Apixaban following discharge in hospitalised adults with COVID-19: Preliminary results from a multicentre, open-label, randomised controlled platform clinical trial. (preprint)

Background The role of thromboprophylaxis in the post-acute phase of COVID-19 is uncertain due to conflicting results from randomised controlled trials and observational studies. We aimed to determine the effectiveness of post-hospital apixaban in reducing the rate of death and hospital readmission of hospitalised adults with COVID-19. Methods HEAL COVID is an adaptive randomised open label multicentre platform trial recruiting participants from National Health Service Hospitals in the United Kingdom. Here we report the preliminary results of apixaban comparison of HEAL-COVID. Participants with a hospital admission related to confirmed COVID-19 and an expected date of discharge in the subsequent five days were randomised to either apixaban 2.5 mg twice daily or standard care (no anticoagulation) for 14 days. The primary outcome was hospital free survival at 12 months obtained through routine data sources. The trial was prospectively registered with ISRCTN (15851697) and Clincialtrials.gov (NCT04801940). Findings Between 19 May 2021 and 21 November 2022, 402 participants from 109 sites were randomised to apixaban and 399 to standard care. Seven participants withdrew from the apixaban group and one from the standard care group. Analysis was undertaken on an intention-to-treat basis. The apixaban arm was stopped on the recommendation of the oversight committees following an interim analysis due to no indication of benefit. Of the 402 participants randomised to apixaban, 117 experienced death or rehospitalisation during a median follow-up of 344.5 days (IQR 125 to 365), and 123 participants receiving standard care experienced death or rehospitalisation during a median follow-up of 349 days (IQR 124 to 365). There was no statistical difference in the rate of death and rehospitalisation (HR: 0.96 99%CI 0.69-1.34; p=0.75). Three participants in the apixaban arm experienced clinically significant bleeding during treatment. Interpretation Fourteen days of post-hospital anticoagulation with the direct oral anticoagulant apixaban did not reduce the rate of death or rehospitalisation of adults hospitalised with COVID-19. These data do not support the use of prophylactic post-hospital anticoagulation in adults with COVID-19. Funding HEAL-COVID is funded by the National Institute for Health and Care Research [NIHR133788] and the NIHR Cambridge Biomedical Research Centre [ BRC-1215-20014*].

Treating intrusive memories after trauma in healthcare workers: a Bayesian adaptive randomised trial developing an imagery-competing task intervention (preprint)

ABSTRACT Intensive care unit (ICU) staff continue to face recurrent work-related traumatic events throughout the COVID-19 pandemic. Intrusive memories (IMs) of such traumatic events comprise sensory image-based memories. Harnessing research on preventing IMs with a novel behavioural intervention on the day of trauma, here we take critical next steps in developing this approach as a treatment for ICU staff who are already experiencing IMs days, weeks, or months post-trauma. To address the urgent need to develop novel mental health interventions, we used Bayesian statistical approaches to optimise a brief imagery-competing task intervention to reduce the number of IMs. We evaluated a digitised version of the intervention for remote, scalable delivery. We conducted a two-arm, parallel-group, randomised, adaptive Bayesian optimisation trial. Eligible participants worked clinically in a UK NHS ICU during the pandemic, experienced at least one work-related traumatic event, and at least three IMs in the week prior to recruitment. Participants were randomised to receive immediate or delayed (after four weeks) access to the intervention. Primary outcome was the number of IMs of trauma during week 4, controlling for baseline week. Analyses were conducted on an intention-to-treat basis as a between-group comparison. Prior to final analysis, sequential Bayesian analyses were conducted ( n =20,23,29,37,41,45) to inform early stopping of the trial prior to the planned maximum recruitment ( n =150). Final analysis ( n =75) showed strong evidence for a positive treatment effect (Bayes factor, BF=1.25 × 10 6 ): the immediate arm reported fewer IMs (median=1, IQR=0-3) than the delayed arm (median=10, IQR=6-16.5). With further digital enhancements, the intervention ( n =28) also showed a positive treatment effect (BF=7.31). Sequential Bayesian analyses provided evidence for reducing IMs of work-related trauma for healthcare workers. This methodology also allowed us to rule out negative effects early, reduced the planned maximum sample size, and allowed evaluation of enhancements. Trial Registration NCT02044809 ( www.clinicaltrials.gov ).

Therapeutic potential of IL6R blockade for the treatment of sepsis and sepsis-related death: Findings from a Mendelian randomisation study (preprint)

Introduction Sepsis is characterised by dysregulated, life-threatening immune responses, which are thought to be driven by cytokines such as interleukin-6 (IL-6). Genetic variants in IL6R known to downregulate IL-6 signalling are associated with improved COVID-19 outcomes, a finding later confirmed in randomised trials of IL-6 receptor antagonists (IL6RA). We hypothesised that blockade of IL6R could also improve outcomes in sepsis. Methods We performed a Mendelian randomisation analysis using single nucleotide polymorphisms (SNPs) in and near IL6R to evaluate the likely causal effects of IL6R blockade on sepsis, sepsis severity, other infections, and COVID-19. We weighted SNPs by their effect on CRP and combined results across them in inverse variance weighted meta-analysis, proxying the effect of IL6RA. Our outcomes were measured in UK Biobank, FinnGen, the COVID-19 Host Genetics Initiative (HGI), and the GenOSept and GainS consortium. We performed several sensitivity analyses to test assumptions of our methods, including utilising variants around CRP in a similar analysis. Results In the UK Biobank cohort (N=485,825, including 11,643 with sepsis), IL6R blockade was associated with a decreased risk of sepsis (OR=0.80; 95% CI 0.66-0.96, per unit of natural log transformed CRP decrease). The size of this effect increased with severity, with larger effects on 28-day sepsis mortality (OR=0.74; 95% CI 0.38-0.70); critical care admission with sepsis (OR=0.48, 95% CI 0.30-0.78) and critical care death with sepsis (OR=0.37, 95% CI 0.14 - 0.98) Similar associations were seen with severe respiratory infection: OR for pneumonia in critical care 0.69 (95% CI 0.49 - 0.97) and for sepsis survival in critical care (OR=0.22; 95% CI 0.04- 1.31) in the GainS and GenOSept consortium. We also confirm the previously reported protective effect of IL6R blockade on severe COVID-19 (OR=0.69, 95% 0.57 - 0.84) in the COVID-19 HGI, which was of similar magnitude to that seen in sepsis. Sensitivity analyses did not alter our primary results. Conclusions IL6R blockade is causally associated with reduced incidence of sepsis, sepsis related critical care admission, and sepsis related mortality. These effects are comparable in size to the effect seen in severe COVID-19, where IL-6 receptor antagonists were shown to improve survival. This data suggests a randomised trial of IL-6 receptor antagonists in sepsis should be considered.

Whole genome sequencing identifies multiple loci for critical illness caused by COVID-19 (preprint)

Critical illness in COVID-19 is caused by inflammatory lung injury, mediated by the host immune system. We and others have shown that host genetic variation influences the development of illness requiring critical care1 or hospitalisation2;3;4 following SARS-Co-V2 infection. The GenOMICC (Genetics of Mortality in Critical Care) study is designed to compare genetic variants in critically-ill cases with population controls in order to find underlying disease mechanisms. Here, we use whole genome sequencing and statistical fine mapping in 7,491 critically-ill cases compared with 48,400 population controls to discover and replicate 22 independent variants that significantly predispose to life-threatening COVID-19. We identify 15 new independent associations with critical COVID-19, including variants within genes involved in interferon signalling (IL10RB, PLSCR1), leucocyte differentiation (BCL11A), and blood type antigen secretor status (FUT2). Using transcriptome-wide association and colocalisation to infer the effect of gene expression on disease severity, we find evidence implicating expression of multiple genes, including reduced expression of a membrane flippase (ATP11A), and increased mucin expression (MUC1), in critical disease. We show that comparison between critically-ill cases and population controls is highly efficient for genetic association analysis and enables detection of therapeutically-relevant mechanisms of disease. Therapeutic predictions arising from these findings require testing in clinical trials.

A Randomized Trial of Otilimab in Severe COVID-19 Pneumonia (OSCAR) (preprint)

BACKGROUNDIncreasing age is a risk factor for COVID-19 severity and mortality; emerging science implicates GM-CSF and dysregulated myeloid cell responses in the pathophysiology of severe COVID-19. METHODSWe conducted a large, global, double-blind, randomized, placebo-controlled study evaluating a single 90 mg infusion of otilimab (human anti-GM-CSF monoclonal) plus standard of care in adults hospitalized with severe COVID-19 respiratory failure and systemic inflammation, stratified by age and clinical status. Primary outcome was the proportion of patients alive and free of respiratory failure at Day 28; secondary endpoints included all-cause mortality at Day 60. RESULTSOverall, 806 patients were randomized (1:1); 71% of patients receiving otilimab were alive and free of respiratory failure at Day 28 versus 67% receiving placebo, although this did not reach statistical significance (model-adjusted difference 5.3% [95% CI -0.8, 11.4]; p=0.09). However, there was a benefit in the pre-defined [≥]70-year age group (model-adjusted difference 19.1% [95% CI 5.2, 33.1]; nominal p=0.009); these patients also had a reduction of 14.4% (95% CI 0.9, 27.9%; nominal p=0.04) in model-adjusted all-cause mortality at Day 60. Safety findings were comparable between otilimab and placebo, and consistent with severe COVID-19. CONCLUSIONSAlthough not statistically significant in the overall population, otilimab demonstrated a substantial benefit in patients aged [≥]70, possibly reflecting a population that could benefit from therapeutic blocking of GM-CSF in severe COVID-19 where myeloid cell dysregulation is predominant. These findings are being confirmed in a further cohort of patients aged [≥]70 in Part 2 of this study. (ClinicalTrials.gov number: NCT04376684).

Genetic mechanisms of critical illness in Covid-19 (preprint)

The subset of patients who develop critical illness in Covid-19 have extensive inflammation affecting the lungs and are strikingly different from other patients: immunosuppressive therapy benefits critically-ill patients, but may harm some non-critical cases. Since susceptibility to life-threatening infections and immune-mediated diseases are both strongly heritable traits, we reasoned that host genetic variation may identify mechanistic targets for therapeutic development in Covid-19. GenOMICC (Genetics Of Mortality In Critical Care, genomicc.org) is a global collaborative study to understand the genetic basis of critical illness. Here we report the results of a genome-wide association study (GWAS) in 2790 critically-ill Covid-19 patients from 208 UK intensive care units (ICUs), representing >95% of all ICU beds. Random controls were drawn from three distinct UK population studies. We identify and replicate several novel genome-wide significant associations including variants chr19p13.3 (rs2109069, P = 3.98 x 10-12), within the gene encoding dipeptidyl peptidase 9 (DPP9), and at chr21q22.1 (rs2236757, P = 4.99 x 10-8) in the interferon receptor IFNAR2. Consistent with our focus on extreme disease in younger patients with less comorbidity, we detect a stronger signal at the known 3p21.31 locus than previous studies (rs73064425, P = 1.2 x 10-27). We identify potential targets for repurposing of existing licensed medications. Using Mendelian randomisation we found evidence in support of a causal link from low expression of IFNAR2, and high expression of TYK2, to life-threatening disease. Transcriptome-wide association in lung tissue revealed that high expression of the monocyte/macrophage chemotactic receptor CCR2 is associated with severe Covid-19. We detected genome-wide significant gene-level associations for genes with central roles in viral restriction (OAS1, OAS2, OAS3). These results identify specific loci associated with life-threatening disease, and potential targets for host-directed therapies. Randomised clinical trials will be necessary before any change to clinical practice.

Supervised Machine Learning for the Early Prediction of Acute Respiratory Distress Syndrome (ARDS) (preprint)

Purpose: Acute respiratory distress syndrome (ARDS) is a serious respiratory condition with high mortality and associated morbidity. The objective of this study is to develop and evaluate a novel application of gradient boosted tree models trained on patient health record data for the early prediction of ARDS. Materials and Methods: 9919 patient encounters were retrospectively analyzed from the Medical Information Mart for Intensive Care III (MIMIC-III) data base. XGBoost gradient boosted tree models for early ARDS prediction were created using routinely collected clinical variables and numerical representations of radiology reports as inputs. XGBoost models were iteratively trained and validated using 10-fold cross validation. Results: On a hold-out test set, algorithm classifiers attained area under the receiver operating characteristic curve (AUROC) values of 0.905, 0.827, 0.810, and 0.790 when tested for the prediction of ARDS at 0-, 12-, 24-, and 48-hour windows prior to onset, respectively. Conclusion: Supervised machine learning predictions may help predict patients with ARDS up to 48 hours prior to onset.