Enoxaparin Thromboprophylaxis in Children Hospitalized for COVID-19: A Phase 2 Trial

The COVID-19 Anticoagulation in Children – Thromboprophylaxis (COVAC-TP) trial (www.clinicaltrials.gov, NCT04354155) was an investigator-initiated, multicenter, open-label, nonrandomized trial conducted at 8 children’s hospitals and academic medical centers in the United States and funded by the Johns Hopkins All Children’s Foundation from June 2020 through June 2021. The protocol (Supplemental Information 1) was approved by the institutional review board (IRB) at Johns Hopkins University (IRB 00247991) using a single IRB system for most centers, and independently reviewed at each participating site. Signed informed consent (and assent where appropriate) were required for all participants. The Clinical Coordinating Center and Data Coordinating Center were located at the Johns Hopkins All Children’s Institute for Clinical and Translational Research (St. Petersburg, FL). The trial monitoring plan is detailed in Supplemental Information 1. The trial was overseen by a Steering Committee and an Independent Data and Safety Monitoring Committee. All primary outcomes and adverse events were adjudicated by a Clinical Endpoint Adjudication Committee. The compositions of each study committee are provided in Supplemental Information 1.

Complete inclusion and exclusion criteria are provided in Supplemental Information 2. In summary, children <18 years of age (inclusive) hospitalized for symptomatic SARS-CoV-2-related illness including MIS-C and primary COVID-19 within an intensive care or acute care unit setting were eligible for enrollment. Main exclusion criteria were an expected length of stay (LOS) <2 days, treatment with therapeutic anticoagulation, concomitant aspirin at doses >5mg/kg per day, presence of clinically-relevant bleeding within 72 hours before enrollment, unwillingness to consent to participate, or any of the following laboratory-based exclusion criteria: platelet count <50 000/uL; prothrombin time ≥2 seconds above, or activated partial thromboplastin time ≥4 seconds above the upper limit of local age-dependent reference range; fibrinogen level <75 mg/dL; or estimated glomerular filtration rate <31 mL/min per 1.73 m² as calculated using the Schwartz formula. Race and ethnicity (as self-identified by participants, using fixed categories) were recorded in accordance with National Institutes of Health requirements for tracking inclusion of minorities in clinical trials, and given racial and ethnic disparities in VTE incidence and outcomes, including hospitalized cases with COVID-19.^13,34,35 

The protocol’s schedule of assessments is provided in Supplemental Table 4 and Supplemental Information 1. Patients received enoxaparin at a starting dose of 0.5 mg per kg (max starting dose: 60 mg per dose) subcutaneously every 12 hours, adjusted to achieve a 4 hour postdose anti-Xa level of 0.20 to 0.49 U/mL following the first or second dose based upon recommendations from the ISTH.³³  For anti-Xa values <0.20 U/mL or ≥0.5 U/mL, enoxaparin dose was increased or decreased by 10% to 20%, respectively, and a repeat 4-hour postdose anti-Xa level was obtained. For patients who were started on enoxaparin twice-daily thromboprophylaxis before enrollment, based on the treating physician’s clinical judgement, the protocol did not require a change in dosing on-study if the anti-Xa level was in the protocol-specified target range. Serious adverse events, safety laboratory values, and pertinent clinical data were captured from patient enrollment through trial completion at 30 days (±2 days) postenrollment. Nonserious adverse events were not collected for this trial, for which an Investigational New Drug Application Waiver was granted by the US Food and Drug Administration.

The primary endpoint was safety—specifically, ISTH-defined clinically-relevant bleeding (ie, major plus clinically-relevant nonmajor bleeding)³⁶  within 30-days of enrollment. Major bleeding included: (1) fatal bleeding; (2) clinically overt bleeding associated with a decrease in hemoglobin (Hgb) of at least 2 g/dL in a 24 hour period; (3) bleeding that is retroperitoneal, pulmonary, intracranial, or otherwise involves the central nervous system; and (5) bleeding that requires surgical intervention in an operating suite. Clinically-relevant nonmajor bleeding included: (1) overt bleeding for which a blood product is administered and that is not directly attributable to the patient’s underlying medical condition; and (2) bleeding that requires medical or surgical intervention to restore hemostasis, other than in an operating suite.

The secondary endpoint was the median twice-daily enoxaparin dose (mg/kg) required to achieve a 4 hour postdose antifactor Xa level of 0.20 to 0.49 anti-Xa U/mL. The cumulative incidence of HA-VTE within 30 days was defined as an exploratory efficacy endpoint. Clinically-relevant bleeding and recurrent VTE events reported by investigators were centrally adjudicated via a Clinical Endpoint Adjudication Committee.

The statistical analysis plan (Supplemental Information 1) was developed by the lead trial biostatistician (E.A.). Patient demographics and baseline characteristics were summarized with counts and percentages for categorical variables, and with means, standard deviations, medians, and interquartile ranges (IQR) as appropriate, for numeric variables. The primary analysis population for the study endpoints was prespecified as all patients received at least 2 doses of enoxaparin postenrollment and before hospital discharge. Cumulative incidences were calculated with corresponding exact (Clopper-Pearson) 90% confidence intervals (90% CIs). According to the protocol, we targeted an enrollment of 38 patients. We calculated with this sample size, the precision of estimates for various cumulative incidences of clinically-relevant bleeding (with corresponding 90% CIs) was as follows: 0 of 38, 0% (0% to 7.6%); 1 of 38, 2.6% (0% to 11.8%); and 2 of 38, 5.3% (1% to 15.7%). The secondary endpoint (enoxaparin dose requirement) was summarized by COVID-19 presentations (primary COVID-19 or MIS-C) with medians and corresponding IQRs. Medians were compared between age groups (<12 versus ≥12 years) and COVID-19 presentations (primary COVID-19 versus MIS-C) using the Kruskal-Wallis test. An age of 12 years was selected for subgroup analysis based on epidemiologic data demonstrating a peak in VTE incidence among adolescents.^15,16,18  Two-sided P values of <.05 were considered to indicate statistical significance. Missing data were not imputed. Adverse events were descriptively analyzed in the safety population, defined as all patients who received at least 1 dose of enoxaparin postenrollment. Analyses were performed with SAS software, version 9.4 (SAS Institute, Cary, NC, USA).

A total of 40 patients from 8 participating United States sites were enrolled in the study, of which 38 were retained in the primary analysis population (Fig 1). One patient was withdrawn by parental request and another discharged from the hospital before receiving 2 doses of enoxaparin postenrollment. Clinical presentation was primary COVID-19 in 18 children and MIS-C in 20. The median age was 12.1 (range 1.3–17.5) years, 52.6% were male, and 47.4% were female. Table 1 summarizes demographic data, body mass index (BMI), and other baseline clinical characteristics by presentation; Table 2 provides these data by patient age <12 or ≥12 years (n = 19 each). As comparedwith children presenting with primary COVID-19, those with MIS-C were younger (median age [IQR]: 9.9 [7.3–12.6] versus 15.4 [12.5–17] years, P = .002), had lower BMI (median [IQR]: 20.1 [16.3–23.3] versus 32.1 [23.1–43.4] kg/m², P < .001), and had higher D-dimer levels (median [IQR]: 3.4 [2.1–4.7] versus 0.8 [0.6–2] µg/mL, P < .001). Adolescents, when compared with younger children, had higher BMIs (median [IQR]: 31.3 [21.7–43.4] versus 18.8 [16.3–24.7] kg/m², P < .001) and lower D-dimer levels (median [IQR] 1.2 [0.7–2] versus 3.4 [2.2–4.3] µg/mL, P = .001). Details on treatments employed for underlying COVID-19 disease are provided in Table 3.

No primary safety endpoint (ie, clinically-relevant bleeding) events were observed (cumulative incidence, 0% [90% CI, 0% to 7.6%]). Enoxaparin dose requirement did not differ by age ≥12 vs <12 years (median [IQR]: 0.5 [0.46–0.52] versus 0.52 [0.49–0.55] mg/kg, P = .51) but was higher among those presenting with MIS-C compared with primary respiratory COVID-19 (0.52 [0.5–0.61] versus 0.48 [0.39–0.51] mg/kg, P = .010). Targeted prophylactic anti-Xa levels were achieved in 35 of 38 children (92.1%). Twenty-eight participants (74.7%) required no dosing adjustments to achieve target anti-Xa levels. Ten participants (26.3%) underwent a single dosing adjustment after initial anti-Xa levels, of which 8 (75%) represented dose increases per protocol.

Two participants with primary COVID-19 developed VTE (on days 1 and 7 postenrollment), yielding a cumulative incidence of 5.3% (90% CI, 1.0% to 15.7%). Both VTEs were central venous catheter-related DVTs of the upper extremities that occurred while receiving the enoxaparin regimen at the targeted prophylactic anti-Xa levels. The indications for central venous catheterization for both subjects were hemodynamic monitoring, frequent blood draws, and access to administer medications following invasive ventilation. No VTE were recorded after hospital discharge at time of 30-day follow up data collection. Of the 36 participants discharged from the hospital by the conclusion of the study period, 13 (36.1%) were prescribed outpatient thromboprophylaxis (aspirin, n = 11; rivaroxaban, n = 2). Participants prescribed aspirin at discharge all had MIS-C.

One death occurred during study in a participant with primary COVID-19. This was deemed by the principal investigator to be related to underlying COVID-19 illness and unrelated to study intervention; the independent data and safety monitoring committee concurred with this assessment. No other serious adverse events were observed.

This multicenter phase 2 trial assessed safety, dose-finding, and exploratory efficacy of subcutaneous enoxaparin twice-daily dosing at 0.5 mg/kg for primary thromboprophylaxis among children hospitalized for SARS-CoV-2-related illness. This trial represents 1 of few prospective pediatric investigations of thromboprophylaxis reported for hospitalized children, and the first among children hospitalized for COVID-19. These analyses demonstrated the safety of enoxaparin based on the absence of clinically-relevant bleeding events within the prespecified range of clinically-relevant bleeding events. Goal anti-Xa levels were achieved in 92.1% of participants with few dosing adjustments. Dosing did not appear to vary by age groups ≥12 versus <12 years of age. Children hospitalized for MIS-C required a statistically (but not clinically-relevant) greater dose to achieve goal anti-Xa levels than those with primary COVID-19. Participants with MIS-C had significantly elevated D-dimer levels (median of 3.4 µg/mL, IQR: 2.1–4.7) as compared with primary COVID-19 and the observed increased enoxaparin dose requirement may be related to the potent proinflammatory hypercoagulable state seen in MIS-C.^5,6,13  Two children (5.3%) developed central venous catheter-related VTE during enoxaparin thromboprophylaxis. These findings question the intensity of enoxaparin exposure employed in this trial and preliminarily recommended by ISTH are sufficiently efficacious thromboprophylaxis regimens (versus therapeutic anticoagulation dosing) for VTE prevention in children hospitalized for COVID-19-related illness. Overall, for children hospitalized for COVID-19 related illness, primary thromboprophylaxis with twice-daily subcutaneous enoxaparin is safe with respect to the risk of clinically-relevant bleeding. Phase 3 trials of enoxaparin thromboprophylaxis are warranted in hospitalized children with primary COVID-19, MIS-C, and other proinflammatory conditions for which VTE risk is increased.

Severe COVID-19 infection results in a dysregulated inflammatory response, endothelial dysfunction, platelet activation, and a resultant prothrombotic state.^4–12,37  This acquired coagulopathy is characterized by an elevation in thrombogenesis biomarkers (ie, D‐dimer and fibrinogen levels).^38,39  Specific to MIS-C, routine as well as viscoelastic coagulation assays have revealed marked hypercoagulability and hyperfibrinolysis as compared with controls and those with COVID-19.^40,41  Prothrombotic risk from MIS-C is enhanced by systemic inflammatory-acquired organ dysfunction, including cardiomyopathy, leading to changes in blood flow states.^42,43  Similar to data reported in this manuscript, retrospective observational evidence suggests that the absence of thromboprophylaxis and presence of elevated inflammatory markers (ie, D-dimer levels) are independently associated with greater odds of HA-VTE in critically ill adults.^9,44–47  Additionally, patients with COVID-19 are at risk for clinically-relevant bleeding from consumptive coagulopathy, platelet dysfunction, and endotheliopathy.^48–51  Therefore, a balanced interpretation of clinical trial data assessing both thrombotic and bleeding risk is essential to distinguish optimal thromboprophylaxis strategies.

Recent randomized controlled trials among adults hospitalized with COVID-19 have evaluated the safety and efficacy of anticoagulation strategies, including therapeutic dosing, standard once-daily thromboprophylaxis (40 mg per day), and intermediate thromboprophylaxis (1 mg/kg per day).⁵²  When considering the safety endpoint of clinically relevant bleeding, adult trial participants receiving thromboprophylaxis experienced comparable findings to this study ranging between 0.9% to 2.3%.^52–55  In contrast, trial participants exposed to therapeutic anticoagulation generally experienced greater rates of clinically relevant bleeding, ranging from 3.8% to 8%.^54–56 

While no clinically relevant bleeding events were observed in the present pediatric phase 2 multicenter trial of enoxaparin thromboprophylaxis, 2 participants with a central venous catheter developed VTE while receiving thromboprophylaxis. This rate is consistent with adult literature; overall, the incidence of major thromboembolism among adult trial participants receiving anticoagulants during hospitalization for COVID-19 ranged from 2% to 7.5%.^53–56  The rather broad range of efficacy findings may be explained by the intensity of anticoagulant regimens studied (ie, therapeutic versus various thromboprophylaxis strategies) and the distribution of acuity or complexity of patients studied. For example, in the HEP-COVID trial conducted by Spyropoulos and colleagues,⁵³  adults hospitalized for COVID-19 with D-dimer levels >4 times the upper limit of normal were randomized to receive therapeutic versus prophylactic heparins. While the incidence of major thromboembolism and death was reduced (relative risk: 0.46, 95% CI:0.27–0.81) among non-ICU patients receiving therapeutic heparin, this was not true for adults hospitalized in an ICU setting (risk ratio: 0.92, 95% CI:0.62–1.39). Alternatively, an open-label, adaptive, randomized clinical trial among critically ill adults with COVID-19 showed reduced rates of pulmonary embolism for participants receiving therapeutic-dose heparinization versus standard-dose thromboprophylaxis (2.5% versus 7.5%).⁵⁵  Taken together, the findings from prior trials in adults with COVID-19 and those from the present report support the need for phase 3 trials to definitively evaluate the efficacy and safety of the use and intensity of enoxaparin thromboprophylaxis in children hospitalized for COVID-19-related illness.

This study has several limitations. First, variations in underlying disease treatments among and within participating sites contributed to the heterogeneity of the study population, which limit the precision of the outcome estimates reported here. This reflects real-world experience that, throughout a rapidly-evolving SARS-CoV-2 pandemic, a variety of algorithms and agents were employed for the underlying treatment of primary COVID-19 and MIS-C. Second, as study enrollment began before knowledge of the MIS-C phenotype, this subgroup analysis was posthoc and should be interpreted with caution. Third, some children may have been hospitalized in the ICU setting as part of institutional cohorting and pandemic-related infection control measures, rather than a need for escalated care or monitoring. Fourth, although the majority of participants had an elevated D-dimer at enrollment, this was not an inclusion criterion as with previously cited adult trials. Additional prospective investigations are needed to elucidate potential relationships between prothrombotic markers (D-dimer, fibrinogen), clinical presentation (MIS-C versus COVID-19), and the development of HA-VTE. The presence of mechanical thromboprophylaxis (ie, sequential compression devices) in addition to primary anticoagulant thromboprophylaxis with enoxaparin was not assessed in this trial. The standard of care for all participating centers were to apply sequential compression devices to adolescents admitted to the ICU setting unless contraindicated. We cannot account for the potential impact of concurrent mechanical prophylaxis and future trials must consider comparative effectiveness. Although extended thromboprophylaxis regimens were prescribed, thromboprophylaxis following hospital discharge and its impact on subsequent VTE development as an outpatient was beyond the scope of this trial and represents a knowledge gap for further inquiry. Lastly, differences in dose requirements for children with MIS-C may have been influenced by factors related to participant BMI (ie, obesity) and pre-enrollment enoxaparin exposure.

In this open-label, multicenter, phase 2 clinical trial, we found twice-daily subcutaneous enoxaparin as the primary thromboprophylaxis among children hospitalized for symptomatic COVID-19-related illness to be safe without evidence of clinically relevant bleeding or other serious adverse events. Further investigation is warranted to definitively assess clinical efficacy.

We thank the COVAC-TP investigators (Supplemental Information 1); the clinical research coordinators and research nurses at all participating centers for their partnership in conducting the multicenter trial; the members of the Data and Safety Monitoring Committee; the teams at the Clinical Coordinating Center for Pediatric Multicenter Studies and Data Coordinating Center for Pediatric Multicenter Studies in the Johns Hopkins All Children’s Institute for Clinical and Translational Research (St. Petersburg, FL, USA); Jade Hanson, RN, CCRP for Trial Master maintenance and monitoring; Laurel McDevitt, MS, Deborah Martino, MA, and Fatima Tariq, MS, for project management; and all patients and parents who participated in the trial.

CI

confidence interval

DVT

deep vein thrombosis

HA-VTE

hospital acquired - venous thromboembolism

IQR

interquartile range

IRB

institutional review board

ISTH

International Society on Thrombosis and Haemostasis

LMWH

low molecular weight heparin

LOS

length of stay

MIS-C

multisystem inflammatory syndrome in children