Clinical profiles of Asians with NAFLD: A systematic review and meta-analysis

Introduction: NAFLD is increasingly prevalent in Asia, where people suffer more metabolic comorbidities at a lower body mass index (BMI), suggesting potential differences in their clinical profile. Therefore, we attempted to characterize the clinical profile of Asians with NAFLD via a meta-analytic approach. Methods: We searched Pubmed, EMBASE, and Cochrane databases from January 1, 2000 to January 17, 2019. Two authors independently reviewed and selected 104 articles (2,247,754 persons) that identified NAFLD in Asians and reported relevant data, especially BMI and ALT, and excluded individuals with other liver disease and excessive alcohol consumption. Individual patient-level data were obtained from seven cohorts in Asia to complement meta-analyzed data. Results: Overall, the mean age was 52.07 (95%CI:51.28-52.85) years with those from Southeast Asia (42.66, 95%CI: 32.23-53.11) being significantly younger. The mean BMI was 26.2 kg/m2, higher in moderate-severe vs. mild hepatic steatosis (28.3 vs. 25.7) patients and NFS ≥-1.455 vs. <-1.455 (27.09 vs. 26.02), with 34% having non-obese NAFLD. The mean ALT was 31.74 U/L, higher in NFS <-1.455 vs. ≥-1.455 (33.74 vs. 27.83), though no differences were found by obesity or steatosis severity. The majority of males (85.7%) and females (60.7%) had normal to minimally elevated ALT (1-1.5x 95% ULN). Individual patient-level data analysis (N=7,668) demonstrated similar results. Conclusion: About one-third of Asians with NAFLD were non-obese and the majority did not have markedly elevated ALT. Therefore, abnormal ALT or BMI are not recommended as a criterion for NAFLD screening in this population. Additionally, there were significant differences in the clinical profiles of NAFLD among the different regions of Asia.

CytoSorb Therapy in COVID-19 (CTC) Patients Requiring Extracorporeal Membrane Oxygenation: A Multicenter, Retrospective Registry

Introduction: CytoSorb extracorporeal blood purification therapy received FDA Emergency Use Authorization (EUA) to suppress hyperinflammation in critically ill COVID-19 patients. The multicenter CTC Registry was established to systematically collect patient-level data, outcomes, and utilization patterns of CytoSorb under the EUA.Methods: Patient-level data was entered retrospectively at participating centers. The primary outcome of the registry was ICU mortality. Patient disposition of death, continuing ICU care, or ICU discharge was analyzed up to Day 90 after start of CytoSorb therapy. Demographics, comorbidities, COVID-19 medications, inflammatory biomarkers, and details on CytoSorb use were compared between survivors and non-survivors in the veno-venous extracorporeal membrane oxygenation (ECMO) cohort.Results: Between April 2020 and April 2021, 52 patients received veno-venous ECMO plus CytoSorb therapy at 5 U.S. centers. ICU mortality was 17.3% (9/52) on day 30, 26.9% (14/52) on day 90, and 30.8% (16/52) at final follow-up of 153 days. Survivors had a trend toward lower baseline D-Dimer levels (2.3 ± 2.5 vs. 19.8 ± 32.2 μg/mL, p = 0.056) compared to non-survivors. A logistic regression analysis suggested a borderline association between baseline D-Dimer levels and mortality with a 32% increase in the risk of death per 1 μg/mL increase (p = 0.055). CytoSorb was well-tolerated without any device-related adverse events reported.Conclusions: CytoSorb therapy for critically ill COVID-19 patients on ECMO was associated with high survival rates suggesting potential therapeutic benefit. Elevated baseline D-Dimer levels may suggest increased risk of mortality. Prospective controlled studies are warranted to substantiate these results.Clinical Trial Registration:https://clinicaltrials.gov/ct2/show/NCT0439192, identifier: NCT04391920.

Thrombectomy With and Without Computed Tomography Perfusion Imaging in the Early Time Window: A Pooled Analysis of Patient-Level Data

Background and Purpose: The optimal imaging paradigm for endovascular thrombectomy (EVT) patient selection in early time window (0–6 hours) treated acute ischemic stroke patients remains uncertain. We aimed to compare post-EVT outcomes between patients who underwent prerandomization basic (noncontrast computed tomography [CT], CT angiography only) versus additional advanced imaging (computed tomography perfusion [CTP] imaging) and to determine the association of performance of prerandomization CTP imaging with clinical outcomes. Methods: The HERMES collaboration (Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials) pooled patient-level data from randomized controlled trials comparing EVT with usual care for acute ischemic stroke due to anterior circulation large vessel occlusion. Good functional outcome, defined as modified Rankin Scale score 0 to 2 at 90 days, was compared between randomized patients with and without CTP baseline imaging. Univariable and multivariable binary logistic regression analysis was performed to determine the association of baseline CTP imaging and good functional outcome. Results: We analyzed 1348 patients 610 (45.3%) of whom underwent CTP prerandomization. The benefit of EVT compared with best medical management was maintained irrespective of the baseline imaging paradigm (90-day modified Rankin Scale score 0–2 in EVT versus control patients: with CTP: 46.0% (137/298) versus 28.9% (88/305), without CTP: 44.1% (162/367) versus 27.3% (100/366). Performance of CTP baseline imaging compared with baseline noncontrast CT and CT angiography only yielded similar rates of good outcome (odds ratio, 1.05 [95% CI, 0.82–1.33], adjusted odds ratio, 1.04, [95% CI, 0.80–1.35]). Conclusions: Rates of good functional outcome were similar among patients in whom CTP was or was not performed, and EVT treatment effect in the 0- to 6-hour time window was similar in patients with and without baseline CTP imaging.

Correlation between Previous Antibiotic Exposure and COVID-19 Severity. A Population-Based Cohort Study

We examined the correlation between previous antibiotic exposure and COVID-19 severity using a population-based observational matched cohort study with patient level data obtained for more than 5.8 million people registered in SIDIAP in Catalonia, Spain. We included all patients newly diagnosed with COVID-19 from March to June 2020 and identified all their antibiotic prescriptions in the previous two years. We used a composite severity endpoint, including pneumonia, hospital admission and death due to COVID-19. We examined the influence of high antibiotic exposure (>4 regimens), exposure to highest priority critically important antimicrobials (HPCIA) and recent exposure. Potential confounders were adjusted by logistic regression. A total of 280,679 patients were diagnosed with COVID-19, 146,656 of whom were exposed to at least one antibiotic course (52.3%) during the preceding two years. A total of 25,222 presented severe COVID-19 infection (9%), and the risk of severity was highest among those exposed to antibiotics (OR 1.12; 95% CI: 1.04–1.21). Among all individuals exposed to antibiotics, high, recent and exposure to HPCIAs were correlated with increased COVID severity (OR 1.19; 95% CI: 1.14–1.26; 1.41; 95% CI: 1.36–1.46; and 1.35; 95% CI: 1.30–1.40, respectively). Our findings confirm a significant correlation between previous antibiotic exposure and increased severity of COVID-19 disease.

Single-Agent Ibrutinib Versus Real-World (RW) Treatments for Patients with Previously Untreated Chronic Lymphocytic Leukemia (CLL): Adjusted Comparison of Resonate-2™ with the Cllear and Lyon-Sud Rw Databases

Introduction Ibrutinib, a once-daily oral Bruton's tyrosine kinase inhibitor, has shown progression-free survival (PFS) or overall-survival (OS) benefit over chemoimmunotherapy (CIT) in multiple phase 3 studies in previously untreated patients with CLL, and significantly longer time to next treatment compared with CIT in previously untreated high-risk patients in a RW setting. We conducted an adjusted comparison of ibrutinib versus RW treatment for previously untreated CLL using patient-level data from the phase 3 RESONATE-2™ (NCT01722487) trial and RW databases from 2 countries. Methods Previously untreated patients with CLL, fulfilling RESONATE-2™ eligibility criteria (age ≥ 65, no del17p) were selected from 2 RW data sources containing electronic medical records for patients with CLL: Centre Hospitalier Lyon-Sud, France; CLLEAR CLL registry from 7 academic centers in the Czech Republic. PFS and OS were compared between patients from the ibrutinib arm of RESONATE-2™ with a median follow-up of 60 months, and those receiving physicians' choice (PC) treatment other than ibrutinib from the RW database, adjusting for differences in baseline characteristics including age, gender, del11q, IGHV status, RAI/BINET disease stage, and Eastern Cooperative Oncology Group (ECOG) score. Hazard ratios (HRs) for ibrutinib versus RW PC treatment were estimated using a multivariable Cox proportional hazards model including available baseline characteristics as covariates, and using an inverse probability weighted (IPW) Cox model with average treatment effect for treatment (ATT) weights derived from propensity scores estimated from a logistic regression including the same covariates. Results The analysis included 136 patients from the RESONATE-2™ study receiving ibrutinib and 920 previously untreated RW patients receiving PC treatment (Lyon-Sud n = 162, CLLEAR n = 758). Baseline characteristics were generally balanced between treatment groups. The most common PC regimens contained CIT and were fludarabine + cyclophosphamide + rituximab (FCR) (n = 227), bendamustine + rituximab (BR) (n = 201), and rituximab + chlorambucil (R + Chlor) (n = 116). Older age, male gender, del11q and advanced disease stage were independent risk factors for PFS and OS. When comparing ibrutinib versus the overall PC cohort, the adjusted HR (95% confidence interval [CI]) was 0.24 (0.16-0.34) for PFS and 0.33 (0.21-0.52) for OS (both p &lt; 0.0001). IPW-based comparative estimates were highly consistent for both PFS (HR = 0.27 [0.18-0.39]) and OS (HR = 0.39 [0.24-0.62]). ATT-weighted survival curves estimating PFS and OS for the RESONATE-2™ population as if treated with PC, showed a median value of 32.1 months and 72.5 months, respectively, while median values for Ibrutinib were not reached. When comparing ibrutinib versus FCR, BR, and R + Chlor, the adjusted HRs (95% CI) were 0.26 (0.17-0.38), 0.27 (0.18-0.41), and 0.27 (0.17-0.42), respectively, for PFS and 0.34 (0.20-0.56), 0.28 (0.16-0.49), and 0.61 (0.32-1.13), respectively, for OS (Figure). Concl usions Adjusted comparisons of the RESONATE-2™ trial and RW patient-level data demonstrates significantly improved PFS and OS for ibrutinib versus physician's choice treatment (predominantly CIT) in previously untreated patients with CLL. PFS and OS benefit for ibrutinib was consistent across a range of common regimens: FCR, BR, and R + Chlor. These results are consistent with data from phase 3 studies and support the use of ibrutinib for first-line CLL treatment. Funding Source: Sponsored by Janssen Pharmaceutica NV, and Pharmacyclics LLC, an AbbVie Company. The RW databases are independently owned. Writing assistance was provided by Emma Fulkes and Liqing Xiao of Parexel and funded by Janssen Pharmaceutica NV. Figure 1 Figure 1. Disclosures Doubek: Janssen-Cilag, AbbVie, AstraZeneca, Amgen, Gilead, Novartis: Honoraria, Research Funding. Smolej: AbbVie, AstraZeneca, Gilead, Janssen-Cilag, and Roche: Consultancy, Honoraria, Other: Travel Grants. Šimkovič: AbbVie, AstraZeneca, Janssen-Cilag, Gilead, Roche: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Travel Grants. Lysak: Novartis, Janssen-Cilag, AbbVie; AstraZeneca: Honoraria, Research Funding. Bachy: Kite, a Gilead Company: Honoraria; Novartis: Honoraria; Daiishi: Research Funding; Roche: Consultancy; Takeda: Consultancy; Incyte: Consultancy. Ferrant: AbbVie: Honoraria, Other: Travel, Accommodations, Expenses; AstraZeneca: Honoraria; Janssen: Other: Travel, Accommodations, Expenses. Diels: Janssen: Current Employment. Cabrieto: Janssen: Current Employment. Nielsen: Janssen: Current Employment. Salles: Allogene: Consultancy; Regeneron: Consultancy, Honoraria; Velosbio: Consultancy; Takeda: Consultancy; Rapt: Consultancy; Genentech/Roche: Consultancy; Epizyme: Consultancy, Honoraria; Debiopharm: Consultancy; Genmab: Consultancy; Incyte: Consultancy; Ipsen: Consultancy; Janssen: Consultancy; Kite/Gilead: Consultancy; Loxo: Consultancy; Miltneiy: Consultancy; Morphosys: Consultancy, Honoraria; Novartis: Consultancy; BMS/Celgene: Consultancy; Beigene: Consultancy; Abbvie: Consultancy, Honoraria; Bayer: Honoraria.

1352. Changing Quality Indicators by Monitoring Veterans Using the Sexually TRansmitted Infection Key Evaluation (STRIKE) Dashboard

Background During the COVID-19 pandemic, there have been multiple reports concerning patients falling out of healthcare. The National VA HIV and Hepatitis and Related Conditions (HHRC) has created the Sexually TRansmitted Infection Key Evaluation (STRIKE) Dashboard to help clinicians identify Veterans who need to complete co-testing for sexually transmitted infections (STIs) or human immunodeficiency virus (HIV) and allows providers to document if the Veteran was offered pre-exposure prophylaxis (PrEP). STRIKE Interface Screen Methods A national VA Veteran dataset was generated from data within the Corporate Data Warehouse (CDW) that included all active PLWH. Positive HIV status is evaluated based on positive antibody test and positive confirmatory result or positive viral load lab result. Negative HIV status is evaluated based on a negative antibody test in the past year. Of the 140 sites, 39 participated but only 9 were active throughout the period of October 1, 2020 to March 31, 2021. Active and nonactive participating sites had metrics assessed across the study period at 3 time points: October 1, 2020, January 1, 2021 and April 1, 2021. Sites with at least 48 visits to report across the 6-month QI period were considered active. Patient level data for review Additional patient level data for review Results Multiple sites had scarcity of supplies due to the national shortage of CT/NG Test Kits during COVID-19. To improve access to CT/NG testing, the dashboard suppress the list of Veterans with + syphlis who were not co-tested for CT/NG. Co-testing improved from 60.2% to 77.2% in active sites and from 61.9% to 68.7% for nonactive across the study period. Percent of Veterans with completed HIV testing on or after STI diagnosis in active sites had an upward trend of 2.1% compared to the nonactive which increased 0.6%. Likewise, new diagnosis of STI for those on PrEP increased 2.6% in active and 0.5% increase in nonactive. On average, active sites increased percent of high risk veterans with active PrEP prescriptions by 2%, compared with nonactive that only increased 1%. STRIKE Co Testing and HIV Screen performed This graph shows average of Syphilis Co Testing and HIV screen performed by active and nonactive sites as measured at three different time point. Number of Veterans on PrEP with STIs This graph shows the average percent of veterans on PrEP with STIs between active versus nonactive sites as measured at three time points Active PrEP Prescriptions Across All Sites This graph shows average percent of active PrEP prescriptions for high risk patients between active versus nonactive sites as measured at three time points Conclusion The STRIKE Dashboard efficiently flags Veterans with STI diagnosis who need completion of STI co-testing, including HIV testing and PrEP offer. Active participating facilities who used the STRIKE Dashboard improved STI Co-testing and PrEP prescription in a short 6 month period even during COVID-19. Disclosures All Authors: No reported disclosures

Clinical knowledge extraction via sparse embedding regression (KESER) with multi-center large scale electronic health record data

The increasing availability of electronic health record (EHR) systems has created enormous potential for translational research. However, it is difficult to know all the relevant codes related to a phenotype due to the large number of codes available. Traditional data mining approaches often require the use of patient-level data, which hinders the ability to share data across institutions. In this project, we demonstrate that multi-center large-scale code embeddings can be used to efficiently identify relevant features related to a disease of interest. We constructed large-scale code embeddings for a wide range of codified concepts from EHRs from two large medical centers. We developed knowledge extraction via sparse embedding regression (KESER) for feature selection and integrative network analysis. We evaluated the quality of the code embeddings and assessed the performance of KESER in feature selection for eight diseases. Besides, we developed an integrated clinical knowledge map combining embedding data from both institutions. The features selected by KESER were comprehensive compared to lists of codified data generated by domain experts. Features identified via KESER resulted in comparable performance to those built upon features selected manually or with patient-level data. The knowledge map created using an integrative analysis identified disease-disease and disease-drug pairs more accurately compared to those identified using single institution data. Analysis of code embeddings via KESER can effectively reveal clinical knowledge and infer relatedness among codified concepts. KESER bypasses the need for patient-level data in individual analyses providing a significant advance in enabling multi-center studies using EHR data.

An efficient and accurate distributed learning algorithm for modeling multi-site zero-inflated count outcomes

Clinical research networks (CRNs), made up of multiple healthcare systems each with patient data from several care sites, are beneficial for studying rare outcomes and increasing generalizability of results. While CRNs encourage sharing aggregate data across healthcare systems, individual systems within CRNs often cannot share patient-level data due to privacy regulations, prohibiting multi-site regression which requires an analyst to access all individual patient data pooled together. Meta-analysis is commonly used to model data stored at multiple institutions within a CRN but can result in biased estimation, most notably in rare-event contexts. We present a communication-efficient, privacy-preserving algorithm for modeling multi-site zero-inflated count outcomes within a CRN. Our method, a one-shot distributed algorithm for performing hurdle regression (ODAH), models zero-inflated count data stored in multiple sites without sharing patient-level data across sites, resulting in estimates closely approximating those that would be obtained in a pooled patient-level data analysis. We evaluate our method through extensive simulations and two real-world data applications using electronic health records: examining risk factors associated with pediatric avoidable hospitalization and modeling serious adverse event frequency associated with a colorectal cancer therapy. In simulations, ODAH produced bias less than 0.1% across all settings explored while meta-analysis estimates exhibited bias up to 12.7%, with meta-analysis performing worst in settings with high zero-inflation or low event rates. Across both applied analyses, ODAH estimates had less than 10% bias for 18 of 20 coefficients estimated, while meta-analysis estimates exhibited substantially higher bias. Relative to existing methods for distributed data analysis, ODAH offers a highly accurate, computationally efficient method for modeling multi-site zero-inflated count data.