, 2000 and October 31, 2019. The median follow‐up was 4.5 (95%CI: 4–5) years. The R‐Benda cohort (n = 83) demonstrated superior overall response rate (ORR: 98%), in comparison to DRC (n = 92, ORR: 78%) or BDR (n = 45, ORR: 84%) cohorts, p = 0.003. Similarly, longer progression‐free survival (PFS) was evident with R‐Benda use [median 5.2 vs. 4.3 (DRC) and 1.8 years (BDR), p < 0.001]. The time‐to‐next therapy (TTNT) favored R‐Benda [median, not‐reached, 4.4 (DRC) and 2.6 years (BDR), p < 0.001). These endpoints were comparable between the DRC and BDR cohorts. Overall survival (OS) was similar across the three cohorts, p = 0.77. In a subset analysis of 142 patients genotyped for MYD88 L265P mutation, the ORR, PFS and TTNT were unaffected by the patients' MYD88 signature within each cohort. In conclusion, ORR, PFS and TTNT with R‐Benda are superior compared to DRC or BDR in treatment‐naïve patients with active WM. The patient outcomes with any one of these three regimens are unaffected by the MYD88 L265P mutation status.

Abstract Purpose: Small-bowel adenocarcinoma (SBA) is rare, and no standard of care exists for metastatic disease beyond first-line FOLFOX/CAPOX. SBA has higher rates of microsatellite instability (MSI-H) and T-lymphocyte infiltration than other gastrointestinal cancers. We hypothesize that pembrolizumab, a PD-1 inhibitor, will induce antitumor response. Patients and Methods: Patients with previously treated advanced SBA received pembrolizumab 200 mg i.v. every 3 weeks until disease progression (PD), toxicity, or 35 doses maximum. Primary endpoint was confirmed overall response rate (ORR) with secondary progression-free survival (PFS), overall survival (OS), and toxicity assessment endpoints. Outcomes were stratified by tumor location, microsatellite stability (MSS) or instability (MSI-H), and PD-L1 level. Results: Forty patients were treated for a median duration of four cycles (range, 1–35). All patients are off study treatment due to PD (75%), death (10%), 35 cycles completed (8%), refusal (3%), and adverse effects (AEs, 5%). Three confirmed partial responses [PRs; 8%; 95% confidence interval (CI), 2–20] did not meet predefined success criteria of ORR 30%. Median OS (7.1 months; 95% CI, 5.1–17.1) and median PFS (2.8 months; 95% CI, 2.7–4.2) were similar across primary tumor sites. One confirmed PR (3%) was seen in patients with low MSS/MSI tumors and correlated with high tumor mutation burden (TMB). Fifty percent of patients with MSI-H tumors achieved PR and remain alive without progression. Twenty-five patients (63%) had grade ≥3 AEs and 11 patients (28%) had grade 4/5 AEs. Conclusions: In the largest study of SBA to date, pembrolizumab did not induce the hypothesized response rate; however, we did identify responses in key biomarker-selected cohorts.