key: cord-0807256-481o93v3 authors: H. Talasaz, A.; Sadeghipour, P.; Kakavand, H.; Aghakouchakzadeh, M.; Van Tassell, B. W.; Kordzadeh-Kermani, E.; Gheymati, A.; Ariannrjad, H.; Hosseini, S. H.; Jamalkhani, S.; Sholzberg, M.; Monreal, M.; Jimenez, D.; Piazza, G.; Parikh, S. A.; Kirtane, A. J.; Eikelboom, J. W.; Connors, J. M.; Hunt, B. J.; Konstantinides, S. V.; Cushman, M.; Weitz, J. I.; Stone, G.; Krumholz, H.; Lip, G. Y. H.; Goldhaber, S. Z.; Bikdeli, B. title: Antithrombotic Therapy in COVID-19: Systematic Summary of Ongoing or Completed Randomized Trials date: 2021-01-06 journal: nan DOI: 10.1101/2021.01.04.21249227 sha: 7657184a34902000a38e0acf24813d7d02f9a717 doc_id: 807256 cord_uid: 481o93v3 Endothelial injury and microvascular/macrovascular thrombosis are common pathophysiologic features of coronavirus disease-2019 (COVID-19). However, the optimal thromboprophylactic regimens remain unknown across the spectrum of illness severity of COVID-19. A variety of antithrombotic agents, doses and durations of therapy are being assessed in ongoing randomized controlled trials (RCTs) that focus on outpatients, hospitalized patients in medical wards, and critically-ill patients with COVID-19. This manuscript provides a perspective of the ongoing or completed RCTs related to antithrombotic strategies used in COVID-19, the opportunities and challenges for the clinical trial enterprise, and areas of existing knowledge, as well as data gaps that may motivate the design of future RCTs. Microvascular and macrovascular thrombotic complications including arterial and especially venous thromboembolism (VTE) appear to be common clinical manifestations of coronavirus disease 2019 (COVID-19), particularly among hospitalized and critically-ill patients (1) (2) (3) (4) . Pooled analyses have helped in providing aggregate estimates of thrombotic events (4, 5) . In a recent systematic review and meta-analysis, the overall incidence of VTE among inpatients with COVID-19 was estimated at 17% (95% confidence interval [CI], 13.4%-20.9%), with variation based on study design and method of ascertainment, with a four-fold higher incidence rate in patients in the intensive care units (ICU) compared with non-ICU settings (28% vs. 7%) (6) . In addition, postmortem studies show frequent evidence of microvascular thrombosis in patients with COVID-19 (7, 8) . The influence of these events on mortality rates remains unknown (9). COVID-19 can potentiate all three components of Virchow's triad and increase the risk of thrombosis ( Figure 1 ). First, SARS-CoV-2 infection may trigger endothelial dysfunction. Using the angiotensin converting enzyme (ACE)-2, expressed on the surface of many cells, SARS-CoV-2 enters endothelial cells and may impair their intrinsic antithrombotic properties. It is proposed that viremia, hypoxia, the inflammatory response, increased expression of tissue factor, and elevated levels of neutrophil extracellular traps (NETs) can together disrupt the hemostasis equilibrium, and promote endothelial activation (10) (11) (12) . This induction of a procoagulant state along with the reduction in plasminogen activators, further results in increased platelet reactivity (13) (14) (15) . Inflammatory cytokines and endothelial activation can lead to downregulation of 5 antithrombin and protein C expression, and an increase in the levels of plasminogen activator inhibitor, fibrinogen, factors V, VII, VIII, and X, in addition to von Willebrand factor (16) . Increased platelet reactivity, NETosis and alterations in the aforementioned hemostatic factors result in a hypercoagulable state (17) (18) (19) (20) (21) (22) . Particularly in COVID-19, it is thought that the excessive inflammatory response plays an important role in the pathogenesis of thrombosis (thromboinflammation), including pulmonary microthrombosis and pulmonary intravascular coagulopathy (7, 8) . Antiphospholipid antibodies have been identified in some patients (23) but their clinical significance is uncertain (24). Finally, venous stasis may synergistically combine with this prothrombotic milieu in patients more severely affected by the clinical syndrome of COVID-19 constituting of extreme fatigue with decreased daily activity, immobility due to oxygen requirements in the hospital. This may be further exacerbated by development of complications such as viral myocarditis and left ventricular systolic dysfunction (25, 26). All the aforementioned mechanisms may increase the risk of arterial and venous thrombosis, thereby impacting the severity of illness. Bedside observations, pathophysiological investigations, and initial epidemiological data led to enthusiasm for antithrombotic prophylaxis in COVID-19 (27-31). The concern for thrombotic risk was heightened by reports of VTE in 13% to 56% of patients despite the use of standard prophylaxis. (32-35) This led some experts to recommend empiric use of escalated doses of anticoagulants (36). However, the risks associated with intensified use of antithrombotic agents such as bleeding should be weighed against the presumptive benefits (22, 27, 31). 6 In addition, there have been variations in methodology and outcomes assessment for thrombotic events -including the concern about counting situ thrombosis in small vessels, a recognized feature of acute lung injury also known as immunothrombosis, as pulmonary emboli. Due to these issues, as well as well as the concerns for excess bleeding, a number of guidance statements have not recommended empiric escalated-dose anticoagulation (27, 37). Multiple ongoing randomized controlled trials (RCTs) are evaluating a variety of antithrombotic regimens in patients with COVID-19 ( Figure 2 ). These include trials of antiplatelet agents, anticoagulants, fibrinolytic agents, or combinations of these agents. In most trials, the intensity of antithrombotic therapy is proportional to the expected thrombotic event rates in the population under study. Less intensive therapies including antiplatelet agents, oral anticoagulants, and standard prophylactic dose of low molecular weight heparin (LMWH) are typically studied in the outpatient or lower acuity hospital settings. In turn, more intensive therapies including intermediate-dose or fully-therapeutic doses of anticoagulants, or even fibrinolytic therapy are under investigation in RCTs of hospitalized critically-ill patients. The aims of this article are to systematically summarize the ongoing and completed RCTs of antithrombotic therapy in patients with COVID-19, to evaluate the strengths and limitations of 7 fibrinolytic agents, and antithrombotic agents. We screened the identified studies and included those that were designed as RCTs with at least one active arm of antithrombotic therapy (date of last search: December 16, 2020) . Supplemental Table 1 summarizes study-level inclusion and exclusion criteria for this review. For the included studies, we searched PubMed and MedRxiv for design papers, study protocols or published results of those studies. The list was complemented by hand-searching and discussion within the author group. After identification of 918 records and manual screening of 180 records, we included 75 RCTs (Supplemental figure 1). In 13 cases, a design paper and/or study protocol was available. Of all ongoing studies, one RCT reported the results in peer-reviewed literature (38) and one shared the findings on a pre-print server (39). For three RCTs, final results are unknown but patient enrollment was paused in critically-ill patients due to concern for futility and potential excess of safety events (40). As of December 16, 2020, 75 RCTs of antithrombotic agents for patients with COVID-19 were registered at ClinicalTrials.gov or WHO ICTRP databases. Figure 2 provides a graphical summary of all RCTs of antithrombotic agents in COVID-19 in a pharmacologic-based approach. Agents used in these trials include antiplatelet agents, unfractionated heparin (UFH) and heparin derivatives, parenteral direct thrombin inhibitors (DTIs), direct oral anticoagulants (DOACs), fibrinolytic agents, sulodexide (a mixture of heparan sulfate and dermatan sulfate) (39), dociparstat (a heparin derivative with anti-inflammatory properties), and nafamostat (a synthetic serine protease inhibitor with anticoagulant activity). A succinct discussion of the design features of these 8 trials is provided below according to the clinical setting (see Supplemental Tables 2 and 3 for more details). Supplemental Table 4 summarizes the RCTs of other investigational agents with antithrombotic properties but not meeting our eligibility criteria for the current manuscript. In each section, the discussion starts with antiplatelet agents, followed by oral anticoagulants, parenteral anticoagulants, fibrinolytic therapy, and investigational agents with antithrombotic properties. This sequence is arbitrary and does not indicate treatment preference. Eleven RCTs of antithrombotic therapy in outpatients with COVID-19 have been registered in clinical trials databases and are studying aspirin, DOACs, enoxaparin, and sulodexide compared with no treatment (6/11) or with placebo (5/11). These trials are mostly (8/11) open-label, with the number of participants ranging from 172 to 7,000 patients, and include patients with a hyperinflammatory or procoagulant profile (including elevated levels of C-reactive protein [CRP] (1/11) or d-dimer [2/11] ) and exclude patients at high risk of bleeding (such as those with history of recent gastrointestinal bleeding or intracranial hemorrhage). Pregnant women and patients with severe kidney dysfunction (creatinine clearance [CrCl] < 30 mL/min) are excluded from 8 and 6 of these trials, respectively. The most common primary outcome in the outpatient trials include the need for hospitalization, incidence of thromboembolic events, mortality or composite outcomes inclusive of these within 21 to 90 days after randomization. Aspirin, DOACs (at both low-intensity and high-intensity), LMWHs (at standard prophylactic dose), and sulodexide are the agents under investigation in the outpatient setting. The 9 impact of low-dose aspirin on the composite rate of hospitalizations and mortality is being evaluated in 3 RCTs with a total of 12,080 patients with COVID-19 (ACTCOVID19, LEAD COVID-19, and ACTIV-4b). Low-intensity rivaroxaban ( SulES-COVID is the only completed trial of antithrombotic therapy in outpatients with COVID-19. This single-center study of 243 participants assessed the efficacy of sulodexide, compared with placebo on 21-day rates of hospitalization and need for use of supplemental oxygen. Use of sulodexide was associated with reduced hospital admissions (relative risk [RR], 0.6; 95% CI, 0.37-0.96; p=0.03) and need for oxygen support (RR, 0.71; 95% CI, 0.5 to 1; p=0.05), without a significant effect on mortality (39). The study has limitations, including frequent (22.1%) post-enrollment exclusions due to negative SARS-CoV2 test results or lost to follow-up. Many of the outpatient antithrombotic therapy trials for COVID-19 are large and the follow-up windows are sufficient to capture the intended primary outcomes. An issue with some All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint 10 of these trials is an open-label design, which is a pragmatic feature facilitating the design and enrollment but potentially limits the internal validity, especially for outcomes that may be less bias-resistant. In addition, the available data do not clarify whether dose adjustments are made for renal or liver dysfunction. We identified 50 ongoing RCTs related to antithrombotic therapy in hospitalized non-ICU patients with COVID-19. Most trials (44/50) are open-label. The antithrombotic agents under investigation include aspirin, P2Y12 inhibitors, dipyridamole, DOACs, heparin (both systemic and inhaled), dociparstat, nafamostat, and a combination of these drugs. The planned sample sizes range between 34 and 20,000 patients. Considering the potential link between elevated D-dimer, micro and macro-thrombosis in COVID-19, and worse outcomes in COVID-19 (42-44), many RCTs (N=16) include patients with elevated D-dimer levels with cut-offs ranging from >500 ng/mL to >1,500 ng/mL (or defined as >2-4 times the upper limit of normal per the local laboratory). Most trials exclude pregnant women (41 studies) and patients with active bleeding or history of intracranial or gastrointestinal bleeding (39 studies). Many trials also excluded patients with CrCl < 30 mL/min (20 studies). In most trials, the time frame for the primary outcome assessment is 28-30 days, although a few studies are designed to assess the primary outcomes at earlier or longer durations. These RCTs are focused on primary efficacy outcomes including all-cause mortality, VTE, arterial thrombosis, requirement for respiratory support, or a composite of these outcomes. The potential protective effect of antiplatelet agents in hospitalized patients with COVID-19 is being evaluated in 11 RCTs. REMAP-CAP is a large global RCT with a multifactorial adaptive design which is planning to randomize 7,100 patients to multiple therapeutic interventions All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; including an anticoagulant arm and antiplatelet agent arm looking ats aspirin, and the P2Y12 inhibitors clopidogrel, ticagrelor or prasugrel (45 The use of DOACs in hospitalized ward patients with COVID-19 is under investigation in 5 RCTs. Low-intensity rivaroxaban is being investigated in 650 planned participants in the ACOVACT and XACT trials of hospitalized patients to assess outcomes such as all-cause mortality, ICU admission, and intubation. High-intensity (but not loading-intensity) DOACs including rivaroxaban and apixaban are being evaluated in large RCTs that will enroll a total of 4,750 participants (ACTION, COVID-PREVENT, FREEDOM COVID, and XACT). 12 C-19-ACS is an adaptive RCT conducted to evaluate the impact of the combination of lowdose rivaroxaban (2.5 mg BID) plus aspirin 75 mg/day plus clopidogrel 75 mg/day along with atorvastatin and omeprazole on 30-day all-cause mortality in 3,170 hospitalized patients with COVID-19. Patients with definite acute coronary syndromes are excluded from this RCT. The effect of dual pathway inhibition using the combination of low-dose rivaroxaban and aspirin is being evaluated in the adaptive ACTCOVID19 inpatient study. In this RCT of 4,000 patients, the rate of invasive mechanical ventilation or death is assessed at 45 days post-randomization. Twenty-eight ongoing studies are being conducted to examine the efficacy of heparinbased regimens on primary outcomes such as all-cause mortality, venous and arterial thrombosis, re-hospitalization, the need for invasive mechanical ventilation, or composite outcomes inclusive of these in hospitalized patients with COVID-19. The majority of these RCTs has chosen standarddose prophylactic anticoagulation regimen as the comparator. Intermediate-dose anticoagulation will be tested in DAWn-Antico (48), X-COVID-19, COVID-19 HD, COVI-DOSE, EMOS-COVID, NCT04360824, and EUCTR2020-001891-14-ES with 4,434 patients in total. On the other hand, a total of 18 RCTs with 19,776 patients will evaluate the efficacy of therapeutic anticoagulation in non-ICU hospitalized patients (49). Only two trials totaling 494 patients (IMPACT and HEP-COVID) will directly compare therapeutic and intermediate doses of heparin. The different intensities of heparin derivatives are summarized in Supplemental Table 3 . Recognizing that heparin has more than an anticoagulant effect but also an antiviral and anti-inflammatory effect, INHALE-HEP and NEBUHEPA are evaluating the impact of nebulized UFH on the rate of intubation in 856 hospitalized patients with COVID-19. PACTR202007606032743 evaluates the impact of nebulized UFH on PaO2/FiO2 ratio in 100 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. High mobility group box protein 1 (HMGB1) is a protein involved in the pathogenesis of inflammation. Elevated levels of HMGB1 is associated with worse the outcomes in COVID-19 (50). Dociparstat, a heparin derivative with presumed anticoagulant and anti-inflammatory properties, inhibits HMGB1 and may reduce the formation of NETs and the risk of thrombosis. The drug is being studied in NCT04389840 to assess its impacts on all-cause mortality and need for mechanical ventilation in 600 patients with severe COVID-19 (51). Nafamostat is a synthetic serine protease inhibitor with anti-viral, anti-inflammatory, and anticoagulant activity previously used for anticoagulation during hemodialysis (52). Nafamostat is under evaluation in hospitalized patients with COVID-19 in 7 RCTs with 826 individuals in total. The primary efficacy outcome in most (5/7) of these trials is time to recovery. The strengths of many of the antithrombotic trials among inpatients with COVID-19 include relatively large sample sizes and ample follow-up for detection of events. With multiple large clinical trials underway, robust evidence should soon be available comparing the intermediate/therapeutic doses of heparinoids versus usual care. However, studies such as PARTISAN and NCT04420299 have relatively small sample sizes and short periods of follow-up (7 days and 10 days, respectively), rendering them susceptible to type II error. There is also variability across the trials in methods for identification and ascertainment of thrombotic outcomes. Lack of blinding and blinded outcome adjudication are practical limitations for some of these trials. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint 14 The risk of thrombotic events appears to be highest among critically-ill patients with COVID-19. A systematic review estimated that VTE event rates in critically-ill patients with COVID-19 would be estimated at 27.9% (95% CI, 22.1%-34.1%) (6) . Currently, there are 33 ongoing RCTs evaluating the role of antithrombotic agents in critically-ill patients with COVID-19 of which 18 RCTs enroll mixed non-ICU and ICU populations and 16 RCTs solely enroll ICU patients. The sample size of these studies range from 15 to 20,000 patients. These trials study the role of antiplatelet agents (aspirin, clopidogrel and dipyridamole), systemic anticoagulants (intermediate to full-therapeutic-dose of heparin and DTIs), inhaled UFH, fibrinolytic agents (tenecteplase and alteplase), and nafamostat. Inclusion criteria in 11 of 34 RCTs require D-dimer cut-offs ranging from >500 ng/ml to >3,000 ng/ml (or defined as >2-6 times the upper limit of normal limit). Allcause mortality, venous and arterial thrombotic complications, and oxygenation (expressed mostly as PaO2/FiO2) status are the most common components of the primary efficacy outcomes. Bleeding complications are the most widely used primary safety outcome among these studies. The role of antiplatelet agents is under investigation in critically-ill patients in 4 trials. As previously described, dipyridamole (TOLD) and aspirin (RECOVERY and CAM-Covid-19) are under evaluation. COVID-PACT is a multicenter, open-label study that will randomize 750 patients with 2 x 2 factorial design trial to full-dose anticoagulation vs. standard-dose prophylactic anticoagulation with heparin-based regimens (first randomization), and to antiplatelet therapy with clopidogrel vs. no antiplatelet therapy (second randomization). The primary efficacy outcome is the incidence of VTE or arterial thrombosis incidence 28 days after enrollment. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. the primary outcome for NCT04397510 is improvement in PaO2/FiO2 ratio within 10 days. 16 The use of parenteral anticoagulants other than UFH and LMWHs in COVID-19 is being studied in two trials. IMPACT will randomize 100 ICU patients with COVID-19 into 4 arms to compare fondaparinux, argatroban, intermediate-dose heparin and therapeutic-dose heparin (UFH/LMWH) with the primary outcomes of 30-day mortality. In ANTI-CO, bivalirudin is being investigated in 100 critically-ill patients for the primary outcome of improvement in oxygenation as determined by the PaO2/FiO2 ratio (55). Research in the ICU faces several challenges for study design, data/sample collection, and patient follow-up (57). In many cases, patients are unconscious and obtaining informed consent requires discussion with healthcare proxies. This is further complicated because visitors are prohibited. The strengths of the aforementioned studies in the ICU include the diversity of studied antithrombotic agents, and sample size in many RCTs. There are also a number of notable limitations to these trials. The most important limitation is the small sample size in several studies, All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint raising the possibility of type II error. The small sample size will mostly influence trials of thrombolytic therapy and non-heparin anticoagulants. ACTIV-4c (NCT04650087) is a double-blind, placebo-controlled RCT which will evaluate the impact of apixaban 2.5 mg BID on the rate of all-cause mortality, and arterial and venous thromboembolism on 5,320 post-discharge patients. MICHELLE (NCT04662684) is an openlabel RCT with 320 participants aims to evaluate the safety and efficacy of rivaroxaban 10 mg QD for 35+/-4 days versus no intervention after hospital discharge with a composite efficacy outcome of VTE and VTE-related death. In addition, there are 7 RCTs with a projected total of 1,452 participants which will continue the already assigned antithrombotic therapy after discharge in patients who were randomized in the general medical wards or in the ICU. In COVID-PREVENT, XACT, and NCT04508439 RCTs, post-discharge thromboprophylaxis with rivaroxaban (10 or 20 mg QD) is being investigated in 680 participants who were enrolled in general medical wards to measure the incidence of VTE at 30-35 days after discharge. In the HERO-19 study, edoxaban 60 mg QD or placebo will continue after discharge in 172 patients who were randomized in the ICU or non-ICU settings to evaluate all-cause mortality rate and VTE incidence at 42 days. In the INSPIRATION study, intermediate or standard prophylactic dose of enoxaparin will be continued after discharge in 600 patients who were randomized in the ICU to evaluate the rate of VTE. Finally, aspirin in the PEAC study, and dipyridamole ER plus aspirin in the ATTAC-19 study, will be continued after discharge in patients randomized in the non-ICU general wards. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint Most of the ongoing RCTs are excluding patients at increased risk of bleeding, or with acute and chronic hepatic failure. In more than 50% of the trials designed to evaluate escalated dose A large number of RCTs will help to delineate the efficacy and safety of antithrombotic agents in patients with COVID-19 (Central Illustration). Until the results accrue, participation in these RCTs is encouraged. Efficacy outcomes vary based on the location of enrollment; i.e., between outpatient trials and inpatient trials. As for safety outcomes, many of the trials are systematically assessing major bleeding using the International Society on Thrombosis and Haemostasis (ISTH) criteria or the Bleeding Academic Research Consortium definitions (58, 59). While observational evidence suggests low rates of major bleeding, (33, 60), observational studies have the potential for underreported outcomes and therefore RCTs with systematic and prospective capture of both thrombotic and bleeding events will help determine the true risk-benefit ratio for treatments. This All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; is especially the case since risk factors for thrombosis in COVID-19, such as D-dimer, may also predict bleeding (33). While results from the individual trials may inform interim practice, some challenges persist. The large number of antithrombotic agents under investigation, the variable dosing regimens tested, and variability in trial conduct as well as methods of outcome detection and adjudication may complicate the identification of the optimal regimens. A prospective metaanalysis of RCTs, ideally with individual participant data, will help to assess the effects of distinct agents across the spectrum of disease severity and may address the clinical and statistical heterogeneity of the upcoming results. Efforts to harmonize endpoints have been advocated, with creation of common data elements for VTE for example, to aid in pooling trial results (61, 62). In addition, there are few head-to-head comparisons for many of the experimental therapies, such as intermediate-dose compared with fully-therapeutic heparin-based regimens. Network metaanalytic techniques might generate insights into the comparative tradeoffs of these regimens (63). Additional biomarker and clinical risk prediction sub-studies can also further elucidate subgroups with more favorable net benefit profiles from distinct regimens. Moreover, the remaining knowledge gaps summarized in Table 1 should be kept in mind so that the design of additional studies could be considered. Anti-inflammatory properties and activity against thromboinflammation have been attributed to several antithrombotic regimens including heparin derivatives and antiplatelet agents (30, 64, 65), with the potential to reduce large-vessel thrombosis and improve outcomes. Another evolving concept is the role of microthrombosis and pulmonary intravascular coagulopathy (7, 8, 66) in the pathophysiology of respiratory failure in COVID-19 (67). Results from the small HESACOVID study suggested improved arterial oxygenation (PaO2/FiO2) with therapeutic All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; versus standard-dose prophylaxis anticoagulation in critically-ill patients with COVID-19 (38). However, combined investigation of three large-scale randomized trials of therapeutic anticoagulation (ACTIV-4a, REMAP CAP, and ATTACC) paused enrollment of critically ill patients for futility; we await further clarifications (40). Therapeutic drug monitoring of the investigational agents is also important. Even when an agent is selected (e.g., UFH), the best method for dose titration or adjustment remains uncertain (68). Some experts recommend measuring anti-Xa levels in those receiving intravenous unfractionated heparin, since the high levels of Factor VIII observed among critically-ill patients with COVID-19 may interfere with aPTT assays. The necessity and optimal method for dosing and monitoring of heparins and LMWHs, in particular for patients with kidney disease or obesity is yet to be elucidated, and is even understudied outside COVID-19 (69). Ideally, future strategy trials should test the merits and limitations of these monitoring tests. The clinical trial enterprise has been significantly affected during the COVID-19 pandemic (70). Patient recruitment in many ongoing pre-COVID-19 trials was temporarily halted. Notable challenges such as barriers to follow-up and site-monitoring persist. However, the desire to provide an evidence-based response has been one of the key drivers of positive changes during the pandemic (71). These changes include multi-specialty study teams, harmonization of multicenter protocols, expedited multi-institutional agreement execution and institutional review board and governmental agency approvals, accelerated informed consent, and enrollment with digital contact-free technology, expeditious outcomes ascertainment, remote monitoring, and All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint 21 dissemination of the findings via fast-track publications, pre-prints, and social media accounts from scientific societies or investigators (Table 2 ) (72-74). Although traditional RCTs have provided a great deal of knowledge for modern medicine, they are confined to testing a limited number of interventions. Since COVID-19 has multi-organ involvement and broad manifestations (including inflammation, ARDS, thrombosis and others), adaptive platform trials -which allow for testing multiple interventions in a single disease based on a decisive algorithm-have gained attention (75). This type of trial has a perpetual and multiarm-multi-stage design.(76) The RECOVERY Trial (77) and the World Health Organization Solidarity Trial (78) have tested different steroid and antiviral regimens, respectively, and have some additional agents under investigation, including aspirin in one of the hypotheses from RECOVERY. REMAP-CAP is testing several interventions including steroids, antivirals, biologic agents, simvastatin, and antiplatelet therapy. ACTIV4 platform is similarly using an adaptive design for antithrombotic agents. Notwithstanding the good will of investigators, the constant pressure to provide rapid pandemic response may pose challenges, as well. In some cases, multiple small single-center RCTs underpowered for their clinical points or using surrogate endpoints with short follow-up have been designed (71, 79) and may compete against larger multicenter, and potentially more definitive, studies. The large numbers of these trials alone, in addition to the intense pressure to present broadly and publish these findings suggests at least some potential for Type I error with amplification of these results through rapid dissemination of the results. Additional methodological aspects deserve attention. Interpretation of these trial results may be limited by underutilization of placebo (perhaps except for the outcome of mortality) (54, 79). Some experts consider that the pressures of working during a global pandemic makes the use All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint 22 of placebo more aspirational than realistic. Nevertheless, when feasible, placebo control improves the internal validity of a trial. Further, appropriate endpoint assessment, including blinded adjudication when feasible, and pre-specified analysis methods will remain of importance (54). IRBs and independent ethics committees may experience the burden of numerous protocol submissions and amendments during the pandemic. Burnout of healthcare systems during the pandemic, and the risks to the research teams are unique challenges that should also be considered when designing and executing study protocols (72). Investigators should attempt to foresee some of the challenges in order to minimize the need for protocol amendments (80-82). Moreover, the informed consent process has become adapted to facilitate discussions by telephone or video conference, followed by verbal confirmation, and documentation of consent using approved software programs and electronic signature, where acceptable (80, 83). Committee (CEC) and Data and Safety Monitoring Board (DSMB) meetings for assessing the adverse events is a fast safe, efficient alternative to face-to-face meetings. If done with appropriate planning to adhere to standards of high-quality CEC and DSMB meetings, such approaches may be considered even when society transitions out of the pandemic (81, 83). Peer-review and dissemination of the studies has had unique challenges and advancements, too. Journal editors and reviewers have been pressured for rapid release of the results of completed studies. This has activated fast-track peer-review process more than ever. Despite its merits, the "COVID-19 fatigue" created by the fast-track review process might negatively impact the quality of peer-review, as noted by occurrence of post-publication major revisions and retractions, including in major journals (84). In a recent study, only 29% of the clinical trials of patients with COVID-19 reviewed on ClinicalTrials.gov met the Oxford Centre for Evidence-Based Medicine All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; (OCEBM) level 2 evidence (85). The process of peer-review remains an imperfect -yet essentialstep in the evaluation and reporting of results (86). Preprint servers include full drafts of research studies shared publicly before peer-review. Pre-prints have the potential benefit of early dissemination and opportunity for feedback and discussion, and could be of substantial benefit during the pandemic. With a preprint, key researchers in the field can discover findings sooner; indicate critical errors, or suggest new studies or data that strengthen the argument (87). The limitations of pre-prints should be also communicated transparently, so that similar weight is not placed on pre-print and peer-reviewed literature by the lay people, the press, healthcare workers, or policy-makers. Indeed many retracted papers were from pre-print servers (84). Prospectively planned meta-analyses would be of particular help during the pandemic. Such studies can help understand the heterogeneity of the findings between interventions, between distinct studies, and within subgroups. Prospective meta-analysis can also help with pooled comparisons for interventions with small individual studies, as well as indirect comparisons for interventions that do not have sufficiently-large head-to-head comparisons in existing studies. Optimal antithrombotic therapy in patients with COVID-19 is yet to be determined. Results of these ongoing RCTs, and prospective meta-analyses of the completed studies will help clarify whether any of the plentiful antithrombotic regimens under investigation can safely mitigate thrombotic complications and improve patient outcomes. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. • The optimal thromboprophylactic regimens still remain unknown in patients with COVID-19. • At this time, dozens of RCTs are evaluating the role of antithrombotic regimens among outpatients and inpatients with COVID-19. • COVID-19 has led to changes in the design, conduct, analysis, and reporting of the RCT results. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; https://doi.org/10.1101/2021.01.04.21249227 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. ; • Specialty-based and multi-specialty collaboration common • Frequent ad hoc collaborations within and between institutions and countries • Diverse research priorities • Patient enrollment over a long time period, recruitment time could be slow of fast • Long-term follow up a routine feature of many trials • Distinct focus on COVID-19-related trials, some adaptations required for pre-COVID-19 trials. • Time-sensitive trial design (to provide rapid access to high quality evidence). Trial design in short period of time may lead to multiple smaller and underpowered trials rather than larger multicenter collaborations • Urgently needed medical solutions necessitate relatively fast patient enrollment • Incorporation of pragmatic design features. • Multiple projects around the world occasionally leading into several smaller trials rather than fewer large-scaler trials • Short-term follow up most common • Applicability for adaptive platform design for multiple aspects of COVID-19 trials. Multiple interventions, quick enrollment, and the possibility of re-estimation of the optimal sample size during the study • Higher certain event rates (death or re-admission) that excepted from protocol • Time-consuming review and approval process for funding allocation • Accelerated review, prioritizing trials that impact the response to the pandemic • Time-consuming process with occasional long delays prior to approval • IRBs meeting more frequently, often resulting in rapid review and approval • More permissive regulations may expedite trial initiation • Based on paper forms, may be cumbersome • In-person or remote electronic informed consent available in many trials • Variable willingness for trial participation by patients • Patients willing to participate and engage in trials as partners • Periodic slowdown or interruptions in enrollment for some non-COVID-19 trials, in COVID-19 trials there may be changes in enrollment rate with COVID-19 disease waves. • On site session for multiple predefined monitoring visits • On site or in-person data audits • Longer peer-review process • More strict criteria for publication • Uncommon use of pre-print servers • Fast-track peer-review process expedites the dissemination of completed studies. However, very quick peer-review has occasionally missed important flaws of submitted reports. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted January 6, 2021. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. To target anti-Xa 13, 14, 22, 26, 27 (level 0.3 -0.7 U/mL) 14, 22 (level 0.5 -0.7 U/mL) 26 To target anti-Xa 14 (level 0.3 -0.7 U/mL) 14 To target anti-Xa 14, 26 (level 0.3 -0.7 U/mL) 14 (level 0.5 -0.7 U/mL) 26 Registry of arterial and venous thromboembolic complications in patients with COVID-19 Anticoagulation, mortality, bleeding and pathology among patients hospitalized with COVID-19: a single health system study Venous Thromboembolism in COVID-19 High prevalence of deep vein thrombosis in mechanically ventilated COVID-19 patients Risk of venous thromboembolism in patients with COVID-19: A systematic review and meta-analysis Incidence of venous thromboembolism and bleeding among hospitalized patients with COVID-19: a systematic review and meta-analysis Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in Covid-19 Autopsy findings and venous thromboembolism in patients with COVID-19: a prospective cohort study Presenting Characteristics, Treatment Patterns, and Outcomes among Patients with Venous Thromboembolism during Hospitalization for COVID-19 COVID-19 is, in the end, an endothelial disease Coronavirus and Cardiovascular Disease, Myocardial Injury, and Arrhythmia: JACC Focus Seminar Complement and tissue factor-enriched COVID-19 and its implications for thrombosis and anticoagulation Potential Role of Platelets in COVID-19: Implications for Thrombosis COVID-19-A vascular disease Marked factor V activity elevation in severe COVID-19 is associated with venous thromboembolism COVID-19 update: Covid-19-associated coagulopathy NT-proBNP: N-Terminal-pro hormone B-type Natriuretic Peptide, O2 Sat: Oxygen Saturation Level, P: Prophylactic, PCI: Percutaneous Coronary Intervention, PCR: Polymerase Chain Reaction, PE: Pulmonary Embolism, P/F ratio (PaO2/FiO2) ratio: The ratio of arterial oxygen partial pressure to fractional inspired oxygen, PHV: Prosthetic Heart Valve, RR: Respiratory Rate, RVD: Right Ventricular Dysfunction, RQ-PCR: Real-time Quantitative Polymerase Chain Reaction, RT-PCR: Reverse Transcription Polymerase Chain Reaction, SARS-CoV-2: Severe Acute Respiratory Syndrome-Coronavirus-2 DOSE trial information in clinicaltrials.gov did not include definite intermediate doses and standard prophylactic doses for LMWH except enoxaparin, and a double dose of standard prophylactic dose of enoxaparin, adjusted to body weight, described as the intermediate dose We created a second section for INSPIRATION because in INSPIRATION trial, 1 mg/kg SC daily of enoxaparin as intermediate dose anticoagulation is only used for CrCl> 30 mL/min and weight< 120 kg, and for other situations (CrCl> 30 mL/min and weight> 120 kg or 15