Continuous glucose monitoring and advanced glycation endproducts for prediction of clinical outcomes and development of cystic fibrosis-related diabetes in adults with CF.

Introductions: Cystic fibrosis-related diabetes (CFRD) is associated with pulmonary decline, compromised nutritional status, and earlier mortality. Onset is often insidious, so screening for early detection of glycemic abnormalities is important. Continuous glucose monitoring (CGM) has been validated in people with CF and has been shown to detect early glycemic variability otherwise missed on 2-hour oral glucose tolerance testing (OGTT). We previously reported that CGM measures of hyperglycemia and glycemic variability are superior to hemoglobin A1c (HbA1c) in distinguishing those with and without CFRD. However, little is known about the long-term predictive value of CGM measures of glycemia for both the development of CFRD and their effect on key clinical outcomes such as weight maintenance and pulmonary function. In addition, there have been no studies investigating advanced glycation endproducts (AGE) assessed by skin autofluorescence in people with CF. Methods: In this prospective observational study, CGM and HbA1c were measured at 2 to 3 time points 3 months apart in 77 adults with CF. Participants who did not have CFRD at the time of enrollment underwent OGTT at the baseline visit, and all participants had AGE readings at baseline. Follow up data including anthropometric measures, pulmonary function and CFRD status were collected by review of medical records 1- and 2-years after the baseline visits. We applied multivariable linear regression models correlating glycemic measures to change in key clinical outcomes (weight, BMI, FEV1) accounting for age, gender and elexacaftor/tezacaftor/ivacaftor (ETI) use. We also conducted logistic regression analyses comparing baseline glycemic data to development of CFRD during the 2-year follow up period. Results: Of the 77 participants, 25 had pre-existing CFRD at the time of enrollment, and six participants were diagnosed with CFRD by the OGTT performed at the baseline visit. When adjusting for age, gender, and ETI use, multiple CGM measures correlated with weight and BMI decline after one year but not after two years. CGM and HbA1c at baseline did not predict decline in FEV1 (p>0.05 for all). In the 46 participants without a diagnosis of CFRD at baseline, two participants were diagnosed with CFRD over the following two years, but CGM measures at baseline did not predict progression to CFRD. Baseline AGE values were higher in individuals with CFRD and correlated with multiple measures of dysglycemia (HbA1c, AG, SD, CV, TIR, % time >140, >180, >250) as well as weight. AGE values also correlated with FEV1 decline at year 1 and weight decline at year 1 and year 2. Conclusions: Several key CGM measures of hyperglycemia and glycemic variability were predictive of future decline in weight and BMI over one year in this population of adults with CF with and without CFRD. None of the baseline glycemic variables predicted progression to CFRD over 2 years. To our knowledge, this is the first report correlating AGE levels with key clinical and glycemic measures in CF. Limitations of these analyses include the small number of participants who developed CFRD (n=2) during the follow up period and the initiation of ETI by many participants, affecting their trajectory in weight and pulmonary function. These results provide additional data supporting the potential role for CGM in identifying clinically significant dysglycemia in CF. Future studies are needed to investigate CGM as a diagnostic and screening tool for CFRD and to understand the implications of AGE measures in this patient population.

Host DNA depletion on frozen human respiratory samples enables successful metagenomic sequencing for microbiome studies.

Background: Most respiratory microbiome studies have focused on amplicon rather than metagenomics sequencing due to high host DNA content. We evaluated efficacy of five host DNA depletion methods on previously frozen human bronchoalveolar lavage (BAL), nasal swabs, and sputum prior to metagenomic sequencing. Results: Median sequencing depth was 76.4 million reads per sample. Untreated nasal, sputum and BAL samples had 94.1%, 99.2%, and 99.7% host-reads. The effect of host depletion differed by sample type. Most treatment methods increased microbial reads, species richness and predicted functional richness; the increase in species and predicted functional richness was mediated by higher effective sequencing depth. For BAL and nasal samples, most methods did not change Morisita-Horn dissimilarity suggesting limited bias introduced by host depletion. Conclusions: Metagenomics sequencing without host depletion will underestimate microbial diversity of most respiratory samples due to shallow effective sequencing depth and is not recommended. Optimal host depletion methods vary by sample type.

Type 2 Diabetes Polygenic Score Predicts the Risk of Glucocorticoid-Induced Hyperglycemia in Patients Without Diabetes.

OBJECTIVE: To assess whether increased genetic risk of type 2 diabetes (T2D) is associated with the development of hyperglycemia after glucocorticoid treatment. RESEARCH DESIGN AND METHODS: We performed a retrospective analysis of individuals with no diagnosis of diabetes who received a glucocorticoid dose of ≥10 mg prednisone. We analyzed the association between hyperglycemia and a T2D global extended polygenic score, which was constructed through a meta-analysis of two published genome-wide association studies. RESULTS: Of 546 individuals who received glucocorticoids, 210 developed hyperglycemia and 336 did not. T2D polygenic score was significantly associated with glucocorticoid-induced hyperglycemia (odds ratio 1.4 per SD of polygenic score; P = 0.038). CONCLUSIONS: Individuals with increased genetic risk of T2D have a higher risk of glucocorticoid-induced hyperglycemia. This finding offers a mechanism for risk stratification as part of a precision approach to medical treatment.

Genome-wide association analysis identifies ancestry-specific genetic variation associated with acute response to metformin and glipizide in SUGAR-MGH.

AIMS/HYPOTHESIS: Characterisation of genetic variation that influences the response to glucose-lowering medications is instrumental to precision medicine for treatment of type 2 diabetes. The Study to Understand the Genetics of the Acute Response to Metformin and Glipizide in Humans (SUGAR-MGH) examined the acute response to metformin and glipizide in order to identify new pharmacogenetic associations for the response to common glucose-lowering medications in individuals at risk of type 2 diabetes. METHODS: One thousand participants at risk for type 2 diabetes from diverse ancestries underwent sequential glipizide and metformin challenges. A genome-wide association study was performed using the Illumina Multi-Ethnic Genotyping Array. Imputation was performed with the TOPMed reference panel. Multiple linear regression using an additive model tested for association between genetic variants and primary endpoints of drug response. In a more focused analysis, we evaluated the influence of 804 unique type 2 diabetes- and glycaemic trait-associated variants on SUGAR-MGH outcomes and performed colocalisation analyses to identify shared genetic signals. RESULTS: Five genome-wide significant variants were associated with metformin or glipizide response. The strongest association was between an African ancestry-specific variant (minor allele frequency [MAFAfr]=0.0283) at rs149403252 and lower fasting glucose at Visit 2 following metformin (p=1.9×10-9); carriers were found to have a 0.94 mmol/l larger decrease in fasting glucose. rs111770298, another African ancestry-specific variant (MAFAfr=0.0536), was associated with a reduced response to metformin (p=2.4×10-8), where carriers had a 0.29 mmol/l increase in fasting glucose compared with non-carriers, who experienced a 0.15 mmol/l decrease. This finding was validated in the Diabetes Prevention Program, where rs111770298 was associated with a worse glycaemic response to metformin: heterozygous carriers had an increase in HbA1c of 0.08% and non-carriers had an HbA1c increase of 0.01% after 1 year of treatment (p=3.3×10-3). We also identified associations between type 2 diabetes-associated variants and glycaemic response, including the type 2 diabetes-protective C allele of rs703972 near ZMIZ1 and increased levels of active glucagon-like peptide 1 (GLP-1) (p=1.6×10-5), supporting the role of alterations in incretin levels in type 2 diabetes pathophysiology. CONCLUSIONS/INTERPRETATION: We present a well-phenotyped, densely genotyped, multi-ancestry resource to study gene-drug interactions, uncover novel variation associated with response to common glucose-lowering medications and provide insight into mechanisms of action of type 2 diabetes-related variation. DATA AVAILABILITY: The complete summary statistics from this study are available at the Common Metabolic Diseases Knowledge Portal ( https://hugeamp.org ) and the GWAS Catalog ( www.ebi.ac.uk/gwas/ , accession IDs: GCST90269867 to GCST90269899).

Classification of neurologic outcomes from medical notes using natural language processing.

Neurologic disability level at hospital discharge is an important outcome in many clinical research studies. Outside of clinical trials, neurologic outcomes must typically be extracted by labor intensive manual review of clinical notes in the electronic health record (EHR). To overcome this challenge, we set out to develop a natural language processing (NLP) approach that automatically reads clinical notes to determine neurologic outcomes, to make it possible to conduct larger scale neurologic outcomes studies. We obtained 7314 notes from 3632 patients hospitalized at two large Boston hospitals between January 2012 and June 2020, including discharge summaries (3485), occupational therapy (1472) and physical therapy (2357) notes. Fourteen clinical experts reviewed notes to assign scores on the Glasgow Outcome Scale (GOS) with 4 classes, namely 'good recovery', 'moderate disability', 'severe disability', and 'death' and on the Modified Rankin Scale (mRS), with 7 classes, namely 'no symptoms', 'no significant disability', 'slight disability', 'moderate disability', 'moderately severe disability', 'severe disability', and 'death'. For 428 patients' notes, 2 experts scored the cases generating interrater reliability estimates for GOS and mRS. After preprocessing and extracting features from the notes, we trained a multiclass logistic regression model using LASSO regularization and 5-fold cross validation for hyperparameter tuning. The model performed well on the test set, achieving a micro average area under the receiver operating characteristic and F-score of 0.94 (95% CI 0.93-0.95) and 0.77 (0.75-0.80) for GOS, and 0.90 (0.89-0.91) and 0.59 (0.57-0.62) for mRS, respectively. Our work demonstrates that an NLP algorithm can accurately assign neurologic outcomes based on free text clinical notes. This algorithm increases the scale of research on neurological outcomes that is possible with EHR data.

Initial Insights into the Genetic Variation Associated with Metformin Treatment Failure in Youth with Type 2 Diabetes.

Metformin is the first-line treatment for type 2 diabetes (T2D) in youth but with limited sustained glycemic response. To identify common variants associated with metformin response, we used a genome-wide approach in 506 youth from the Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) study and examined the relationship between T2D partitioned polygenic scores (pPS), glycemic traits, and metformin response in these youth. Several variants met a suggestive threshold (P < 1 × 10-6), though none including published adult variants reached genome-wide significance. We pursued replication of top nine variants in three cohorts, and rs76195229 in ATRNL1 was associated with worse metformin response in the Metformin Genetics Consortium (n = 7,812), though statistically not being significant after Bonferroni correction (P = 0.06). A higher ß-cell pPS was associated with a lower insulinogenic index (P = 0.02) and C-peptide (P = 0.047) at baseline and higher pPS related to two insulin resistance processes were associated with increased C-peptide at baseline (P = 0.04,0.02). Although pPS were not associated with changes in glycemic traits or metformin response, our results indicate a trend in the association of the ß-cell pPS with reduced ß-cell function over time. Our data show initial evidence for genetic variation associated with metformin response in youth with T2D.

Development and Evaluation of a Natural Language Processing Annotation Tool to Facilitate Phenotyping of Cognitive Status in Electronic Health Records: Diagnostic Study.

BACKGROUND: Electronic health records (EHRs) with large sample sizes and rich information offer great potential for dementia research, but current methods of phenotyping cognitive status are not scalable. OBJECTIVE: The aim of this study was to evaluate whether natural language processing (NLP)-powered semiautomated annotation can improve the speed and interrater reliability of chart reviews for phenotyping cognitive status. METHODS: In this diagnostic study, we developed and evaluated a semiautomated NLP-powered annotation tool (NAT) to facilitate phenotyping of cognitive status. Clinical experts adjudicated the cognitive status of 627 patients at Mass General Brigham (MGB) health care, using NAT or traditional chart reviews. Patient charts contained EHR data from two data sets: (1) records from January 1, 2017, to December 31, 2018, for 100 Medicare beneficiaries from the MGB Accountable Care Organization and (2) records from 2 years prior to COVID-19 diagnosis to the date of COVID-19 diagnosis for 527 MGB patients. All EHR data from the relevant period were extracted; diagnosis codes, medications, and laboratory test values were processed and summarized; clinical notes were processed through an NLP pipeline; and a web tool was developed to present an integrated view of all data. Cognitive status was rated as cognitively normal, cognitively impaired, or undetermined. Assessment time and interrater agreement of NAT compared to manual chart reviews for cognitive status phenotyping was evaluated. RESULTS: NAT adjudication provided higher interrater agreement (Cohen κ=0.89 vs κ=0.80) and significant speed up (time difference mean 1.4, SD 1.3 minutes; P<.001; ratio median 2.2, min-max 0.4-20) over manual chart reviews. There was moderate agreement with manual chart reviews (Cohen κ=0.67). In the cases that exhibited disagreement with manual chart reviews, NAT adjudication was able to produce assessments that had broader clinical consensus due to its integrated view of highlighted relevant information and semiautomated NLP features. CONCLUSIONS: NAT adjudication improves the speed and interrater reliability for phenotyping cognitive status compared to manual chart reviews. This study underscores the potential of an NLP-based clinically adjudicated method to build large-scale dementia research cohorts from EHRs.

Proteomic analysis of cardiometabolic biomarkers and predictive modeling of severe outcomes in patients hospitalized with COVID-19.

BACKGROUND: The high heterogeneity in the symptoms and severity of COVID-19 makes it challenging to identify high-risk patients early in the disease. Cardiometabolic comorbidities have shown strong associations with COVID-19 severity in epidemiologic studies. Cardiometabolic protein biomarkers, therefore, may provide predictive insight regarding which patients are most susceptible to severe illness from COVID-19. METHODS: In plasma samples collected from 343 patients hospitalized with COVID-19 during the first wave of the pandemic, we measured 92 circulating protein biomarkers previously implicated in cardiometabolic disease. We performed proteomic analysis and developed predictive models for severe outcomes. We then used these models to predict the outcomes of out-of-sample patients hospitalized with COVID-19 later in the surge (N = 194). RESULTS: We identified a set of seven protein biomarkers predictive of admission to the intensive care unit and/or death (ICU/death) within 28 days of presentation to care. Two of the biomarkers, ADAMTS13 and VEGFD, were associated with a lower risk of ICU/death. The remaining biomarkers, ACE2, IL-1RA, IL6, KIM1, and CTSL1, were associated with higher risk. When used to predict the outcomes of the future, out-of-sample patients, the predictive models built with these protein biomarkers outperformed all models built from standard clinical data, including known COVID-19 risk factors. CONCLUSIONS: These findings suggest that proteomic profiling can inform the early clinical impression of a patient's likelihood of developing severe COVID-19 outcomes and, ultimately, accelerate the recognition and treatment of high-risk patients.

Proteomic analysis of cardiometabolic biomarkers and predictive modeling of severe outcomes in patients hospitalized with COVID-19.

Background: The high heterogeneity in the symptoms and severity of COVID-19 makes it challenging to identify high-risk patients early in the disease. Cardiometabolic comorbidities have shown strong associations with COVID-19 severity in epidemiologic studies. Cardiometabolic protein biomarkers, therefore, may provide predictive insight regarding which patients are most susceptible to severe illness from COVID-19. Methods: In plasma samples collected from 343 patients hospitalized with COVID-19 during the first wave of the pandemic, we measured 92 circulating protein biomarkers previously implicated in cardiometabolic disease. We performed proteomic analysis and developed predictive models for severe outcomes. We then used these models to predict the outcomes of out-of-sample patients hospitalized with COVID-19 later in the surge (N=194). Results: We identified a set of seven biomarkers predictive of admission to the intensive care unit and/or death (ICU/death) within 28 days of presentation to care. Two of the biomarkers, ADAMTS13 and VEGFD, were associated with a lower risk of ICU/death. The remaining biomarkers, ACE2, IL-1RA, IL6, KIM1, and CTSL1, were associated with higher risk. When used to predict the outcomes of the future, out-of-sample patients, the predictive models built with these biomarkers outperformed all models built from standard clinical data, including known COVID-19 risk factors. Conclusions: These findings suggest that proteomic profiling can inform the early clinical impression of a patient’s likelihood of developing severe COVID-19 outcomes and, ultimately, accelerate the recognition and treatment of high-risk patients.

Immunotherapy-based targeting of MSLN+ activated portal fibroblasts is a strategy for treatment of cholestatic liver fibrosis.

We investigated the role of mesothelin (Msln) and thymocyte differentiation antigen 1 (Thy1) in the activation of fibroblasts across multiple organs and demonstrated that Msln-/- mice are protected from cholestatic fibrosis caused by Mdr2 (multidrug resistance gene 2) deficiency, bleomycin-induced lung fibrosis, and UUO (unilateral urinary obstruction)-induced kidney fibrosis. On the contrary, Thy1-/- mice are more susceptible to fibrosis, suggesting that a Msln-Thy1 signaling complex is critical for tissue fibroblast activation. A similar mechanism was observed in human activated portal fibroblasts (aPFs). Targeting of human MSLN+ aPFs with two anti-MSLN immunotoxins killed fibroblasts engineered to express human mesothelin and reduced collagen deposition in livers of bile duct ligation (BDL)-injured mice. We provide evidence that antimesothelin-based therapy may be a strategy for treatment of parenchymal organ fibrosis.

Prolonged Intubation in Patients With Prior Cerebrovascular Disease and COVID-19.

Objectives: Patients with comorbidities are at increased risk for poor outcomes in COVID-19, yet data on patients with prior neurological disease remains limited. Our objective was to determine the odds of critical illness and duration of mechanical ventilation in patients with prior cerebrovascular disease and COVID-19. Methods: A observational study of 1,128 consecutive adult patients admitted to an academic center in Boston, Massachusetts, and diagnosed with laboratory-confirmed COVID-19. We tested the association between prior cerebrovascular disease and critical illness, defined as mechanical ventilation (MV) or death by day 28, using logistic regression with inverse probability weighting of the propensity score. Among intubated patients, we estimated the cumulative incidence of successful extubation without death over 45 days using competing risk analysis. Results: Of the 1,128 adults with COVID-19, 350 (36%) were critically ill by day 28. The median age of patients was 59 years (SD: 18 years) and 640 (57%) were men. As of June 2nd, 2020, 127 (11%) patients had died. A total of 177 patients (16%) had a prior cerebrovascular disease. Prior cerebrovascular disease was significantly associated with critical illness (OR = 1.54, 95% CI = 1.14-2.07), lower rate of successful extubation (cause-specific HR = 0.57, 95% CI = 0.33-0.98), and increased duration of intubation (restricted mean time difference = 4.02 days, 95% CI = 0.34-10.92) compared to patients without cerebrovascular disease. Interpretation: Prior cerebrovascular disease adversely affects COVID-19 outcomes in hospitalized patients. Further study is required to determine if this subpopulation requires closer monitoring for disease progression during COVID-19.

Cardiometabolic risk factors for COVID-19 susceptibility and severity: A Mendelian randomization analysis.

BACKGROUND: Epidemiological studies report associations of diverse cardiometabolic conditions including obesity with COVID-19 illness, but causality has not been established. We sought to evaluate the associations of 17 cardiometabolic traits with COVID-19 susceptibility and severity using 2-sample Mendelian randomization (MR) analyses. METHODS AND FINDINGS: We selected genetic variants associated with each exposure, including body mass index (BMI), at p < 5 × 10-8 from genome-wide association studies (GWASs). We then calculated inverse-variance-weighted averages of variant-specific estimates using summary statistics for susceptibility and severity from the COVID-19 Host Genetics Initiative GWAS meta-analyses of population-based cohorts and hospital registries comprising individuals with self-reported or genetically inferred European ancestry. Susceptibility was defined as testing positive for COVID-19 and severity was defined as hospitalization with COVID-19 versus population controls (anyone not a case in contributing cohorts). We repeated the analysis for BMI with effect estimates from the UK Biobank and performed pairwise multivariable MR to estimate the direct effects and indirect effects of BMI through obesity-related cardiometabolic diseases. Using p < 0.05/34 tests = 0.0015 to declare statistical significance, we found a nonsignificant association of genetically higher BMI with testing positive for COVID-19 (14,134 COVID-19 cases/1,284,876 controls, p = 0.002; UK Biobank: odds ratio 1.06 [95% CI 1.02, 1.10] per kg/m2; p = 0.004]) and a statistically significant association with higher risk of COVID-19 hospitalization (6,406 hospitalized COVID-19 cases/902,088 controls, p = 4.3 × 10-5; UK Biobank: odds ratio 1.14 [95% CI 1.07, 1.21] per kg/m2, p = 2.1 × 10-5). The implied direct effect of BMI was abolished upon conditioning on the effect on type 2 diabetes, coronary artery disease, stroke, and chronic kidney disease. No other cardiometabolic exposures tested were associated with a higher risk of poorer COVID-19 outcomes. Small study samples and weak genetic instruments could have limited the detection of modest associations, and pleiotropy may have biased effect estimates away from the null. CONCLUSIONS: In this study, we found genetic evidence to support higher BMI as a causal risk factor for COVID-19 susceptibility and severity. These results raise the possibility that obesity could amplify COVID-19 disease burden independently or through its cardiometabolic consequences and suggest that targeting obesity may be a strategy to reduce the risk of severe COVID-19 outcomes.

Classification of the Disposition of Patients Hospitalized with COVID-19: Reading Discharge Summaries Using Natural Language Processing.

BACKGROUND: Medical notes are a rich source of patient data; however, the nature of unstructured text has largely precluded the use of these data for large retrospective analyses. Transforming clinical text into structured data can enable large-scale research studies with electronic health records (EHR) data. Natural language processing (NLP) can be used for text information retrieval, reducing the need for labor-intensive chart review. Here we present an application of NLP to large-scale analysis of medical records at 2 large hospitals for patients hospitalized with COVID-19. OBJECTIVE: Our study goal was to develop an NLP pipeline to classify the discharge disposition (home, inpatient rehabilitation, skilled nursing inpatient facility [SNIF], and death) of patients hospitalized with COVID-19 based on hospital discharge summary notes. METHODS: Text mining and feature engineering were applied to unstructured text from hospital discharge summaries. The study included patients with COVID-19 discharged from 2 hospitals in the Boston, Massachusetts area (Massachusetts General Hospital and Brigham and Women's Hospital) between March 10, 2020, and June 30, 2020. The data were divided into a training set (70%) and hold-out test set (30%). Discharge summaries were represented as bags-of-words consisting of single words (unigrams), bigrams, and trigrams. The number of features was reduced during training by excluding n-grams that occurred in fewer than 10% of discharge summaries, and further reduced using least absolute shrinkage and selection operator (LASSO) regularization while training a multiclass logistic regression model. Model performance was evaluated using the hold-out test set. RESULTS: The study cohort included 1737 adult patients (median age 61 [SD 18] years; 55% men; 45% White and 16% Black; 14% nonsurvivors and 61% discharged home). The model selected 179 from a vocabulary of 1056 engineered features, consisting of combinations of unigrams, bigrams, and trigrams. The top features contributing most to the classification by the model (for each outcome) were the following: "appointments specialty," "home health," and "home care" (home); "intubate" and "ARDS" (inpatient rehabilitation); "service" (SNIF); "brief assessment" and "covid" (death). The model achieved a micro-average area under the receiver operating characteristic curve value of 0.98 (95% CI 0.97-0.98) and average precision of 0.81 (95% CI 0.75-0.84) in the testing set for prediction of discharge disposition. CONCLUSIONS: A supervised learning-based NLP approach is able to classify the discharge disposition of patients hospitalized with COVID-19. This approach has the potential to accelerate and increase the scale of research on patients' discharge disposition that is possible with EHR data.

Lung Histopathology in Coronavirus Disease 2019 as Compared With Severe Acute Respiratory Sydrome and H1N1 Influenza: A Systematic Review.

BACKGROUND: Patients with severe coronavirus disease 2019 (COVID-19) have respiratory failure with hypoxemia and acute bilateral pulmonary infiltrates, consistent with ARDS. Respiratory failure in COVID-19 might represent a novel pathologic entity. RESEARCH QUESTION: How does the lung histopathology described in COVID-19 compare with the lung histopathology described in SARS and H1N1 influenza? STUDY DESIGN AND METHODS: We conducted a systematic review to characterize the lung histopathologic features of COVID-19 and compare them against findings of other recent viral pandemics, H1N1 influenza and SARS. We systematically searched MEDLINE and PubMed for studies published up to June 24, 2020, using search terms for COVID-19, H1N1 influenza, and SARS with keywords for pathology, biopsy, and autopsy. Using PRISMA-Individual Participant Data guidelines, our systematic review analysis included 26 articles representing 171 COVID-19 patients; 20 articles representing 287 H1N1 patients; and eight articles representing 64 SARS patients. RESULTS: In COVID-19, acute-phase diffuse alveolar damage (DAD) was reported in 88% of patients, which was similar to the proportion of cases with DAD in both H1N1 (90%) and SARS (98%). Pulmonary microthrombi were reported in 57% of COVID-19 and 58% of SARS patients, as compared with 24% of H1N1 influenza patients. INTERPRETATION: DAD, the histologic correlate of ARDS, is the predominant histopathologic pattern identified in lung pathology from patients with COVID-19, H1N1 influenza, and SARS. Microthrombi were reported more frequently in both patients with COVID-19 and SARS as compared with H1N1 influenza. Future work is needed to validate this histopathologic finding and, if confirmed, elucidate the mechanistic underpinnings and characterize any associations with clinically important outcomes.

Development and Prospective Validation of a Deep Learning Algorithm for Predicting Need for Mechanical Ventilation.

BACKGROUND: Objective and early identification of hospitalized patients, and particularly those with novel coronavirus disease 2019 (COVID-19), who may require mechanical ventilation (MV) may aid in delivering timely treatment. RESEARCH QUESTION: Can a transparent deep learning (DL) model predict the need for MV in hospitalized patients and those with COVID-19 up to 24 h in advance? STUDY DESIGN AND METHODS: We trained and externally validated a transparent DL algorithm to predict the future need for MV in hospitalized patients, including those with COVID-19, using commonly available data in electronic health records. Additionally, commonly used clinical criteria (heart rate, oxygen saturation, respiratory rate, Fio2, and pH) were used to assess future need for MV. Performance of the algorithm was evaluated using the area under receiver operating characteristic curve (AUC), sensitivity, specificity, and positive predictive value. RESULTS: We obtained data from more than 30,000 ICU patients (including more than 700 patients with COVID-19) from two academic medical centers. The performance of the model with a 24-h prediction horizon at the development and validation sites was comparable (AUC, 0.895 vs 0.882, respectively), providing significant improvement over traditional clinical criteria (P < .001). Prospective validation of the algorithm among patients with COVID-19 yielded AUCs in the range of 0.918 to 0.943. INTERPRETATION: A transparent deep learning algorithm improves on traditional clinical criteria to predict the need for MV in hospitalized patients, including in those with COVID-19. Such an algorithm may help clinicians to optimize timing of tracheal intubation, to allocate resources and staff better, and to improve patient care.

CoVA: An Acuity Score for Outpatient Screening that Predicts Coronavirus Disease 2019 Prognosis.

BACKGROUND: We sought to develop an automatable score to predict hospitalization, critical illness, or death for patients at risk for coronavirus disease 2019 (COVID-19) presenting for urgent care. METHODS: We developed the COVID-19 Acuity Score (CoVA) based on a single-center study of adult outpatients seen in respiratory illness clinics or the emergency department. Data were extracted from the Partners Enterprise Data Warehouse, and split into development (n = 9381, 7 March-2 May) and prospective (n = 2205, 3-14 May) cohorts. Outcomes were hospitalization, critical illness (intensive care unit or ventilation), or death within 7 days. Calibration was assessed using the expected-to-observed event ratio (E/O). Discrimination was assessed by area under the receiver operating curve (AUC). RESULTS: In the prospective cohort, 26.1%, 6.3%, and 0.5% of patients experienced hospitalization, critical illness, or death, respectively. CoVA showed excellent performance in prospective validation for hospitalization (expected-to-observed ratio [E/O]: 1.01; AUC: 0.76), for critical illness (E/O: 1.03; AUC: 0.79), and for death (E/O: 1.63; AUC: 0.93). Among 30 predictors, the top 5 were age, diastolic blood pressure, blood oxygen saturation, COVID-19 testing status, and respiratory rate. CONCLUSIONS: CoVA is a prospectively validated automatable score for the outpatient setting to predict adverse events related to COVID-19 infection.

Analysis of Glucocorticoid-Related Genes Reveal CCHCR1 as a New Candidate Gene for Type 2 Diabetes.

Glucocorticoids have multiple therapeutic benefits and are used both for immunosuppression and treatment purposes. Notwithstanding their benefits, glucocorticoid use often leads to hyperglycemia. Owing to the pathophysiologic overlap in glucocorticoid-induced hyperglycemia (GIH) and type 2 diabetes (T2D), we hypothesized that genetic variation in glucocorticoid pathways contributes to T2D risk. To determine the genetic contribution of glucocorticoid action on T2D risk, we conducted multiple genetic studies. First, we performed gene-set enrichment analyses on 3 collated glucocorticoid-related gene sets using publicly available genome-wide association and whole-exome data and demonstrated that genetic variants in glucocorticoid-related genes are associated with T2D and related glycemic traits. To identify which genes are driving this association, we performed gene burden tests using whole-exome sequence data. We identified 20 genes within the glucocorticoid-related gene sets that are nominally enriched for T2D-associated protein-coding variants. The most significant association was found in coding variants in coiled-coil α-helical rod protein 1 (CCHCR1) in the HLA region (P = .001). Further analyses revealed that noncoding variants near CCHCR1 are also associated with T2D at genome-wide significance (P = 7.70 × 10-14), independent of type 1 diabetes HLA risk. Finally, gene expression and colocalization analyses demonstrate that variants associated with increased T2D risk are also associated with decreased expression of CCHCR1 in multiple tissues, implicating this gene as a potential effector transcript at this locus. Our discovery of a genetic link between glucocorticoids and T2D findings support the hypothesis that T2D and GIH may have shared underlying mechanisms.

Cardiometabolic Risk Factors for COVID-19 Susceptibility and Severity: A Mendelian Randomization Analysis.

IMPORTANCE: Early epidemiological studies report associations of diverse cardiometabolic conditions especially body mass index (BMI), with COVID-19 susceptibility and severity, but causality has not been established. Identifying causal risk factors is critical to inform preventive strategies aimed at modifying disease risk. OBJECTIVE: We sought to evaluate the causal associations of cardiometabolic conditions with COVID-19 susceptibility and severity. DESIGN: Two-sample Mendelian Randomization (MR) Study. SETTING: Population-based cohorts that contributed to the genome-wide association study (GWAS) meta-analysis by the COVID-19 Host Genetics Initiative. PARTICIPANTS: Patients hospitalized with COVID-19 diagnosed by RNA PCR, serologic testing, or clinician diagnosis. Population controls defined as anyone who was not a case in the cohorts. Exposures: Selected genetic variants associated with 17 cardiometabolic diseases, including diabetes, coronary artery disease, stroke, chronic kidney disease, and BMI, at p<5 x 10-8 from published largescale GWAS. MAIN OUTCOMES: We performed an inverse-variance weighted averages of variant-specific causal estimates for susceptibility, defined as people who tested positive for COVID-19 vs. population controls, and severity, defined as patients hospitalized with COVID-19 vs. population controls, and repeated the analysis for BMI using effect estimates from UKBB. To estimate direct and indirect causal effects of BMI through obesity-related cardiometabolic diseases, we performed pairwise multivariable MR. We used p<0.05/17 exposure/2 outcomes=0.0015 to declare statistical significance. RESULTS: Genetically increased BMI was causally associated with testing positive for COVID-19 [6,696 cases / 1,073,072 controls; p=6.7 x 10-4, odds ratio and 95% confidence interval 1.08 (1.03, 1.13) per kg/m2] and a higher risk of COVID-19 hospitalization [3,199 cases/897,488 controls; p=8.7 x 10-4, 1.12 (1.04, 1.21) per kg/m2]. In the multivariable MR, the direct effect of BMI was abolished upon conditioning on the effect on type 2 diabetes but persisted when conditioning on the effects on coronary artery disease, stroke, chronic kidney disease, and c-reactive protein. No other cardiometabolic exposures tested were associated with a higher risk of poorer COVID-19 outcomes. CONCLUSIONS AND RELEVANCE: Genetic evidence supports BMI as a causal risk factor for COVID-19 susceptibility and severity. This relationship may be mediated via type 2 diabetes. Obesity may have amplified the disease burden of the COVID-19 pandemic either single-handedly or through its metabolic consequences.

COVID-19 Outpatient Screening: a Prediction Score for Adverse Events.

BACKGROUND: We sought to develop an automatable score to predict hospitalization, critical illness, or death in patients at risk for COVID-19 presenting for urgent care during the Massachusetts outbreak. METHODS: Single-center study of adult outpatients seen in respiratory illness clinics (RICs) or the emergency department (ED), including development (n = 9381, March 7-May 2) and prospective (n = 2205, May 3-14) cohorts. Data was queried from Partners Enterprise Data Warehouse. Outcomes were hospitalization, critical illness or death within 7 days. We developed the COVID-19 Acuity Score (CoVA) using automatically extracted data from the electronic medical record and learning-to-rank ordinal logistic regression modeling. Calibration was assessed using predicted-to-observed event ratio (E/O). Discrimination was assessed by C-statistics (AUC). RESULTS: In the development cohort, 27.3%, 7.2%, and 1.1% of patients experienced hospitalization, critical illness, or death, respectively; and in the prospective cohort, 26.1%, 6.3%, and 0.5%. CoVA showed excellent performance in the development cohort (concurrent validation) for hospitalization (E/O: 1.00, AUC: 0.80); for critical illness (E/O: 1.00, AUC: 0.82); and for death (E/O: 1.00, AUC: 0.87). Performance in the prospective cohort (prospective validation) was similar for hospitalization (E/O: 1.01, AUC: 0.76); for critical illness (E/O 1.03, AUC: 0.79); and for death (E/O: 1.63, AUC=0.93). Among 30 predictors, the top five were age, diastolic blood pressure, blood oxygen saturation, COVID-19 testing status, and respiratory rate. CONCLUSIONS: CoVA is a prospectively validated automatable score to assessing risk for adverse outcomes related to COVID-19 infection in the outpatient setting.