UCDCRC/20/03

University College Dublin

Belfield, Dublin 4, Dublin, Ireland,

Clinical Trials Information, University College Dublin, +353 1716 4593, crc.monitoring@ucd.ie

Clinical Trials Information, University College Dublin, +353 1716 4593, crc.monitoring@ucd.ie

No

No

No

Final

09 Jun 2021

Yes

10 May 2021

Yes

14 Oct 2021

No

Primary Objective: To determine the effect of therapeutic anticoagulation, with LMWH (enoxaparin sodium), compared to standard care in hospitalized patients admitted for COVID-19 with an elevated D-dimer on the composite outcome of intensive care unit (ICU) admission, non-invasive positive pressure ventilation, invasive mechanical ventilation or death at 28 days.

This study was conducted according to Good Clinical Practice and the EU CT Directive 2001/20/EC and GCP Commission Directive 2005/28/EC. All participants or their legal representatives provided written informed consent before undergoing any trial related procedures. Authorized research ethics committees approved the trial at all participating sites. The global study was monitored by an independent data safety monitoring board (DSMB) including a biostatistician, a hematologist, a general internist and intensive care specialist assigned by the global trial sponsor United Health Toronto – St. Michael’s Hospital.

01 Feb 2021

No

Yes

Ireland: 23

Brazil: 105

Canada: 150

Saudi Arabia: 147

United Arab Emirates: 13

United States: 27

465

23

0

0

0

0

0

0

276

164

25

Overall trial (overall period)

Randomised - controlled

Blinding of participants, clinical research staff, and clinicians was not possible due to the nature of the intervention. However, important clinical outcomes were adjudicated by independent, blinded, clinical content experts.

Yes

Low molecular weight heparin (LMWH) or unfractionated heparin (UFH)

Concentrate for solution for injection/infusion

Intravenous use, Subcutaneous use

Therapeutic heparin was administered until hospital discharge, death, day 28 or study withdrawal. Choice of LMWH vs UFH was at clinician’s discretion and dependent on local institutional supply. LMWH dose regimens were dependent on CrCl (creatinine clearance) and BMI. e.g. for Enoxaparin: CrCl ≥30 and BMI <40 1 mg/kg SC q12h OR 1.5 SC mg/kg q24h Enoxaparin CrCl ≥30 and BMI ≥40 1 mg/kg q12h* Enoxaparin CrCl <30 UFH IV bolus, with continuous infusion to titrate to institution specific anti-Xa or aPTT values* or LMWH per hospital protocol taking BMI into consideration as above *For BMI above 40, measurement of anti-Xa to confirm therapeutic effect could be used. UFH, if used in the experimental arm, was administered using a weight- based nomogram (bolus plus continuous infusion) with activated partial thromboplastin time (aPTT) or UFH anti-Xa titration according to the center-specific institutional protocols as per venous thromboembolism treatment (i.e. high dose nomogram)

Low molecular weight heparin (LMWH), unfractionated heparin (UFH) or fondaparinux

Concentrate for solution for injection/infusion

Subcutaneous use

Administration of LMWH, UFH or fondaparinux at thromboprophylactic doses for acutely ill hospitalized medical patients, in the absence of contraindication, is generally considered standard care. Prophylactic dose level was defined based on the best available evidence from clinical trials and expert consensus, and took BMI and creatinine clearance (CrCl) into consideration. e.g. for Enoxaparin: ≥30 CrCl and BMI <40: 40 mg SC q24h ≥30 CrCl and BMI ≥40: 40 mg SC q12h <30 CrCl and BMI <40: UFH 5000 units SC q8-12h or LMWH per hospital protocol taking BMI into consideration <30 CrCl and BMI ≥40: UFH 7500 units SC q8h or LMWH per hospital protocol taking BMI into consideration as above Full therapeutic dose anticoagulation (therapeutic dose LMWH) was permitted as rescue therapy in the event of suspected or confirmed thromboembolism. Additional rescue therapy, in the form of thrombolysis (with tissue plasminogen activator), was also permitted if deemed clinically warranted.

Therapeutic Heparin

Prophylactic Heparin

Therapeutic Heparin

Prophylactic Heparin

If a patient was discharged alive before 28 days, vital status was determined using a telephone follow- up. If a patient was discharged alive on mechanical ventilation (invasive or non-invasive) prior to day 28, a call to the patient or a doctor/nurse from the rehabilitation health facility was made to confirm ventilation status on day 28 and their last day of mechanical ventilation.

Primary

From randomisation to 28 days

Primary analyses were by the intention-to-treat population of all randomized patients in accordance with the allocated intervention. A chi-square test was conducted to derive a two-sided p-value for the main analysis of the primary outcome. A logistic regression model was fitted to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.69

2-sided

The per protocol set was restricted those who received the experimental or control intervention as allocated during the first 48 hours after randomization. A logistic regression model was fitted to derive an odds ratio with 95% confidence intervals. The per protocol set includes 216 patients in the therapeutic heparin arm (34 met the primary endpoint) and 227 patients in the prophylactic heparin arm (47 met the primary endpoint).

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.72

2-sided

Sensitivity analysis 1 excluded patients who did not meet a component of the primary composite outcome and did not have a follow-up up to day 28; 11 patients in therapeutic heparin group and 12 patients in the prophylactic heparin group. 37/217 met this endpoint in the therapeutic heparin arm and 52/225 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.68

2-sided

Sensitivity analysis 2 excluded those who did not satisfy all eligibility criteria (i.e. those with a negative d-dimer; 6 patients in the therapeutic heparin group and 5 in the prophylactic heparin group). 36/222 met this endpoint in the therapeutic heparin arm and 48/231 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.74

2-sided

Sensitivity analysis 3 was conducted using logistic regression to derive odds ratios with 95% confidence intervals, excluding patients who did not meet a component of the primary composite outcome, did not have a follow-up up to day 28 and those who did not satisfy all eligibility criteria; 17 patients in the therapeutic heparin group and 18 patients in the prophylactic heparin group. 36/211 met the primary endpoint in the therapeutic heparin arm and 48/219 in the prophylactic heparin arm

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.73

2-sided

Intention-to-treat analysis of the primary endpoint adjusted for age, taking into account that randomization was stratified by age. An odds ratio with 95% confidence interval was estimated by logistic regression.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.68

2-sided

This intention-to-treat analysis of the primary endpoint was repeated to adjust for time. To address changes in co-interventions over time due to emerging evidence from Covid-19 clinical trials, a logistic regression model was used to fit a time by treatment interaction where time was days since first randomized subject. Time was modelled with a restricted cubic spline having 3 knots. Three knots were chosen because of the modest number of events. The model with splines and interactions revealed

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.69

2-sided

If a patient was discharged alive before 28 days, vital status was determined using a telephone follow- up.

Secondary

From randomisation to 28 days

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.22

2-sided

The per protocol set was restricted those who received the experimental or control intervention as allocated during the first 48 hours after randomization. A logistic regression model was fitted to derive an odds ratio with 95% confidence intervals. The per protocol set includes 216 patients in the therapeutic heparin arm (4 experienced death from any cause within 28 days) and 227 patients in the prophylactic heparin arm (17 experienced death from any cause within 28 days)

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.23

2-sided

Sensitivity analysis 1 excluded patients who did not meet a component of the primary composite outcome and did not have a follow-up up to day 28; 11 patients in therapeutic heparin group and 12 patients in the prophylactic heparin group. 4/217 met this endpoint in the therapeutic heparin arm and 18/225 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.22

2-sided

Sensitivity analysis 2 excluded those who did not satisfy all eligibility criteria (i.e. those with a negative d-dimer; 6 patients in the therapeutic heparin group and 5 in the prophylactic heparin group). 4/222 met this endpoint in the therapeutic heparin arm and 17/231 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.23

2-sided

Sensitivity analysis 3 was conducted using logistic regression to derive an odds ratio with a 95% confidence interval, excluding patients who did not meet a component of the primary composite outcome, did not have a follow-up up to day 28 and those who did not satisfy all eligibility criteria; 17 patients in the therapeutic heparin group and 18 patients in the prophylactic heparin group. 4/211 met the endpoint within the therapeutic heparin arm and 17/219 in the prophylactic heparin arm

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.23

2-sided

Intention-to-treat analysis of the primary endpoint components were adjusted for age, taking into account that randomization was stratified by age. An odds ratio with 95% confidence interval was estimated by logistic regression.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.19

2-sided

Secondary

Up to 28 days post-randomisation

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.7

2-sided

The per protocol set was restricted those who received the experimental or control intervention as allocated during the first 48 hours after randomization. A logistic regression model was fitted to derive an odds ratio with 95% confidence intervals. The per protocol set includes 216 patients in the therapeutic heparin arm (9 met this endpoint) and 227 patients in the prophylactic heparin arm (13 met this endpoint).

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.72

2-sided

Sensitivity analysis 1 excluded patients who did not meet a component of the primary composite outcome and did not have a follow-up up to day 28; 11 patients in therapeutic heparin group and 12 patients in the prophylactic heparin group. 11/217 met this endpoint in the therapeutic heparin arm and 16/225 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.7

2-sided

Sensitivity analysis 2 excluded those who did not satisfy all eligibility criteria (i.e. those with a negative d-dimer; 6 patients in the therapeutic heparin group and 5 in the prophylactic heparin group). 10/222 met this endpoint in the therapeutic heparin arm and 14/231 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.73

2-sided

Sensitivity analysis 3 excluded patients who did not meet a component of the primary composite outcome, did not have a follow-up up to day 28 and those who did not satisfy all eligibility criteria; 17 patients in the therapeutic heparin group and 18 patients in the prophylactic heparin group. 10/211 received invasive mechanical ventilation in the therapeutic heparin group and 14/219 in the prophylactic heparin group.

Prophylactic Heparin v Therapeutic Heparin

465

Pre-specified

0.73

2-sided

Intention-to-treat analysis of invasive mechanical ventilation adjusted for age taking into account that randomization was stratified by age.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.69

2-sided

Invasive or non-invasive (bilevel or continuous positive airway pressure) mechanical ventilation.

Secondary

Up to 28 days post-randomisation

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.82

2-sided

The per protocol set was restricted those who received the experimental or control intervention as allocated during the first 48 hours after randomization. A logistic regression model was fitted to derive an odds ratio with 95% confidence intervals. The per protocol set includes 216 patients in the therapeutic heparin arm (18 with any mechanical ventilation) and 227 patients in the prophylactic heparin arm (22 with any mechanical ventilation).

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.85

2-sided

Sensitivity analysis 1 excluded patients who did not meet a component of the primary composite outcome and did not have a follow-up up to day 28; 11 patients in therapeutic heparin group and 12 patients in the prophylactic heparin group. 21/217 met this endpoint in the therapeutic heparin arm and 26/225 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.82

2-sided

Sensitivity analysis 2 excluded those who did not satisfy all eligibility criteria (i.e. those with a negative d-dimer; 6 patients in the therapeutic heparin group and 5 in the prophylactic heparin group). 20/222 met this endpoint in the therapeutic heparin arm and 23/231 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.9

2-sided

Sensitivity analysis 3 was conducted using logistic regression to derive an odds ratio with a 95% confidence interval, excluding patients who did not meet a component of the primary composite outcome, did not have a follow-up up to day 28 and those who did not satisfy all eligibility criteria; 17 patients in the therapeutic heparin group and 18 patients in the prophylactic heparin group.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.89

2-sided

Intention-to-treat analysis of any mechanical ventilation adjusted for age taking into account that randomization was stratified by age.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.82

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.79

2-sided

The per protocol set was restricted those who received the experimental or control intervention as allocated during the first 48 hours after randomization. A logistic regression model was fitted to derive an odds ratio with 95% confidence intervals. The per protocol set includes 216 patients in the therapeutic heparin arm (30 admitted to the ICU) and 227 patients in the prophylactic heparin arm (37 admitted to ICU).

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.83

2-sided

Sensitivity analysis 1 excluded patients who did not meet a component of the primary composite outcome and did not have a follow-up up to day 28; 11 patients in therapeutic heparin group and 12 patients in the prophylactic heparin group. 33/217 met this endpoint in the therapeutic heparin arm and 42/225 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.78

2-sided

Sensitivity analysis 2 excluded those who did not satisfy all eligibility criteria (i.e. those with a negative d-dimer; 6 patients in the therapeutic heparin group and 5 in the prophylactic heparin group). 32/222 met this endpoint in the therapeutic heparin arm and 39/231 met this endpoint in the prophylactic heparin arm. Logistic regression was used to derive odds ratios with 95% confidence intervals.

Therapeutic Heparin v Prophylactic Heparin

465

Post-hoc

0.83

2-sided

Sensitivity analysis 3 excluded patients who did not meet a component of the primary composite outcome, did not have a follow-up up to day 28 and those who did not satisfy all eligibility criteria; 17 patients in the therapeutic heparin group and 18 patients in the prophylactic heparin group. 32/211 subjects were admitted to the ICU in the therapeutic heparin arm and 39/219 in the prophylactic heparin arm

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.83

2-sided

Intention-to-treat analysis ICU admission adjusted for age taking into account that randomization was stratified by age.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.78

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.59

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.7

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Ordinal logistic regression was used to derive odds ratios with 95% confidence intervals. Death up to 28 days was assigned the worst outcome (a value of -1) in these analyses. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

1.77

2-sided

Organ support was defined as receipt of non-invasive mechanical ventilation, high flow nasal cannula oxygen, invasive mechanical ventilation, or vasopressor therapy. Any patient who died during the acute hospital stay was assigned 28 Day Organ-Support Free Days of –1. If there was intervening time in which a patient was free of organ support, but went back on organ support, the intervening time did not count toward the organ support free days endpoint. Only time before organ support and after the last use of organ support was counted as “free days”. If a patient was discharged alive without mechanical ventilation prior to Day 28, the patient was assumed to be free of organ support after hospital discharge for the remainder of the 28 days. If a patient was discharged alive on mechanical ventilation (invasive or non-invasive) prior to day 28, a call to the patient or a doctor/nurse from the rehabilitation health facility was made to confirm ventilation status.

Secondary

Defined as the number of days that a patient was alive and free of organ support through 28 days after trial entry.

This analysis was conducted on the ITT population. Ordinal logistic regression was used to derive odds ratios with 95% confidence intervals. Death up to 28 days was assigned the worst outcome (a value of -1) in these analyses. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Prophylactic Heparin v Therapeutic Heparin

465

Pre-specified

1.41

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Ordinal logistic regression was used to derive odds ratios with 95% confidence intervals. Death up to 28 days was assigned the worst outcome (a value of -1) in these analyses. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

1.51

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Ordinal logistic regression was used to derive odds ratios with 95% confidence intervals. Death up to 28 days was assigned the worst outcome (a value of -1) in these analyses. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

1.09

2-sided

Renal replacement therapy was defined as continuous renal replacement therapy {CRRT} or intermittent hemodialysis {IHD};

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.41

2-sided

Thromboembolism was diagnostically confirmed except for 1 symptomatic deep vein thrombosis in the prophylactic heparin group, which could not be definitively confirmed as diagnostic imaging was not done during acute symptomatic period.

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.29

2-sided

Thromboembolism was all diagnostically confirmed except for 1 symptomatic deep vein thrombosis in the prophylactic heparin group, which could not be definitively confirmed as diagnostic imaging was not done during acute symptomatic period.

Secondary

Up to 28 days post-randomization.

Secondary

Up to 28 days post-randomization.

Since D-dimer assays differed across sites, D-dimer levels were analyzed as the logarithm of D-dimer x ULN by taking the natural logarithm of the ratio of the actual d-dimer value divided by the ULN for the assay used. SD reported is for the natural logarithm of D-dimer levels x ULN. The day 2±24 hours D-dimer was missing for 66 in the therapeutic heparin group and 64 in the prophylactic heparin group. In accordance with the statistical analysis plan, because this outcome had missing data for more than 5% of the subjects, the main analysis was a complete case analysis , with sensitivity analysis by inverse probability weighted analysis and multiple imputation.

Secondary

D-dimer level at 2 days ± 24 hours post-randomization

Ratio of geometric means of D-dimer levels x ULN of day 2±24h post-randomization, adjusted for baseline geometric means of D-dimer levels x ULN using analysis of covariance. The day 2±24 hours D-dimer was missing for 66 in the therapeutic heparin group and 64 in the prophylactic heparin group.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.88

2-sided

Sensitivity analysis. D-dimer levels at day 2±24 hours post-randomization were missing for 66 (29.0%) in the therapeutic heparin group and 64 (27.0%) in the prophylactic heparin groups. Ratio of geometric means of D-dimer level x ULN of day 2±24h post-randomization, adjusted for baseline geometric means of D-dimer levels x ULN using analysis of covariance. SD for the natural logarithm of D-dimer levels x ULN. Inverse probability weighted analysis used to account for missing data.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.87

2-sided

Sensitivity analysis. D-dimer levels at day 2±24 hours post-randomization were missing for 66 (29.0%) in the therapeutic heparin group and 64 (27.0%) in the prophylactic heparin groups. Ratio of geometric means of D-dimer level x ULN of day 2±24h post-randomization, adjusted for baseline geometric means of D-dimer levels x ULN using analysis of covariance. SD for the natural logarithm of D-dimer levels x ULN. Multiple imputation used to account for missing data.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.91

2-sided

Secondary

ISTH: International Society on Thrombosis and Haemostasis. Major bleeding defined by the ISTH Scientific and Standardization Committee.

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.52

2-sided

Secondary

Up to 28 days post-randomization

This analysis was conducted on the ITT population. Logistic regression was used to derive odds ratios with 95% confidence intervals. Secondary outcomes were exploratory and were not adjusted for multiple comparisons.

Therapeutic Heparin v Prophylactic Heparin

465

Pre-specified

0.34

2-sided

Secondary

Up to 28 days post-randomization

The AE reporting period began on the day of randomization and ended with the final study (follow-up) visit.

Investigators and study staff assessed the occurrence of AEs and SAEs at all subject evaluation time points during the study. Adverse events were recorded in the subject’s medical records and on applicable AE source documents. Regulatory authorities, REBs and investigators were notified of SAEs in accordance with applicable requirements.

N/A

Therapeutic Heparin

Prophylactic Heparin