The prevalence of postacute sequelae of coronavirus disease 2019 in solid organ transplant recipients: Evaluation of risk in the National COVID Cohort Collaborative.

Postacute sequelae after the coronavirus disease (COVID) of 2019 (PASC) is increasingly recognized, although data on solid organ transplant (SOT) recipients (SOTRs) are limited. Using the National COVID Cohort Collaborative, we performed 1:1 propensity score matching (PSM) of all adult SOTR and nonimmunosuppressed/immunocompromised (ISC) patients with acute COVID infection (August 1, 2021 to January 13, 2023) for a subsequent PASC diagnosis using International Classification of Diseases, 10th Revision, Clinical Modification codes. Multivariable logistic regression was used to examine not only the association of SOT status with PASC, but also other patient factors after stratifying by SOT status. Prior to PSM, there were 8769 SOT and 1 576 769 non-ISC patients with acute COVID infection. After PSM, 8756 SOTR and 8756 non-ISC patients were included; 2.2% of SOTR (n = 192) and 1.4% (n = 122) of non-ISC patients developed PASC (P value < .001). In the overall matched cohort, SOT was independently associated with PASC (adjusted odds ratio [aOR], 1.48; 95% confidence interval [CI], 1.09-2.01). Among SOTR, COVID infection severity (aOR, 11.6; 95% CI, 3.93-30.0 for severe vs mild disease), older age (aOR, 1.02; 95% CI, 1.01-1.03 per year), and mycophenolate mofetil use (aOR, 2.04; 95% CI, 1.38-3.05) were each independently associated with PASC. In non-ISC patients, only depression (aOR, 1.96; 95% CI, 1.24-3.07) and COVID infection severity were. In conclusion, PASC occurs more commonly in SOTR than in non-ISC patients, with differences in risk profiles based on SOT status.

Reusable Solid-form Phase-Selective Organogelators for Rapid and Efficient Remediation of Crude Oil Spill.

Phase-selective organogelators (PSOGs) are considered as a prospective tool for their application in oil spill remediation. However, the number of reports on the PSOGs that can be used in powder form for prompt phase-selective gelation of crude oils is still limited. In this study, a series of compounds with l-mandelic acid as the scaffold bearing different amino acid fragments have been prepared. Also, the gelation behaviors and properties of these derivatives toward organic liquids, product oils, and a type of Chinese crude oil were investigated via heating-and-cooling process, stirring, or resting operation. Besides, the micromorphologies of the resulting gels and the driving forces for the gel formation have been studied by scanning electron microscopy, Fourier transform infrared, UV spectroscopy, concentration-dependent 1H NMR, and X-ray diffraction. Particularly, gelator C15-Phe-Mac-Nap was shown to have the capability of congealing the Chinese crude oil selectively from water in powder form with a relatively lower gelator dosage, as compared with the other gelators we reported in the current and previous works. Moreover, gelator C15-Phe-Mac-Nap displayed some advantageous behaviors such as the reusability of gelator, excellent mechanical and chemical stability of the crude oil gels, and nontoxicity of the gelator in the aquatic environment, indicating its great potential application value for marine oil spill remediation.

Controlling false discovery rate for mediator selection in high-dimensional data.

The need to select mediators from a high dimensional data source, such as neuroimaging data and genetic data, arises in much scientific research. In this work, we formulate a multiple-hypothesis testing framework for mediator selection from a high-dimensional candidate set, and propose a method, which extends the recent development in false discovery rate (FDR)-controlled variable selection with knockoff to select mediators with FDR control. We show that the proposed method and algorithm achieved finite sample FDR control. We present extensive simulation results to demonstrate the power and finite sample performance compared with the existing method. Lastly, we demonstrate the method for analyzing the Adolescent Brain Cognitive Development (ABCD) study, in which the proposed method selects several resting-state functional magnetic resonance imaging connectivity markers as mediators for the relationship between adverse childhood events and the crystallized composite score in the NIH toolbox.

Risk factors for severe outcomes of coronavirus disease 2019 through the waves of the pandemic: Comparing patients with and without solid organ transplantation.

BACKGROUND: While coronavirus disease 2019 (COVID-19) is no longer a public health emergency, certain patients remain at risk of severe outcomes. To better understand changing risk profiles, we studied the risk factors for patients with and without solid organ transplantation (SOT) through the various waves of the pandemic. METHODS: Using the National COVID Cohort Collaborative we studied a cohort of adult patients testing positive for COVID-19 between January 1, 2020, and May 2, 2022. We separated the data into waves of COVID-19 as defined by the Centers for Disease Control. In our primary outcome, we used multivariable survival analysis to look at various risk factors for hospitalization in those with and without SOT. RESULTS: A total of 3,570,032 patients were captured. We found an overall risk attenuation of adverse COVID-19-associated outcomes over time. In both non-SOT and SOT populations, diabetes, chronic kidney disease, and congestive heart failure were risk factors for hospitalization. For SOT specifically, longer time periods between transplant and COVID-19 were protective and age was a risk factor. Notably, asthma was not a risk factor for major adverse renal cardiovascular events, hospitalization, or mortality in either group. CONCLUSIONS: Our study provides a longitudinal view of the risks associated with adverse COVID-related outcomes amongst SOT and non-SOT patients, and how these risk factors evolved over time. Our work will help inform providers and policymakers to better target high-risk patients.

Initial experience with an electron FLASH research extension (FLEX) for the Clinac system.

PURPOSE: Radiotherapy delivered at ultra-high-dose-rates (≥40 Gy/s), that is, FLASH, has the potential to effectively widen the therapeutic window and considerably improve the care of cancer patients. The underlying mechanism of the FLASH effect is not well understood, and commercial systems capable of delivering such dose rates are scarce. The purpose of this study was to perform the initial acceptance and commissioning tests of an electron FLASH research product for preclinical studies. METHODS: A linear accelerator (Clinac 23EX) was modified to include a non-clinical FLASH research extension (the Clinac-FLEX system) by Varian, a Siemens Healthineers company (Palo Alto, CA) capable of delivering a 16 MeV electron beam with FLASH and conventional dose rates. The acceptance, commissioning, and dosimetric characterization of the FLEX system was performed using radiochromic film, optically stimulated luminescent dosimeters, and a plane-parallel ionization chamber. A radiation survey was conducted for which the shielding of the pre-existing vault was deemed sufficient. RESULTS: The Clinac-FLEX system is capable of delivering a 16 MeV electron FLASH beam of approximately 1 Gy/pulse at isocenter and reached a maximum dose rate >3.8 Gy/pulse near the upper accessory mount on the linac gantry. The percent depth dose curves of the 16 MeV FLASH and conventional modes for the 10 × 10 cm2 applicator agreed within 0.5 mm at a range of 50% of the maximum dose. Their respective profiles agreed well in terms of flatness but deviated for field sizes >10 × 10 cm2 . The output stability of the FLASH system exhibited a dose deviation of <1%. Preliminary cell studies showed that the FLASH dose rate (180 Gy/s) had much less impact on the cell morphology of 76N breast normal cells compared to the non-FLASH dose rate (18 Gy/s), which induced large-size cells. CONCLUSION: Our studies characterized the non-clinical Clinac-FLEX system as a viable solution to conduct FLASH research that could substantially increase access to ultra-high-dose-rate capabilities for scientists.

Hormone replacement therapy and COVID-19 outcomes in solid organ transplant recipients compared with the general population.

Exogenous estrogen is associated with reduced coronavirus disease (COVID) mortality in nonimmunosuppressed/immunocompromised (non-ISC) postmenopausal females. Here, we examined the association of estrogen or testosterone hormone replacement therapy (HRT) with COVID outcomes in solid organ transplant recipients (SOTRs) compared to non-ISC individuals, given known differences in sex-based risk in these populations. SOTRs ≥45 years old with COVID-19 between April 1, 2020 and July 31, 2022 were identified using the National COVID Cohort Collaborative. The association of HRT use in the last 24 months (exogenous systemic estrogens for females; testosterone for males) with major adverse renal or cardiac events in the 90 days post-COVID diagnosis and other secondary outcomes were examined using multivariable Cox proportional hazards models and logistic regression. We repeated these analyses in a non-ISC control group for comparison. Our study included 1135 SOTRs and 43 383 immunocompetent patients on HRT with COVID-19. In non-ISC, HRT use was associated with lower risk of major adverse renal or cardiac events (adjusted hazard ratio [aHR], 0.61; 95% confidence interval [CI], 0.57-0.65 for females; aHR, 0.70; 95% CI, 0.65-0.77 for males) and all secondary outcomes. In SOTR, HRT reduced the risk of acute kidney injury (aHR, 0.79; 95% CI, 0.63-0.98) and mortality (aHR, 0.49; 95% CI, 0.28-0.85) in males with COVID but not in females. The potentially modifying effects of immunosuppression on the benefits of HRT requires further investigation.

False discovery rate-controlled multiple testing for union null hypotheses: a knockoff-based approach.

False discovery rate (FDR) controlling procedures provide important statistical guarantees for replicability in signal identification based on multiple hypotheses testing. In many fields of study, FDR controling procedures are used in high-dimensional (HD) analyses to discover features that are truly associated with the outcome. In some recent applications, data on the same set of candidate features are independently collected in multiple different studies. For example, gene expression data are collected at different facilities and with different cohorts, to identify the genetic biomarkers of multiple types of cancers. These studies provide us with opportunities to identify signals by considering information from different sources (with potential heterogeneity) jointly. This paper is about how to provide FDR control guarantees for the tests of union null hypotheses of conditional independence. We present a knockoff-based variable selection method (Simultaneous knockoffs) to identify mutual signals from multiple independent datasets, providing exact FDR control guarantees under finite sample settings. This method can work with very general model settings and test statistics. We demonstrate the performance of this method with extensive numerical studies and two real-data examples.

Sex and organ-specific risk of major adverse renal or cardiac events in solid organ transplant recipients with COVID-19.

While older males are at the highest risk for poor coronavirus disease 2019 (COVID-19) outcomes, it is not known if this applies to the immunosuppressed recipient of a solid organ transplant (SOT), nor how the type of allograft transplanted may impact outcomes. In a cohort study of adult (>18 years) patients testing positive for COVID-19 (January 1, 2020-June 21, 2021) from 56 sites across the United States identified using the National COVID Cohort Collaborative (N3C) Enclave, we used multivariable Cox proportional hazards models to assess time to MARCE after COVID-19 diagnosis in those with and without SOT. We examined the exposure of age-stratified recipient sex overall and separately in kidney, liver, lung, and heart transplant recipients. 3996 (36.4%) SOT and 91 646 (4.8%) non-SOT patients developed MARCE. Risk of post-COVID outcomes differed by transplant allograft type with heart and kidney recipients at highest risk. Males with SOT were at increased risk of MARCE, but to a lesser degree than the non-SOT cohort (HR 0.89, 95% CI 0.81-0.98 for SOT and HR 0.61, 95% CI 0.60-0.62 for non-SOT [females vs. males]). This represents the largest COVID-19 SOT cohort to date and the first-time sex-age-stratified and allograft-specific COVID-19 outcomes have been explored in those with SOT.

The risk and consequences of breakthrough SARS-CoV-2 infection in solid organ transplant recipients relative to non-immunosuppressed controls.

Clinical outcomes in solid organ transplant (SOT) recipients with breakthrough COVID (BTCo) after two doses of mRNA vaccination compared to the non-immunocompromised/immunosuppressed (ISC) general population, are not well described. In a cohort of adult patients testing positive for COVID-19 between December 10, 2020 and April 4, 2022, we compared the cumulative incidence of BTCo in a non-ISC population to SOT recipients (overall and by organ type) using the National COVID Cohort Collaborative (N3C) including data from 36 sites across the United States. We assessed the risk of complications post-BTCo in vaccinated SOT recipients versus SOT with unconfirmed vaccination status (UVS) using multivariable Cox proportional hazards and logistic regression. BTCo occurred in 4776 vaccinated SOT recipients over a median of 149 days (IQR 99-233), with the highest cumulative incidence in heart recipients. The relative risk of BTCo was greatest in SOT recipients (relative to non-ISC) during the pre-Delta period (HR 2.35, 95% CI 1.80-3.08). The greatest relative benefit with vaccination for both non-ISC and SOT cohorts was in BTCo mortality (HR 0.37, 95% CI 0.36-0.39 for non-ISC; HR 0.67, 95% 0.57-0.78 for SOT relative to UVS). While the relative benefit of vaccine was less in SOT than non-ISC, SOT patients still exhibited significant benefit with vaccination.

Instrumental variable with competing risk model.

In this paper, we discuss causal inference on the efficacy of a treatment or medication on a time-to-event outcome with competing risks. Although the treatment group can be randomized, there can be confoundings between the compliance and the outcome. Unmeasured confoundings may exist even after adjustment for measured covariates. Instrumental variable methods are commonly used to yield consistent estimations of causal parameters in the presence of unmeasured confoundings. On the basis of a semiparametric additive hazard model for the subdistribution hazard, we propose an instrumental variable estimator to yield consistent estimation of efficacy in the presence of unmeasured confoundings for competing risk settings. We derived the asymptotic properties for the proposed estimator. The estimator is shown to be well performed under finite sample size according to simulation results. We applied our method to a real transplant data example and showed that the unmeasured confoundings lead to significant bias in the estimation of the effect (about 50% attenuated). Copyright © 2017 John Wiley & Sons, Ltd.

Enterococcus faecium HDRsEf1 Protects the Intestinal Epithelium and Attenuates ETEC-Induced IL-8 Secretion in Enterocytes.

The probiotic Enterococcus faecium HDRsEf1 (Ef1) has been shown to have positive effects on piglet diarrhoea, but the mechanism has not yet been elucidated. In this study, using the IPEC-J2 cell line to mimic intestinal epithelial cells and enterotoxigenic Escherichia coli (ETEC) K88ac as a representative intestinal pathogen, the mechanism underlying Ef1 protection against an enteropathogen was investigated. The results demonstrated that Ef1 was effective in displacing K88ac from the IPEC-J2 cell layer. Moreover, Ef1 and its cell-free supernatant (S-Ef1) modulate IL-8 released by IPEC-J2 cells. Ef1 and its cell-free supernatant showed the potential to protect enterocytes from an acute inflammatory response. In addition, Ef1 and its cell-free supernatant increased the transepithelial electrical resistance (TEER) of the enterocyte monolayer, thus strengthening the intestinal barrier against ETEC. These results may contribute to the development of therapeutic interventions using Ef1 in intestinal disorders of piglets.

Inhibition of Mycobacterium tuberculosis transaminase BioA by aryl hydrazines and hydrazides.

7,8-Diaminopelargonic acid synthase (BioA) of Mycobacterium tuberculosis is a recently validated target for therapeutic intervention in the treatment of tuberculosis (TB). Using biophysical fragment screening and structural characterization of compounds, we have identified a potent aryl hydrazine inhibitor of BioA that reversibly modifies the pyridoxal-5'-phosphate (PLP) cofactor, forming a stable quinonoid. Analogous hydrazides also form covalent adducts that can be observed crystallographically but are incapable of inactivating the enzyme. In the X-ray crystal structures, small molecules induce unexpected conformational remodeling in the substrate binding site. We compared these conformational changes to those induced upon binding of the substrate (7-keto-8-aminopelargonic acid), and characterized the inhibition kinetics and the X-ray crystal structures of BioA with the hydrazine compound and analogues to unveil the mechanism of this reversible covalent modification.

USING SIMULTANEOUS REGRESSION CALIBRATION TO STUDY THE EFFECT OF MULTIPLE ERROR-PRONE EXPOSURES ON DISEASE RISK UTILIZING BIOMARKERS DEVELOPED FROM A CONTROLLED FEEDING STUDY.

Systematic measurement error in self-reported data creates important challenges in association studies between dietary intakes and chronic disease risks, especially when multiple dietary components are studied jointly. The joint regression calibration method has been developed for measurement error correction when objectively measured biomarkers are available for all dietary components of interest. Unfortunately, objectively measured biomarkers are only available for very few dietary components, which limits the application of the joint regression calibration method. Recently, for single dietary components, controlled feeding studies have been performed to develop new biomarkers for many more dietary components. However, it is unclear whether the biomarkers separately developed for single dietary components are valid for joint calibration. In this paper, we show that biomarkers developed for single dietary components cannot be used for joint regression calibration. We propose new methods to utilize controlled feeding studies to develop valid biomarkers for joint regression calibration to estimate the association between multiple dietary components simultaneously with the disease of interest. Asymptotic distribution theory for the proposed estimators is derived. Extensive simulations are performed to study the finite sample performance of the proposed estimators. We apply our methods to examine the joint effects of sodium and potassium intakes on cardiovascular disease incidence using the Women's Health Initiative cohort data. We identify positive associations between sodium intake and cardiovascular diseases as well as negative associations between potassium intake and cardiovascular disease.

Genetic and environmental influence on alcohol intent and alcohol sips among U.S. children-Effects across sex, race, and ethnicity.

INTRODUCTION: Alcohol intent (the susceptibility to initiating alcohol use) and alcohol sips (the initiation of alcohol) in youth are a multifactorial puzzle with many components. This research aims to examine the connection between genetic and environmental factors across sex, race and ethnicity. METHODS: Data was obtained from the twin hub of the Adolescent Brain Cognitive Development (ABCD) study at baseline (2016-2018). Variance component models were conducted to dissect the additive genetic (A), common (C) and unique environmental (E) effects on alcohol traits. The proportion of the total alcohol phenotypic variation attributable to additive genetic factors is reported as heritability (h2). RESULTS: The sample (n = 1,772) included an approximately equal male-female distribution. The 886 same-sex twin pairs were 60.4% dizygotic (DZ), 39.6% monozygotic (MZ), 65.4% non-Hispanic Whites, 13.9% non-Hispanic Blacks, 10.8% of Hispanics with a mean age of 121.2 months. Overall, genetic predisposition was moderate for alcohol intent (h2 = 28%, p = .006) and low for alcohol initiation (h2 = 4%, p = 0.83). Hispanics (h2 = 53%, p < .0001) and Blacks (h2 = 48%, p < .0001) demonstrated higher alcohol intent due to additive genetic factors than Whites (h2 = 34%, p < .0001). Common environmental factors explained more variation in alcohol sips in females (c2 = 63%, p = .001) than in males (c2 = 55%, p = .003). Unique environmental factors largely attributed to alcohol intent, while common environmental factors explained the substantial variation in alcohol initiation. CONCLUSION: Sex and racial/ethnic disparities in genetic and environmental risk factors for susceptibility to alcohol initiation can lead to significant health disparities. Certain populations may be at greater risk for alcohol use due to their genetic and ecological factors at an early age.

[Distribution Characteristics and Risk Assessment of Organophosphates in Water and Sediment in Dongting Lake].

Organophosphates （OPEs） are widely used as flame retardants and additives and thus are commonly detected in the environment. In order to explore their environmental behavior, the concentrations of 13 OPEs in the surface water and sediment of Dongting Lake were analyzed using UPLC-MS/MS. The results showed that 11 OPEs were detected, with detection frequencies of 5.26%-100% and 58.3%-100%, and the concentrations of OPEs were 2.06-2 028 ng·L-1 and 19.6-2 232 ng·g-1 in water and sediment, respectively. Overall, contamination concentrations were ranked in descending order as follows： inflowing rivers, lake area, and outlet, whereas the spatial distribution of concentrations in sediment was inversely proportional to hydrodynamics. The concentration of OPEs in Dongting Lake was at a high level compared with that of domestic and foreign lakes. Among the detected 11 OPEs, tri-iso-butyl phosphate （TnBP） and （TiBP） were dominant in water, accounting for 52.3% and 22.4% of ∑OPEs, respectively. TPhP was the dominant OPEs in sediment, accounting for 31.2% of ∑OPEs. The correlation and principal component analysis indicated that OPEs pollution in Dongting Lake was mainly affected by industrial production emissions, fishery aquaculture, and atmospheric deposition. The assessment results of the risk entropy showed that most of the detected OPEs in water had relatively low ecological risks, whereas the ecological risk of 2-ethylhexyl diphenyl phosphate （EHDPP） at some sampling points requires further attention.

Chronic Lung Disease as a Risk Factor for Long COVID in Patients Diagnosed With Coronavirus Disease 2019: A Retrospective Cohort Study.

Background: Patients with coronavirus disease 2019 (COVID-19) often experience persistent symptoms, known as postacute sequelae of COVID-19 or long COVID, after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. Chronic lung disease (CLD) has been identified in small-scale studies as a potential risk factor for long COVID. Methods: This large-scale retrospective cohort study using the National COVID Cohort Collaborative data evaluated the link between CLD and long COVID over 6 months after acute SARS-CoV-2 infection. We included adults (aged ≥18 years) who tested positive for SARS-CoV-2 during any of 3 SARS-CoV-2 variant periods and used logistic regression to determine the association, considering a comprehensive list of potential confounding factors, including demographics, comorbidities, socioeconomic conditions, geographical influences, and medication. Results: Of 1 206 021 patients, 1.2% were diagnosed with long COVID. A significant association was found between preexisting CLD and long COVID (adjusted odds ratio [aOR], 1.36). Preexisting obesity and depression were also associated with increased long COVID risk (aOR, 1.32 for obesity and 1.29 for depression) as well as demographic factors including female sex (aOR, 1.09) and older age (aOR, 1.79 for age group 40-65 [vs 18-39] years and 1.56 for >65 [vs 18-39] years). Conclusions: CLD is associated with higher odds of developing long COVID within 6 months after acute SARS-CoV-2 infection. These data have implications for identifying high-risk patients and developing interventions for long COVID in patients with CLD.

A novel nanoplatform-based circCSNK1G3 affects CBX7 protein and promotes glioma cell growth.

Bioenvironmental and biological factors have the potential to contribute to the development of glioma, a type of brain tumor. Recent studies have suggested that a unique circular RNA called circCSNK1G3 could play a role in promoting the growth of glioma cells. It does this by stabilizing a specific microRNA called miR-181 and reducing the expression of a tumor-suppressor gene known as chromobox protein homolog 7 (CBX7). To further investigate circCSNK1G3 and its effects on glioma, we utilized a nanoplatform called adeno-associated virus (AAV)-RNAi.To explore the functional implications of circCSNK1G3, we employed siRNA to silence its expression. Along with these effects, the silencing of circCSNK1G3 led to a depletion of miR-181d and an upregulation of CBX7. When we introduced miR-181d mimics, which artificially increase the levels of miR-181d, the anti-glioma cell activity induced by circCSNK1G3 siRNA was almost completely reversed. Conversely, inhibiting miR-181d mimicked the effects of circCSNK1G3 silencing. Moreover, when we overexpressed circCSNK1G3 in glioma cells, we observed an elevation of miR-181d and a depletion of CBX7. We found that the growth of A172 xenografts (tumors) carrying circCSNK1G3 shRNA was significantly inhibited. In these xenograft tissues, we detected a depletion of circCSNK1G3 and miR-181d, as well as an upregulation of CBX7.

Description of two new species of the leafhopper genus Pediopsis Burmeister (Hemiptera, Cicadellidae, Eurymelinae, Macropsini) from China.

Two new leafhopper species of Pediopsis Burmeister, Pediopsisalbopicta Li & Dai, sp. nov. from Hunan and Guizhou provinces of central China and Pediopsispianmaensis Li & Dai, sp. nov. from Yunnan Province of southwestern China, are described and illustrated. Ambiguity in the original description of P.bannaensis Yang & Zhang is discussed, and figures of the female holotype of P.femorata Hamilton are provided for the first time. A checklist and key to Chinese species of Pediopsis are also given.

On High-Dimensional Covariate Adjustment for Estimating Causal Effects in Randomized Trials with Survival Outcomes.

The purpose of this work is to improve the efficiency in estimating the average causal effect (ACE) on the survival scale where right-censoring exists and high-dimensional covariate information is available. We propose new estimators using regularized survival regression and survival Random Forest (RF) to adjust for the high-dimensional covariate to improve efficiency. We study the behavior of the adjusted estimators under mild assumptions and show theoretical guarantees that the proposed estimators are more efficient than the unadjusted ones asymptotically when using RF for the adjustment. In addition, these adjusted estimators are n - consistent and asymptotically normally distributed. The finite sample behavior of our methods is studied by simulation. The simulation results are in agreement with the theoretical results. We also illustrate our methods by analyzing the real data from transplant research to identify the relative effectiveness of identical sibling donors compared to unrelated donors with the adjustment of cytogenetic abnormalities.

A Flexible Ensemble Learning Method for Survival Extrapolation.

OBJECTIVES: Survival extrapolation is an important statistical concept for estimating long-term survival from short-term clinical trial data. It is widely used in health technology assessment (HTA). Survival extrapolation is often performed by fitting one or two parametric models selected based on experience or selecting a model based on some goodness of fit statistics from a predefined collection of models. The main challenge in survival extrapolation is that the result is sensitive to model misspecification. In this study, we aim to propose a new approach that has a robust performance for survival extrapolation. METHODS: We propose a new method called Ensemble Learning for Survival Extrapolation (ELSE). Instead of selecting one best model from a predefined collection, ELSE builds an ensemble model based on a collection of models from the model library. Under this framework, we construct a point estimate of the long-term survival with a weighted average of the estimates of all candidate models and derive confidence intervals using nonparametric bootstrap. RESULTS: With our extensive numerical simulation studies, the proposed ELSE method shows better performance than the traditionally used model selection procedure based on Akaike Information Criterion (AIC). With a real data application to the Therapeutically Applicable Research to Generate Effective Treatment Wilms Tumor project (TARGET-WT) data, the ELSE method produces better survival extrapolation results in point estimate accuracy and confidence interval coverage. CONCLUSIONS: We developed an ensemble learning method for survival extrapolation (ELSE) which is robust for the underline data model and has good real data performance.