Characterization of a Novel Mouse Model for Fuchs Endothelial Corneal Dystrophy

Purpose

Fuchs endothelial corneal dystrophy (FECD) is a progressive blinding disorder, characterized by increased corneal endothelial excrescences (guttae), corneal endothelial cell loss, and edema. These symptoms are hypothesized to be caused by changes in the extracellular matrix (ECM) and mitochondrial dysfunction in the corneal endothelium. Despite this clinical and biological relevance, a comprehensive animal model that recapitulates all the major disease characteristics is currently unavailable. In this study, we develop such a model to improve our understanding of the signaling pathways involved in the FECD progression and develop strategies for early intervention.

Method

To generate a comprehensive FECD model, we generated a double mutant mouse bearing tamoxifen-inducible knockdown of Slc4a11 and the Col8a2 (Q455K) mutation. We performed optical coherence tomography (OCT) and in vivo confocal microscopy using the Heidelberg Retinal Tomography 3 - Rostock Cornea module (HRT3-RCM) on the mice at 5 weeks of age before tamoxifen feeding to establish baseline values for corneal thickness, endothelial cell density, and test for the presence of guttae. We measured these parameters again post-tamoxifen treatment at 16 weeks of age. We collected corneas at 16 weeks to perform histopathology, immunofluorescence staining for tight junctions, adherens junctions, and oxidative stress. We evaluated endothelial pump function using a lactate assay.

Results

The double mutant tamoxifen-fed animals showed the presence of guttae, and displayed increased corneal thickness and decreased endothelial cell density. Endothelial cells showed altered morphology with disrupted adherens junctions and elevated reactive oxygen species (ROS). Finally, we found that stromal lactate concentrations were elevated in the double mutant mice, indicative of compromised endothelial pump function.

Conclusions

Overall, this mouse model recapitulates all the important phenotypic features associated with FECD.

Pan-Asian adapted ESMO Clinical Practice Guidelines for the diagnosis, treatment and follow-up of patients with gastric cancer

The European Society for Medical Oncology (ESMO) Clinical Practice Guidelines for the diagnosis, treatment and follow-up of patients with gastric cancer (GC), published in late 2022 and the updated ESMO Gastric Cancer Living Guideline published in July 2023, were adapted in August 2023, according to previously established standard methodology, to produce the Pan-Asian adapted (PAGA) ESMO consensus guidelines for the management of Asian patients with GC. The adapted guidelines presented in this manuscript represent the consensus opinions reached by a panel of Asian experts in the treatment of patients with GC representing the oncological societies of China (CSCO), Indonesia (ISHMO), India (ISMPO), Japan (JSMO), Korea (KSMO), Malaysia (MOS), the Philippines (PSMO), Singapore (SSO), Taiwan (TOS) and Thailand (TSCO), coordinated by ESMO and the Japanese Society of Medical Oncology (JSMO). The voting was based on scientific evidence and was independent of the current treatment practices, drug access restrictions and reimbursement decisions in the different Asian regions represented by the 10 oncological societies. The latter are discussed separately in the manuscript. The aim is to provide guidance for the optimisation and harmonisation of the management of patients with GC across the different regions of Asia, drawing on the evidence provided by both Western and Asian trials, whilst respecting the differences in screening practices, molecular profiling and age and stage at presentation. Attention is drawn to the disparity in the drug approvals and reimbursement strategies, between the different regions of Asia.

Qualitative systematic review: The lived experiences of males in the nursing profession on gender discrimination encounters

AIMS

This review aims to synthesise the best available evidence exploring the lived experiences of males in the nursing profession on gender discrimination encounters.

BACKGROUND

The underrepresentation of males within the nursing profession could potentially cause discrimination and stereotypes against them, causing barriers to entry into, and retention within, the profession. With a shortage in manpower, the deterrence of males to nursing might exacerbate the existing manpower shortage. This review consolidates and highlights the lived experiences of males in the profession.

METHODS

The review was synthesised according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist. Employing the meta-synthesis approach, two independent reviewers conducted critical appraisal and data extraction of included studies, using the standard Joanna Briggs Institute (JBI) Critical Appraisal Instrument for Qualitative Research and a standardised data extraction form, respectively.

RESULTS

Twenty-three studies were included. Extracted findings were aggregated into 14 categories and synthesised into three themes: stereotypes, discrimination and challenges faced by males.

DISCUSSION

The synthesised findings showed that male nurses faced gender discrimination in multiple facets. These caused them to have difficulty in carrying out their nursing duties, missing out of clinical opportunities and causing them multiple social and mental stressors.

CONCLUSION

This review informed the discriminatory experiences faced by males within the nursing profession. With males already underrepresented within this profession, it is imperative for interventions, policies and research to be made to address this issue to improve experiences of male nurses, thereby limiting the deterrence of males into the profession.

IMPLICATIONS

Findings of this study showed that males are being stereotyped and discriminated, resulting in challenges faced solely because of their gender. Effective interventions and policies could be implemented to alleviate the negative experiences.

Location of Local Recurrences Post-Prostatectomy Detected by 18F-DCPFyL and Implications for Salvage Radiotherapy Target Volumes

PURPOSE/OBJECTIVE(S)

The use of prostate-specific membrane antigen (PSMA) PET scans have improved detection of recurrent prostate cancer for biochemical failure post-prostatectomy. The primary aim of this study is to assess the location of local recurrences for PSMA PET positive disease post-prostatectomy relative to anatomical landmarks and to inform salvage radiotherapy contouring guidelines.

MATERIALS/METHODS

This is a retrospective study of patients who underwent a staging PSMA PET-CT scan with 18F-DCFPyL tracer. Eligible patients included men with biochemical recurrence (PSA >0.2ng/mL) post-prostatectomy for prostate adenocarcinoma. PSMA positive local disease (GTV) was marked on the CT scan of the PSMA PET/CT, using the PSMA PET scan fused as the secondary dataset. Reference structures were contoured to compare the positional relationship to the GTV local recurrence. Centroids of each GTV position were measured to the closest border of the reference structures and distance measured.

RESULTS

Forty-five patients had a local recurrence on PSMA PET, 28/45 with prostate bed recurrence and 17/45 with seminal vesicle bed recurrence. The median age was 72.7 years, median pre-treatment PSA 0.515 ng/ml (0.2-11.1), pT2 in 37.8% and pT3 in 62.2, ISUP G2-3 was 66.7%, ISUP 4-5 29.6%, 67% ECE, 15.4% SV involvement and 40% positive margin. For prostate bed recurrences, 93.6% of GTVs were within 15mm from the inferior border of pubic symphysis (median 3.3mm [-9.8-15.4]), 83.9% within 10mm from the VUA (median 3.3mm [-17.4-15.9]), and 80.6% were greater than 10mm from the most superior slice of the penile bulb (median 16.4mm [1-29.5]). Laterally, 83.9% of recurrences were within 10mm of midline and 41.9% within 5mm (median 3.3mm [-13.56-11.59]) and all were contained within levator, 35.5% within 5mm (median 7.7 [1-22.6]) of the medial muscle border. 29.0% were within 5mm from the anterior edge of the rectal wall (median -7mm [-32.3-0.5]). Seminal vesicle bed recurrences were within 17.5mm of upper border of pubic symphysis for 88.9% of patients (median -4.69mm[-30.8to18.2]), 100% were more than 7.5mm inferiorly from the vas deferens (median-22.1mm[-32.7to-7.7]). Laterally, 100% recurrences were more than 7.5mm medial from obturator internus (median 5.37mm [-23.1to37.8]), and 88.9% were more than 10mm from midline (median 23.1mm [2.0-36.6]).

CONCLUSION

The 18F-DCFPyL PSMA PET local recurrences can provide detailed localization of prostate bed and seminal vesicle recurrences. Our data can inform contouring guidelines for salvage radiotherapy to the prostate bed, particularly when using anatomical landmarks to define cranio-caudal extent of target volumes. For target volume laterally, the medial wall of levator ani would encompass all prostate bed recurrences, and no seminal vesicles recurrences were seen at the medial border of the obturator internus muscle.

Estimating Uncertainty in Neural Networks for Cardiac MRI Segmentation: A Benchmark Study

OBJECTIVE

Convolutional neural networks (CNNs) have demonstrated promise in automated cardiac magnetic resonance image segmentation. However, when using CNNs in a large real-world dataset, it is important to quantify segmentation uncertainty and identify segmentations which could be problematic. In this work, we performed a systematic study of Bayesian and non-Bayesian methods for estimating uncertainty in segmentation neural networks.

METHODS

We evaluated Bayes by Backprop, Monte Carlo Dropout, Deep Ensembles, and Stochastic Segmentation Networks in terms of segmentation accuracy, probability calibration, uncertainty on out-of-distribution images, and segmentation quality control.

RESULTS

We observed that Deep Ensembles outperformed the other methods except for images with heavy noise and blurring distortions. We showed that Bayes by Backprop is more robust to noise distortions while Stochastic Segmentation Networks are more resistant to blurring distortions. For segmentation quality control, we showed that segmentation uncertainty is correlated with segmentation accuracy for all the methods. With the incorporation of uncertainty estimates, we were able to reduce the percentage of poor segmentation to 5% by flagging 31-48% of the most uncertain segmentations for manual review, substantially lower than random review without using neural network uncertainty (reviewing 75-78% of all images).

CONCLUSION

This work provides a comprehensive evaluation of uncertainty estimation methods and showed that Deep Ensembles outperformed other methods in most cases.

SIGNIFICANCE

Neural network uncertainty measures can help identify potentially inaccurate segmentations and alert users for manual review.

What is hemoglobin, albumin, lymphocyte, platelet (HALP) score? A comprehensive literature review of HALP's prognostic ability in different cancer types

Since its inception, the Hemoglobin, Albumin, Lymphocyte, Platelet Score (HALP) has gained attention as a new prognostic biomarker to predict several clinical outcomes in a multitude of cancers. In our review, we searched PubMed for articles between the first paper on HALP in 2015 through September 2022, yielding 32 studies in total that evaluated HALP's association with various cancers, including Gastric, Colorectal, Bladder, Prostate, Kidney, Esophageal, Pharyngeal, Lung, Breast, and Cervical cancers, among others. This review highlights the collective association HALP has with demographic factors such as age and sex in addition to TNM staging, grade, and tumor size. Furthermore, this review summarizes HALP's prognostic ability to predict overall survival, progression-free survival, recurrence-free survival, among other outcomes. In some studies, HALP has also been able to predict response to immunotherapy and chemotherapy. This review article also aims to serve as a comprehensive and encyclopedic report on the literature that has evaluated HALP as a biomarker in various cancers, highlighting the heterogeneity surrounding HALP's utilization. Because HALP requires only a complete blood count and albumin - already routinely collected for cancer patients - HALP shows potential as a cost-effective biomarker to aid clinicians in improving outcomes for immuno-nutritionally deficient patients.

A Deep Learning Segmentation Pipeline for Cardiac T1 Mapping Using MRI Relaxation-based Synthetic Contrast Augmentation

PURPOSE

To design and evaluate an automated deep learning method for segmentation and analysis of cardiac MRI T1 maps with use of synthetic T1-weighted images for MRI relaxation-based contrast augmentation.

MATERIALS AND METHODS

This retrospective study included MRI scans acquired between 2016 and 2019 from 100 patients (mean age ± SD, 55 years ± 13; 72 men) across various clinical abnormalities with use of a modified Look-Locker inversion recovery, or MOLLI, sequence to quantify native T1 (T1_native), postcontrast T1 (T1_post), and extracellular volume (ECV). Data were divided into training (n = 60) and internal (n = 40) test subsets. "Synthetic" T1-weighted images were generated from the T1 exponential inversion-recovery signal model at a range of optimal inversion times, yielding high blood-myocardium contrast, and were used for contrast-based image augmentation during training and testing of a convolutional neural network for myocardial segmentation. Automated segmentation, T1, and ECV were compared with experts with use of Dice similarity coefficients (DSCs), correlation coefficients, and Bland-Altman analysis. An external test dataset (n = 147) was used to assess model generalization.

RESULTS

Internal testing showed high myocardial DSC relative to experts (0.81 ± 0.08), which was similar to interobserver DSC (0.81 ± 0.08). Automated segmental measurements strongly correlated with experts (T1_native, R = 0.87; T1_post, R = 0.91; ECV, R = 0.92), which were similar to interobserver correlation (T1_native, R = 0.86; T1_post, R = 0.94; ECV, R = 0.95). External testing showed strong DSC (0.80 ± 0.09) and T1_native correlation (R = 0.88) between automatic and expert analysis.

CONCLUSION

This deep learning method leveraging synthetic contrast augmentation may provide accurate automated T1 and ECV analysis for cardiac MRI data acquired across different abnormalities, centers, scanners, and T1 sequences.Keywords: MRI, Cardiac, Tissue Characterization, Segmentation, Convolutional Neural Network, Deep Learning Algorithms, Machine Learning Algorithms, Supervised Learning Supplemental material is available for this article. © RSNA, 2022.

A nutritional screening threshold for optimal outcomes after Hartmann's reversal

PURPOSE

Determining is nutritionally replete enough for Hartmann's reversal (HR) can be controversial and multifactorial. While there are many preoperative nutritional screening tools, the impact of malnourishment on HR has not been evaluated. The study aims to clarify how often patients undergoing HR are high risk for malnourishment at the time of surgery and how this impacts postoperative outcomes.

METHODS

From 2012-2019, all elective HRs were identified in ACS-NSQIP. Patients were categorized in a malnourished group if they met one of the following criteria: (1) BMI < 18.5 kg/m², (2) albumin < 3.5 g/dL, or (3) > 10% body weight loss in the last 6 months. Bivariate associations of preoperative demographics and postoperative outcomes were analyzed. Multivariable logistic regression was performed to identify independent predictors for 30-day mortality and organ space wound infection.

RESULTS

8878 procedures were evaluated (well-nourished = 7116 and malnourished = 1762). The malnourished group had higher mortality (p < 0.001), shorter operating time (p < .001), longer length of stay (p = 0.016), and higher rates of infection (p = 0.011), reintubation (p = 0.002), bleeding (p < 0.001), sepsis (p = 0.001), and reoperation (p = 0.018). In multivariate regression models, malnourishment was an independent predictor for mortality (OR = 2.72, p < 0.001) and wound infection (OR = 1.19, p = 0.028).

CONCLUSION

A large percentage of patients undergoing HR were classified as being high-risk for malnutrition. Malnourishment was associated with some worse postoperative compilations including death and wound infection. Surgeons should routinely use preoperative screening for malnutrition to identify and attempt to optimize nutritional status prior to undergoing Hartmann's Reversal.

The effect of clinical interventions on the oral health-related quality of life in older adults

This paper reviews the literature to assess the effectiveness of oral health interventions in influencing oral health-related quality of life (OHRQoL) in older adults. Clinical studies published in 2010 onwards reporting oral health interventions in older adults and subsequent changes in OHRQoL were systematically searched in Medline, Web of Science and EMBASE databases. Reporting followed Preferred Reporting Items for Systematic Review and Meta-Analysis criteria. Eligibility evaluation and data abstraction from eligible studies were guided by predefined PICO questions and bias assessment by the Quality in Prognosis Studies tool. The search retrieved 144 articles. Of the 20 articles that met the eligibility criteria, 45% (9) reported a statistically significant improvement in OHRQoL. The majority of studies were conducted in adults with complete edentulism (70%) and mainly focused on implant provision to support a removable prosthesis (30%). The search indicates the limited scope of current literature on geriatric dental interventions, focusing mainly on denture realignments, denture fabrication and implant placement. Although the majority of the oral health interventions focused on edentulism and removable prostheses, there are many areas of oral health interventions and their potential impact on OHRQoL in this demographic group that have yet to be explored.

A phase 1b study of OXIRI in pancreatic adenocarcinoma patients and its immunomodulatory effects

Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal and debilitating disease with limited therapeutic options. The aim of this clinical study was to evaluate the safety, efficacy and pharmacokinetics of a novel regimen comprised of metronomic oxaliplatin (O), chronomodulated capecitabine (X) and UGT1A1 genotype-guided dosing of irinotecan (IRI) [OXIRI] as well as its immunomodulatory effects. Thirty-six patients were enrolled into either dose-escalation or expansion cohorts. In the dose escalation phase, capecitabine doses (2000, 2650, 3500 and 4500 mg/day) were administered at midnight on days 1 to 14 while oxaliplatin and irinotecan were intravenously infused at fixed doses of 50 and 75 mg/m2 respectively on days 1, 8 in a 21-day cycle. The maximum tolerated dose of capecitabine was 2650 mg/day and the most common grade 3 adverse events were neutropenia (30.6%) and diarrhea (13.9%). No grade 4 toxicity was observed. UGT1A1-genotype directed dosing resulted in similar exposure levels of irinotecan, SN-38 and SN-38G in all patients. Objective response rate was 22.2%. Median overall survival and progression-free survival were 8.1 and 5.2 months, respectively. Exploratory immunoprofiling by flow cytometry and quantitative spatial localization analysis of infiltrated immune cells performed on biopsy and plasma samples revealed significant declines in CCL22, CCL2 and TNFα levels at end of first cycle and an active host immune response. Our study showed that OXIRI was well-tolerated and exhibited good efficacy, with immunomodulatory effects. It may be considered as an alternative to FOLFIRINOX in patients intolerant to the latter.

Cardiac MRI segmentation with sparse annotations: Ensembling deep learning uncertainty and shape priors

The performance of deep learning for cardiac magnetic resonance imaging (MRI) segmentation is oftentimes degraded when using small datasets and sparse annotations for training or adapting a pre-trained model to previously unseen datasets. Here, we developed and evaluated an approach to addressing some of these issues to facilitate broader use of deep learning for short-axis cardiac MRI segmentation. We developed a globally optimal label fusion (GOLF) algorithm that enforced spatial smoothness to generate consensus segmentation from segmentation predictions provided by a deep learning ensemble algorithm. The GOLF consensus was entered into an uncertainty-guided coupled continuous kernel cut (ugCCKC) algorithm that employed normalized cut, image-grid continuous regularization, and "nesting" and circular shape priors of the left ventricular myocardium (LVM) and cavity (LVC). In addition, the uncertainty measurements derived from the segmentation predictions were used to constrain the similarity of GOLF and final segmentation. We optimized ugCCKC through upper bound relaxation, for which we developed an efficient coupled continuous max-flow algorithm implemented in an iterative manner. We showed that GOLF yielded average symmetric surface distance (ASSD) 0.2-0.8 mm lower than an averaging method with higher or similar Dice similarity coefficient (DSC). We also demonstrated that ugCCKC incorporating the shape priors improved DSC by 0.01-0.05 and reduced ASSD by 0.1-0.9 mm. In addition, we integrated GOLF and ugCCKC into a deep learning ensemble algorithm by refining the segmentation of an unannotated dataset and using the refined segmentation to update the trained models. With the proposed framework, we demonstrated the capability of using relatively small datasets (5-10 subjects) with sparse (5-25% slices labeled) annotations to train a deep learning algorithm, while achieving DSC of 0.871-0.893 for LVM and 0.933-0.959 for LVC on the LVQuan dataset, and these were 0.844-0.871 for LVM and 0.923-0.931 for LVC on the ACDC dataset. Furthermore, we showed that the proposed approach can be adapted to substantially alleviate the domain shift issue. Moreover, we calculated a number of commonly used LV function measurements using the derived segmentation and observed strong correlations (Pearson r=0.77-1.00, p<0.001) between algorithm and manual LV function analyses. These results suggest that the developed approaches can be used to facilitate broader application of deep learning in research and clinical cardiac MR imaging workflow.

Demographic and clinical factors associated with psychiatric inpatient admissions during the COVID-19 pandemic

OBJECTIVE

The COVID-19 pandemic may cause a major mental health impact. We aimed to identify demographic or clinical factors associated with psychiatric admissions where COVID-19 was attributed to contribute to mental state, compared to admissions which did not.

METHODS

A retrospective cohort study was undertaken of inpatients admitted to Northern Psychiatric Unit 1, Northern Hospital in Melbourne, Victoria, Australia during 27/02/2020 to 08/07/2020. Data were extracted for participants who identified COVID-19 as a stressor compared to participants who did not. Fisher's exact test and Mann-Whitley rank sum test were used.

RESULTS

Thirty six of 242 inpatients reported the COVID-19 pandemic contributed to mental ill health and subsequent admission. Reasons given included social isolation, generalized distress about the pandemic, barriers to support services, disruption to daily routine, impact on employment, media coverage, re-traumatization, cancelled ECT sessions, loss of loved ones, and increased drug use during the lockdown. Chronic medical conditions or psychiatric multimorbidity were positively associated and smoking status was negatively associated with reporting the COVID-19 pandemic as a contributor to mental ill health.

CONCLUSION

Screening and identifying vulnerable populations during and after the global disaster is vital for timely and appropriate interventions to reduce the impact of the pandemic worldwide.

Modeling attachment and compressive loading of locking and non-locking plate fixation: a finite element investigation of a supracondylar femur fracture model

This study developed a finite element (FE) model of simulated locking plate fixation to examine the strain response following supracondylar femoral plate attachment and under compressive loading. An implicit FE model of a synthetic femur with a distal fracture gap stabilized with a lateral plate was evaluated following attachment and 500 N loading, considering locking and non-locking proximal screws configurations. Screw pre-tension values of 60 N for both distal and proximal non-locking screws yielded good agreement with plate experimental strain data in attached (unloaded) and loaded conditions. The results highlight the importance of pre-tensioning in modeling plate attachment using non-locking screws.

1540 Genome-Wide Association Analysis In 401,583 Individuals Identifies Novel Therapeutic Targets for Haemorrhoids

Aim

To elucidate the genetic architecture of haemorrhoids and identify genes and biological pathways central to their pathobiology.

Method

We report the first ever genome-wide association study of haemorrhoids in 31,652 cases and 369,931 controls from UK Biobank. Genes and biological pathways were prioritised using several bioinformatic approaches, and potential therapeutic targets were identified in the Open Targets Platform. A weighted genetic risk score (wGRS) for haemorrhoids was constructed to compare genetic susceptibility in surgical vs non-surgical haemorrhoids patients.

Results

Twelve novel genome-wide significant susceptibility loci were discovered to be associated with haemorrhoids. Seventeen genes were mapped to these loci, and gene sets in biological pathways relating to extracellular matrix regulation and TGF-β signalling were strongly implicated. Seven gene-products (41.2%) were predicted tractable to antibody and/or small molecule targeting, and three products (17.6%) have known pharmaceutical interactions (ACHE, ADRA2B, ELN). The wGRS analysis demonstrated that haemorrhoid patients requiring surgery have a higher inherent genetic susceptibility than those managed non-surgically (P = 4.58 × 10-27).

Conclusions

This study has advanced our understanding of haemorrhoids pathobiology with the identification of several biologically plausible genes and pathways, many of which demonstrate strong therapeutic potential. The wGRS correlated with disease severity, representing a first step in personalised medicine approaches to haemorrhoids.

1625 Genome-Wide Association Analysis and Replication In 810,625 Individuals Identifies Novel Therapeutic Targets for Varicose Veins

Aim

To elucidate the genetic architecture of varicose veins (VVs) and identify genes and biological pathways central to their pathobiology.

Method

We performed hitherto the largest two-stage genome-wide association study of VVs in 401,656 subjects from UK Biobank, and replication in 408,969 subjects from 23andMe, Inc (total 135,514 VVs cases and 675,111 controls). Genes and biological pathways were prioritised using several bioinformatic approaches, and potential therapeutic targets were identified in the Open Targets Platform. A weighted genetic risk score (wGRS) for VVs was constructed to compare genetic susceptibility in surgical vs non-surgical VVs patients.

Results

109 genome-wide significant (P ≤ 5 × 10-8) loci were identified in UK Biobank, 46 of which successfully replicated in the 23andMe cohort. Twenty-eight loci have not been previously reported. We mapped 237 genes to these loci, many of which are biologically relevant and tractable to therapeutic targeting or repurposing (notably VEGFA, COL27A1, EFEMP1, PPP3R1 and NFATC2). Tissue enrichment analyses implicated vascular tissue, and several genes were enriched in biological pathways relating to extracellular matrix biology, inflammation, angiogenesis, lymphangiogenesis, vascular smooth muscle cell migration, and apoptosis. The wGRS analysis demonstrated that VVs patients requiring surgery have a higher inherent genetic susceptibility than those managed non-surgically (P = 2.46 × 10−13).

Conclusions

This study has advanced our understanding of VVs pathobiology with the identification of several biologically plausible genes and pathways, many of which demonstrate strong therapeutic potential. The wGRS correlated with disease severity, representing a first step in personalised medicine approaches to VVs.

Left Ventricle Quantification Challenge: A Comprehensive Comparison and Evaluation of Segmentation and Regression for Mid-Ventricular Short-Axis Cardiac MR Data

Automatic quantification of the left ventricle (LV) from cardiac magnetic resonance (CMR) images plays an important role in making the diagnosis procedure efficient, reliable, and alleviating the laborious reading work for physicians. Considerable efforts have been devoted to LV quantification using different strategies that include segmentation-based (SG) methods and the recent direct regression (DR) methods. Although both SG and DR methods have obtained great success for the task, a systematic platform to benchmark them remains absent because of differences in label information during model learning. In this paper, we conducted an unbiased evaluation and comparison of cardiac LV quantification methods that were submitted to the Left Ventricle Quantification (LVQuan) challenge, which was held in conjunction with the Statistical Atlases and Computational Modeling of the Heart (STACOM) workshop at the MICCAI 2018. The challenge was targeted at the quantification of 1) areas of LV cavity and myocardium, 2) dimensions of the LV cavity, 3) regional wall thicknesses (RWT), and 4) the cardiac phase, from mid-ventricle short-axis CMR images. First, we constructed a public quantification dataset Cardiac-DIG with ground truth labels for both the myocardium mask and these quantification targets across the entire cardiac cycle. Then, the key techniques employed by each submission were described. Next, quantitative validation of these submissions were conducted with the constructed dataset. The evaluation results revealed that both SG and DR methods can offer good LV quantification performance, even though DR methods do not require densely labeled masks for supervision. Among the 12 submissions, the DR method LDAMT offered the best performance, with a mean estimation error of 301 mm 2 for the two areas, 2.15 mm for the cavity dimensions, 2.03 mm for RWTs, and a 9.5% error rate for the cardiac phase classification. Three of the SG methods also delivered comparable performances. Finally, we discussed the advantages and disadvantages of SG and DR methods, as well as the unsolved problems in automatic cardiac quantification for clinical practice applications.

Management of traumatic laryngotracheal separation: Case series and review

Laryngotracheal separation injuries are a rare but serious condition, as survival from such injuries relies on proper airway management. As a result, recommendations for management have been based on small case reports and expert opinion. We reviewed our last 10 years of experience with managing laryngotracheal separation injuries and identified 6 cases for chart review. Awake tracheostomy or videolaryngobronchoscopy was used in each case to initially obtain the airway. Surgical repair was then performed immediately using nonabsorbable monofilament suture or a miniplate, and a low fenestrated tracheostomy was placed. All of our patients who followed up were decannulated, eating regular diets, and had satisfactory voice quality at 3 months postoperatively. Review of the literature revealed that, while management strategies have changed over time, treatment still varies widely depending on surgeon preference and the details of each injury. Outcomes from our series suggest that our described techniques and management strategies can be used with good outcomes. We believe that this is due to securing a safe airway, early surgical intervention with no unnecessary tissue dissection, effective reconstruction of the airway, and the fenestrated tracheostomy technique.

Comparing outcomes for non-metastatic rectal cancer in academic vs. community centers: A propensity-matched analysis of the National Cancer Database

BACKGROUND

Little is known regarding the impact of hospital academic status on outcomes following rectal cancer surgery. We compare these outcomes for nonmetastatic rectal adenocarcinoma at academic versus community institutions.

METHODS

The National Cancer Database (2010-2016) was queried for patients with nonmetastatic rectal adenocarcinoma who underwent resection. Propensity score matching was performed across facility cohorts to balance confounding covariates. Kaplan-Meier estimation and Cox-proportional hazards regression were used to analyze survival, other short and long-term outcomes were analyzed by way of logistic regression.

RESULTS

After matching, 15,096 patients were included per cohort. Academic centers were associated with significantly decreased odds of conversion and positive margins with significantly increased odds of ≥12 regional nodes examined. Academic programs also had decreased odds of 30 and 90-day mortality and decreased 5-year mortality hazard. After matching for facility volume, no significant differences in outcomes between centers was seen.

CONCLUSIONS

No difference between academic and community centers in outcomes following surgery for non-metastatic rectal cancer was seen after matching for facility procedural volume.

O21: GENOME-WIDE ASSOCIATION STUDY OF VARICOSE VEINS IN 810,625 INDIVIDUALS IDENTIFIES 45 GENETIC RISK LOCI

Introduction

Varicose veins (VV) impact a third of the UK adult population; 10% of patients develop lipodermatosclerosis and ulceration. VV often requires surgical management, however, there is a high-risk of recurrence. VV is a complex disease, where genetic and non-genetic components contribute to overall phenotypic expression. The genetic architecture of VV is poorly understood; we aimed to uncover its genetic basis.

Method

We conducted hitherto the largest genome-wide association study of VV. In stage one, using UK Biobank, we compared 22,473 VV patients and 379,183 controls. In stage two, replication and meta-analysis were performed in an independent cohort of 113,041 VV cases and 295,928 controls from 23&Me (California, USA). In-silico analysis was conducted in FUMA, MAGMA, and XGR.

Result

109 genome-wide significant (P≤ 5×10-8) loci were identified in UK Biobank, 45 of which successfully replicated in the 23&Me cohort. Twenty-seven loci have not been previously reported. FUMA positionally-mapped 128 genes at the replicated loci, with 84 having a combined annotation-dependent depletion score (CADD) &gt;12.37, suggesting functional, deleterious variants. MAGMA analysis implicated pathways involved in cardiovascular system development (P=1.57×10-08) and tube morphogenesis (P=9.35×10-08). Furthermore, XGR revealed enriched pathways in downstream signalling in naive CD8+ T cells (P=0.0017), and encoding structural and core extracellular glycoproteins (both P=0.007).

Conclusion

We identified 45 variants conferring risk of VV, which provide insights into disease biology. Implicated genes are enriched in pathways involved in vascular development, immune cell activity and extracellular matrix function, and provide new targets for therapeutic development.

Take-home message

Unravelling the genetic architecture of varicose veins may facilitate our understanding of the disease and guide therapeutic approaches.

Loss odyssey in medical image segmentation

The loss function is an important component in deep learning-based segmentation methods. Over the past five years, many loss functions have been proposed for various segmentation tasks. However, a systematic study of the utility of these loss functions is missing. In this paper, we present a comprehensive review of segmentation loss functions in an organized manner. We also conduct the first large-scale analysis of 20 general loss functions on four typical 3D segmentation tasks involving six public datasets from 10+ medical centers. The results show that none of the losses can consistently achieve the best performance on the four segmentation tasks, but compound loss functions (e.g. Dice with TopK loss, focal loss, Hausdorff distance loss, and boundary loss) are the most robust losses. Our code and segmentation results are publicly available and can serve as a loss function benchmark. We hope this work will also provide insights on new loss function development for the community.

The impact of surgical approach on short- and long-term outcomes after rectal cancer resection in elderly patients: a national cancer database propensity score matched comparison of robotic, laparoscopic, and open approaches

BACKGROUND

Elderly patients are underrepresented in studies demonstrating the advantages of laparoscopy for the management of colorectal diseases. Moreover, few studies have examined the robotic approach in this population. In this retrospective analysis, we compare outcomes for open, laparoscopic, and robotic approaches in elderly patients with nonmetastatic rectal cancer.

METHODS

The U.S. National Cancer Database was queried for patients aged ≥ 65 with nonmetastatic adenocarcinoma of the rectum who underwent surgical resection from 2010 to 2016. Groups were separated based on approach (open, laparoscopic, robotic). One-to-one nearest neighbor propensity score matching (PSM) ± 1% caliper was performed across surgical approach cohorts to balance potential confounding covariates. Kaplan-Meier estimation and Cox-proportional hazards regression were used to analyze the primary outcome of survival. Secondary outcomes were analyzed by way of logistic regression.

RESULTS

Inclusion criteria and PSM identified 1891 patients per approach (n = 5673). PSM provided adequate discrimination between cohorts (0.6 < AUC < 0.8), and potential confounding covariates did not significantly differ (respective P > 0.05). After PSM, robotic and laparoscopic approaches were associated with decreased odds of 90 day mortality compared to the open approach (P < 0.05). Compared to laparoscopy, a robotic approach was associated with increased odds of ≥ 12 regional lymph nodes examined and negative circumferential resection margin (P < 0.05). No differences were seen in 30 day or 90 day mortality between robotic and laparoscopic approaches. Cox proportional hazards regression showed that both robotic and laparoscopic approaches were significantly associated with decreased mortality hazards relative to open.

CONCLUSION

Our study demonstrates that in elderly patients, minimally invasive surgery for rectal adenocarcinoma was associated with equivalent or improved short- and long-term mortality over open surgery. Compared to laparoscopy, the robotic approach showed no survival disadvantage and greater odds of an appropriate oncological resection. Our study adds evidence to the conclusion that robotic rectal surgery can be safely performed in patients regardless of age.

Six-month and one-year outcomes with the PASCAL transcatheter valve repair system for patients with mitral regurgitation from the multicentre, prospective CLASP study

Background

Transcatheter mitral valve repair has emerged as a viable option for treating mitral regurgitation (MR). We report results from the multicentre, prospective, single arm CLASP study with the PASCAL transcatheter valve repair system.

Methods

109 patients with clinically significant MR deemed candidates for transcatheter repair by the local heart team were treated in the CLASP study. The study evaluated safety, performance, clinical and echocardiographic outcomes and included an independent clinical events committee and echocardiographic core lab. The primary safety endpoint was a composite MAE rate at 30 days of cardiovascular mortality, stroke, MI, new need for renal replacement therapy, severe bleeding, and re-intervention for study device-related complications.

Results

Mean age was 76 years, 54% male, 57% NYHA Class III/IV, 100% MR grade ≥3+ with 62% functional, 31% degenerative, 7% mixed etiology. Successful implantation was achieved in 95% of patients. At 30 days, the MAE rate was 8.3% including one cardiovascular mortality due to cardiogenic shock as a result of severe bleeding at the contralateral arterial access site for hemodynamic monitoring further complicated by disseminated intravascular coagulation, one stroke, and one conversion to mitral valve replacement surgery. In paired analysis, 88% of patients were in NYHA Class I/II (p&lt;0.001), MR grade was ≤1+ in 79% of patients and ≤2+ in 96% of patients. Significant improvements in 6MWD (+27 m, p&lt;0.001) and KCCQ (+16 points, p&lt;0.001) were observed. The six-month data will be available for presentation. In addition, we report one-year follow up of the first 62 patients (ITT): 93% one-year survival rate (Kaplan-Meier estimate), no stroke, no late reintervention, one late MI. In paired analysis, MR grade was ≤1+ in 82% of patients and ≤2+ in 100% of patients. 88% of patients were in NYHA Class I/II (p&lt;0.001), 6MWD improved by 21 m (p=0.124) and KCCQ improved by 13 points (p&lt;0.001).

Conclusions

This study demonstrates the PASCAL transcatheter valve repair system is safe and resulted in robust MR reduction with 100% of patients achieving MR ≤2+, and ∼ 80% MR ≤1+, sustained at one year. Results show high survival and low complication rates, and sustained improvements in functional status, exercise capacity, and quality of life at one year in patients with clinically significant, symptomatic MR. The CLASP IID/IIF pivotal trial is underway.

Funding Acknowledgement

Type of funding source: Private company. Main funding source(s): Edwards Lifesciences (Irvine, CA)

Size of low wall shear stress areas in coronary arteries is related to epicardial physiology and not to microcirculatory resistance

Background

Wall shear stress (WSS) plays an important role in coronary atherosclerosis. Low WSS is associated with inflammation, endothelial dysfunction, and progression of atherosclerosis; while high WSS leads to vulnerable plaque transformation and future myocardial infarction. Defining the relationship between WSS and the currently available coronary physiology indices would provide valuable insights into potential mechanisms for predicting future adverse cardiac events.

Purpose

To investigate the relationship between WSS and the coronary epicardial/microvascular physiology indices.

Methods

Patients undergoing coronary angiography and physiology testing were prospectively recruited. Physiology measurements were performed under resting and hyperaemic conditions using a pressure/temperature sensor guidewire. Fractional flow reserve (FFR), non-hyperaemic pressure ratios (NHPRs), coronary flow reserve (CFR), and corrected index of microcirculatory resistance (IMR) were measured. The NHPRs including resting full-cycle ratio (RFR), resting distal/aortic pressure (Pd/Pa), instantaneous wave-free ratio (iFR), diastolic pressure ratio (dPR) and diastolic hyperaemia-free ratio (DFR) were derived offline by a blinded expert reader. Values of FFR ≤0.80, NHPR ≤0.89, CFR &lt;2, or IMR &gt;25 were considered ischaemic/abnormal. Computational fluid dynamics analysis was performed and fluid motion equations were solved using finite-volume based software. The inlet and outlet boundary conditions were set to the patient-specific Pa and Pd respectively. The lesion WSS, mean WSS in five segments (WSSupstream = 5mm proximal to lesion, WSSprox = proximal third of lesion, WSSmid = middle third of lesion, WSSdistal = distal third of lesion, WSSdownstream = 5mm distal to lesion), and the total area of low WSS (defined as &lt;1 Pa) along the entire vessel were calculated (Figure A).

Results

A total of 112 vessels from 93 patients were included in the study. The total area of low WSS was significantly larger in lesions with ischaemic FFR, NHPRs, and CFR values (Figure B), and not significantly different in lesions with abnormal IMR values. There was no significant difference in lesion WSS between groups stratified by all physiology indices. Within the lesion WSS sub-segments, WSSprox was significant higher in ischaemic lesions stratified by normal/abnormal FFR, iFR, and dPR (4.2 vs 3.3 Pa, 4.2 vs 3.3 Pa, 4.3 vs 3.3 Pa respectively, all p=0.04), and not significantly different when stratified by RFR, DFR, Pd/Pa, CFR, and IMR.

Conclusion

Functionally significant coronary lesions classified by the epicardial physiology indices have significantly larger total area of low WSS and higher WSSprox which may explain the higher cardiovascular event rates in patients with ischaemic lesions. There was no significant relationship between WSS and coronary microcirculatory resistance.

Figure 1

Funding Acknowledgement

Type of funding source: Foundation. Main funding source(s): National Heart Foundation of Australia Health Professional Scholarship

A real-world comparison of outcomes between fractional flow reserve-guided versus angiography-guided percutaneous coronary intervention

Background

Fractional flow reserve (FFR)-guided percutaneous coronary intervention (PCI) has been shown to be superior to angiography-guided PCI in randomized controlled studies. However, real-world data on the use and outcomes of FFR-guided PCI remain limited.

Purpose

To investigate the outcomes of patients undergoing FFR-guided PCI compared to angiography-guided PCI in a large, state-wide unselected cohort.

Methods

All patients undergoing PCI between June 2017 and June 2018 recorded by the Centre for Health Record Linkage (CHeReL) were included in the study. The CHeReL database is one of the largest data linkage systems in Australia, capturing health data from ≥97% of all healthcare facilities in the state of New South Wales, which has a population of 7.5 million people. The PCI cohort was stratified into the FFR-guided group when a concomitant FFR procedure was performed, and the angiography-guided group when no FFR was performed. The primary endpoint was a combined endpoint of death or myocardial infarction (MI). Secondary endpoints included all-cause death, cardiovascular (CV) death, and MI.

Results

The cohort comprised 10,304 patients, of which 542 (5%) underwent FFR-guided PCI. There were no significant differences in age, gender, or comorbidities between the two groups. During a mean follow-up of 12±4 months, the FFR-guided PCI group had reduced occurrence of the primary endpoint (3% vs 8%, P&lt;0.001), all-cause death (1% vs 4%, P=0.001), CV death (1% vs 3%, P=0.01), and MI (2% vs 4%, P=0.01) (Figure). Multivariable Cox regression analysis demonstrated FFR-guidance to be an independent predictor of the primary endpoint (hazard ratio [HR] 0.47, 95% confidence interval [CI] 0.28–0.78, P=0.004), after adjusting for age, clinical presentation, comorbidities, and multi-vessel PCI. A sensitivity analysis was performed excluding patients that presented with acute MI, leading to a smaller cohort of 5,850 patients, of which 448 (8%) underwent FFR-guided PCI. FFR-guidance remained an independent predictor of the primary endpoint in this cohort of stable patients (HR 0.36, 95% CI 0.17–0.77, P=0.01).

Conclusion

In this real-world study of patients undergoing PCI, FFR-guidance was associated with improved clinical outcomes, driven by the lower hard endpoint of death or MI. The use of FFR-guided PCI remains limited worldwide, and efforts should be directed to increase adoption of this technique in future.

Figure 1

Funding Acknowledgement

Type of funding source: Foundation. Main funding source(s): National Heart Foundation of Australia Health Professional Scholarship

Sound levels with aural suctioning: Effects of suction size, canal moisture, and distance from the eardrum

OBJECTIVE

To determine sound levels resulting from aural suctioning of the external auditory canal.

METHODS

Unweighted decibels (dB) and A-weighted decibels (dBA) sound pressure level measurements were recorded using a retrotympanic microphone in cadaveric human temporal bones. Sound measurements were made with common otologic suctions, size 3, 5, and 7 French, within the external ear canal at the tympanic membrane, 5, and 10 mm from the tympanic membrane in the dry condition. In the wet condition, the ear canal was filled with fluid and completely suctioned clear to determine sound effects of suctioning liquid from the ear canal.

RESULTS

Sound levels generated from ear canal suctioning ranged from 68.3 to 97 dB and 62.6 to 95.1 dBA. Otologic suctions positioned closer to the tympanic membrane resulted in louder sound levels, but was not statistically significant (P > .05). Using larger diameter suctions generated louder dB and dBA sound levels (P < .001) and the addition of liquid in the ear canal during the suction process generated louder dB and dBA sound levels (P < .001).

CONCLUSIONS

Smaller caliber suction sizes and nonsuctioning techniques should be utilized for in-office aural toilet to reduce noise trauma and patient discomfort.

LEVEL OF EVIDENCE

5.

Thromboelastogram-Guided Resuscitation for Patients with Traumatic Brain Injury on Novel Anticoagulants

Traumatic brain injuries in patients on antithrombotic agents carry significant morbidity. Initial therapy is centered around reversal of these agents. The thromboelastogram (TEG) maps the clotting cascade to guide reversal. A retrospective chart review was conducted for 118 patients presenting with a traumatic brain injury while on antithrombotics. Patients were divided between those who received a TEG on arrival and those who did not. The primary endpoint was overall mortality. Secondary endpoints included blood product utilization, and outcomes associated with specific novel anticoagulants. Mortality in the control group was 20.3 per cent compared with 18.5 per cent in the TEG group ( P = 0.81). For less severe injuries, the control group mortality was 3.8 per cent and the TEG group mortality was 8.7 per cent ( P = 0.64). For more severe injuries, mortality in the control versus TEG groups were 31.6 per cent and 25.8 per cent, respectively ( P = 0.73). Blood product utilization was significantly lower in the TEG group ( P = 0.002). Overall mortality was not significantly different between the groups. However, when stratified by severity of injury, mortality was reduced in the TEG-guided group in severely injured patients. Blood product utilization was significantly reduced with TEG-guided reversal. Trauma centers can improve the utilization of blood products in reversal of antithrombotics with the use of TEG.

Using zebrafish to assess the effect of chronic, early developmental exposure to environmentally relevant concentrations of 5-fluorouracil and leucovorin

Environmental contaminants can deleteriously affect aquatic animals. One such contaminant is 5-fluorouracil (5-FU), a long-prescribed chemotherapeutic drug. Leucovorin (LV) is co-administered with 5-FU, potentiating its effects. Zebrafish (Danio rerio) larvae were reared in ng/L treatments of either 5-FU, LV, or a combined 5-FU/LV mixture for 8 dy. Survival was measured daily and swimming behavior assessed every other day. After 8 dy, larval length was measured, and densitometry of p53-labeled cryostat sections determined the extent of apoptosis. No significant differences in survival or apoptosis were found; larvae in the highest concentrations were largest. Changes in behavior of 5-FU-treated larvae were based on exposure duration; changes in LV-treated larvae were affected by drug concentration and duration. Larvae co-exposed to 5-FU/LV had responses like 5-FU-treated larvae. Overall, early developmental exposure of zebrafish larvae to environmentally-relevant concentrations of 5-FU and LV did not adversely affect survival, growth, and behavior suggesting realistic concentrations are sublethal and non-toxic.

A unique case of a digital tourniquet in ichthyosis vulgaris

We report a unique case of a digital tourniquet in a patient with ichthyosis vulgaris. We have identified no previous case reports documenting the occurrence of a digital tourniquet in patients caused by this condition. Ichthyosis vulgaris is a skin condition which causes increased scaling of the skin and in this case, resulted in the formation of a tourniquet-like circumferential constriction to one of the patient's digits.