Citations

e13552 Background: Real-world data from Electronic Health Records (EHR) have been widely used for patient identification to build study cohorts for clinical research. Traditionally, diagnosis codes in the EHR, such International Classification of Diseases (ICD), are used to identify the target patients. However, the accuracy of this approach is dependent on the accuracy of ICD coding, with potential errors especially for tumor types that are frequent locations for metastases (which may contribute to mis-coding). In this study, we attempted to develop a Machine Learning (ML) based approach on EHR data to improve the accuracy of identification of patients with lung cancer. Methods: We used survey respondents in the Enhanced, EHR-facilitated Cancer Symptom Control (E2C2, NCT03892967) cluster-randomized trial at Mayo Clinic as our initial pan-cancer cohort. E2C2 includes adults receiving Medical Hematology/Oncology care for a solid or liquid tumor at Mayo Clinic. We collected cancer diagnoses from the individually abstracted Mayo Clinic Cancer Registry to annotate the cancer type for the patients. Lung cancer related ICD-9 (162.X) and ICD-10 (C34.X) codes were used to build a search query on the Mayo’s EHR to find target patients from the E2C2 cohort, and to investigate the ICD-based lung cancer patient identification performance. Diagnosis, radiation oncology treatment (CPT 77261 - 77799), and antineoplastic drug administration data were collected from EHR as variables. Logistic Regression (LR), Support vector machine (SVM), Random Forest (RF), and eXtreme Gradient Boosting (XGB) were selected to build models for lung cancer patient identification. 10-fold cross-validation was implemented to assess the models. Precision, Recall, F1 Score, and area under the curve (AUC) were selected to measure the performance. Results: We collected 13,893 patients with a specific cancer diagnosis, and 1,394 were identified as having lung cancer. The identification performance across different methods is shown. The ICD-based method only had 0.65 precision. It means we collected a lot of false positive cases (other cancer but no lung cancer patient), as we mentioned in the background. SVM gets the best precision results, but its recall and F1 score are not good enough. XGB shows the best F1 Score and AUC, which also means this method achieved the best and most balanced performance. Conclusions: In this study, we found that XGB-based methods achieved the best identification performance for lung cancer. In the future, we will investigate if this is also true for the identification of other cancer types. [Table: see text]

Abstract Background Carbapenemase-producing (CP) Escherichia coli (CP-Ec) are a global public health threat. We aimed to describe the clinical and molecular epidemiology and outcomes of patients from several countries with CP-Ec isolates obtained from a prospective cohort. Methods Patients with CP-Ec were enrolled from 26 hospitals in 6 countries. Clinical data were collected, and isolates underwent whole-genome sequencing. Clinical and molecular features and outcomes associated with isolates with or without metallo-β-lactamases (MBLs) were compared. The primary outcome was desirability of outcome ranking (DOOR) at 30 days after the index culture. Results Of the 114 CP-Ec isolates in Consortium on resistance against carbapenems in Klebsiella and other Enterobacterales-2 (CRACKLE-2), 49 harbored an MBL, most commonly blaNDM-5 (38/49, 78%). Strong regional variations were noted with MBL-Ec predominantly found among patients in China (23/49). Clinically, MBL-Ec were more often from urine sources (49% vs 29%), less often met criteria for infection (39% vs 58%, P = .04), and had lower acuity of illness when compared with non–MBL-Ec. Among patients with infection, the probability of a better DOOR outcome for a randomly selected patient with MBL-Ec as compared with non–MBL-Ec was 62% (95% CI: 48.2–74.3%). Among infected patients, non–MBL-Ec had increased 30-day (26% vs 0%; P = .02) and 90-day (39% vs 0%; P = .001) mortality compared with MBL-Ec. Conclusions Emergence of CP-Ec was observed with important geographic variations. Bacterial characteristics, clinical presentations, and outcomes differed between MBL-Ec and non–MBL-Ec. Mortality was higher among non-MBL isolates, which were more frequently isolated from blood, but these findings may be confounded by regional variations.

Objectives:Mucosal injury in celiac disease (CD) patients can be patchy, and up to 12% of CD patients can have mucosal changes limited to the duodenal bulb. Hence, recent guidelines recommend obtaining bulb biopsies in addition to distal duodenum. This study aimed to describe a cohort of children with isolated bulb CD and assess the benefit of separating bulb biopsies.Methods:A retrospective chart review between January 2011 and January 2022 at 2 medical centers was conducted. We included children with CD who underwent endoscopy with separated biopsies from the bulb and distal duodenum. A blinded pathologist performed Marsh‐Oberhuber grading on selected cases.Results:We identified 224 CD patients, of which 33 (15%) had histologically confirmed isolated bulb CD. Patients with isolated bulb CD were older at diagnosis (10 vs 8 years; P = 0.03). Median anti‐tissue transglutaminase immunoglobulin A (TTG IgA) level was lower in isolate bulb CD (2.8 vs 16.7 times the upper limit of normal [ULN], P < 0.001). Almost 88% (29/33) of isolated bulb CD patients had an anti‐TTG IgA value of less than 10 times the ULN. Time to anti‐TTG IgA normalization (mean 14 months) was similar between the 2 groups. A pathologist review of diagnostic biopsies could not distinguish between the bulb and distal duodenum biopsies in approximately one‐third of the reviewed samples.Conclusions:Separating bulb from distal duodenum biopsies can be considered during CD diagnosis, particularly in children with anti‐TTG IgA levels less than 10 times the ULN. Larger prospective cohorts are needed to decide whether isolated bulb CD is a unique cohort or an early stage of the conventional CD.

Over the last decades, epidemiology and functional studies have started to reveal a pivotal role of vitamin D in both type 1 and type 2 diabetes pathogenesis. Acting through the vitamin D receptor (VDR), vitamin D regulates insulin secretion in pancreatic islets and insulin sensitivity in multiple peripheral metabolic organs. In vitro studies and both T1D and T2D animal models showed that vitamin D can improve glucose homeostasis by enhancing insulin secretion, reducing inflammation, reducing autoimmunity, preserving beta cell mass, and sensitizing insulin action. Conversely, vitamin D deficiency has been shown relevant in increasing T1D and T2D incidence. While clinical trials testing the hypothesis that vitamin D improves glycemia in T2D have shown conflicting results, subgroup and meta-analyses support the idea that raising serum vitamin D levels may reduce the progression from prediabetes to T2D. In this review, we summarize current knowledge on the molecular mechanisms of vitamin D in insulin secretion, insulin sensitivity, and immunity, as well as the observational and interventional human studies investigating the use of vitamin D as a treatment for diabetes.

AbstractAn increasing number of studies have reported using natural language processing (NLP) to assist observational research by extracting clinical information from electronic health records (EHRs). Currently, no standardized reporting guidelines for NLP‐assisted observational studies exist. The absence of detailed reporting guidelines may create ambiguity in the use of NLP‐derived content, knowledge gaps in the current research reporting practices, and reproducibility challenges. To address these issues, we conducted a scoping review of NLP‐assisted observational clinical studies and examined their reporting practices, focusing on NLP methodology and evaluation. Through our investigation, we discovered a high variation regarding the reporting practices, such as inconsistent use of references for measurement studies, variation in the reporting location (reference, appendix, and manuscript), and different granularity of NLP methodology and evaluation details. To promote the wide adoption and utilization of NLP solutions in clinical research, we outline several perspectives that align with the six principles released by the World Health Organization (WHO) that guide the ethical use of artificial intelligence for health.

Hexanucleotide G 4 C 2 repeat expansions in the C9orf72 gene are the most common genetic cause of amyotrophic lateral sclerosis and frontotemporal dementia. Dipeptide repeat proteins (DPRs) generated by translation of repeat-containing RNAs show toxic effects in vivo as well as in vitro and are key targets for therapeutic intervention. We generated human antibodies that bind DPRs with high affinity and specificity. Anti-GA antibodies engaged extra- and intra-cellular poly-GA and reduced aggregate formation in a poly-GA overexpressing human cell line. However, antibody treatment in human neuronal cultures synthesizing exogenous poly-GA resulted in the formation of large extracellular immune complexes and did not affect accumulation of intracellular poly-GA aggregates. Treatment with antibodies was also shown to directly alter the morphological and biochemical properties of poly-GA and to shift poly-GA/antibody complexes to more rapidly sedimenting ones. These alterations were not observed with poly-GP and have important implications for accurate measurement of poly-GA levels including the need to evaluate all centrifugation fractions and disrupt the interaction between treatment antibodies and poly-GA by denaturation. Targeting poly-GA and poly-GP in two mouse models expressing G 4 C 2 repeats by systemic antibody delivery for up to 16 mo was well-tolerated and led to measurable brain penetration of antibodies. Long-term treatment with anti-GA antibodies produced improvement in an open-field movement test in aged C9orf72 450 mice. However, chronic administration of anti-GA antibodies in AAV-(G 4 C 2 ) 149 mice was associated with increased levels of poly-GA detected by immunoassay and did not significantly reduce poly-GA aggregates or alleviate disease progression in this model.

Reproducibility is an important quality criterion for the secondary use of electronic health records (EHRs). However, multiple barriers to reproducibility are embedded in the heterogeneous EHR environment. These barriers include complex processes for collecting and organizing EHR data and dynamic multi-level interactions occurring during information use (e.g., inter-personal, inter-system, and cross-institutional). To ensure reproducible use of EHRs, we investigated four information quality dimensions and examine the implications for reproducibility based on a real-world EHR study. Four types of IQ measurements suggested that barriers to reproducibility occurred for all stages of secondary use of EHR data. We discussed our recommendations and emphasized the importance of promoting transparent, high-throughput, and accessible data infrastructures and implementation best practices (e.g., data quality assessment, reporting standard).

Abstract Internal validation is the most popular evaluation strategy used for drug–target predictive models. The simple random shuffling in the cross-validation, however, is not always ideal to handle large, diverse and copious datasets as it could potentially introduce bias. Hence, these predictive models cannot be comprehensively evaluated to provide insight into their general performance on a variety of use-cases (e.g. permutations of different levels of connectiveness and categories in drug and target space, as well as validations based on different data sources). In this work, we introduce a benchmark, BETA, that aims to address this gap by (i) providing an extensive multipartite network consisting of 0.97 million biomedical concepts and 8.5 million associations, in addition to 62 million drug–drug and protein–protein similarities and (ii) presenting evaluation strategies that reflect seven cases (i.e. general, screening with different connectivity, target and drug screening based on categories, searching for specific drugs and targets and drug repurposing for specific diseases), a total of seven Tests (consisting of 344 Tasks in total) across multiple sampling and validation strategies. Six state-of-the-art methods covering two broad input data types (chemical structure- and gene sequence-based and network-based) were tested across all the developed Tasks. The best-worst performing cases have been analyzed to demonstrate the ability of the proposed benchmark to identify limitations of the tested methods for running over the benchmark tasks. The results highlight BETA as a benchmark in the selection of computational strategies for drug repurposing and target discovery.