Pharmacological thromboprophylaxis is mandatory in hospitalized patients with COVID-19,
unless contraindicated.[1]
[2] However, different series reported a high incidence of thrombotic events among patients
with COVID-19 despite its use, particularly if intensive care unit (ICU) admission
was required.[3]
[4] Consequently, the use of higher doses of anticoagulants for venous thromboembolism
prevention among COVID-19 patients with additional risk factors was encouraged by
several scientific societies guidances.[5]
[6]
[7] Subsequent retrospective studies showed conflicting results regarding the efficacy
and safety of “supraprophylactic” doses.[8] Interestingly, heparins interfere with cellular invasiveness of severe acute respiratory
syndrome coronavirus 2[9] and, together with other anti-inflammatory properties, could favorably impact the
outcome of COVID-19 patients beyond the prevention of thrombotic events, particularly
if administered early in the course of the disease. This hypothesis was the rationale
for the design of numerous randomized clinical trials (RCTs) ([Supplementary Material], available in the online version).
The BEMICOP study (NCT04604327) was an investigator-initiated, open-label, multicenter,
randomized, controlled trial in patients with COVID-19 hospitalized in a conventional
ward, conducted at five Spanish hospitals. The Clínica Universidad de Navarra served
as sponsor and coordinating center. The study protocol was approved by the Agencia
Española de Medicamentos y Productos Sanitarios and the Drug Research Ethics Committee
of the Hospital Universitario Puerta de Hierro. All patients provided a written informed
consent prior to participation.
We included adult patients who required admission due to nonsevere (CURB65 ≤ 2 points
and baseline oxygen saturation ≥ 90%) COVID-19 pneumonia, with baseline D-dimer > 500 ng/mL.
The study design and the full list of eligibility criteria are provided in [Supplementary Appendix] (available in the online version).
Randomization was performed in a 1:1 ratio using a central, electronic, automated
system with permuted blocks of 4. There was no blinding of patients or investigators
to group allocation. Patients allocated to the control arm received standard prophylaxis
with subcutaneous bemiparin 3,500 IU once daily. Patients in the experimental arm
received bemiparin 115 IU/kg once daily, adjusted to body weight (7,500 IU for patients
between 50 and 70 kg; 10,000 IU for patients weighing > 70–100 kg; 12,500 IU for patients
who weighed > 100 kg). The assigned treatments were planned for a 10-day period, independently
of early hospital discharge. After that period, thromboprophylaxis use was left at
investigators' choice. In case of ICU requirement during the study treatment period,
it was at the discretion of the treating physician to continue the study drug or not,
according to local practices. Except for the assigned anticoagulant therapy, all other
clinical care was provided according to local protocols.
The primary efficacy outcome was a composite of death, ICU admission, need of mechanical
ventilation support, development of moderate/severe acute respiratory distress syndrome,
and venous or arterial thrombosis within 10 days of enrollment. Secondary efficacy
outcomes included those same endpoints separately at 10 and 30 days, as well as hospital
discharge and negativization of the polymerase chain reaction test at 10 days. Safety
outcomes were major bleeding and nonmajor clinically relevant bleeding (NMCRB), as
defined by the International Society on Thrombosis and Haemostasis[10]
[11] and any adverse event not related with COVID-19 itself. The full list of study outcomes
and definitions is provided in [Supplementary Appendix] (available in the online version). In this study, there was not an independent Endpoint
Adjudication Committee.
Assuming an incidence of the main efficacy outcome of 40% in the control group and
20% in the experimental arm, with a two-sided p-value of 0.05 and 80% statistical power, a total of 164 patients, 82 in each arm,
were needed. The study protocol included an interim analysis when 40% of the target
population was reached. After the results of this interim analysis, presented herein,
the Steering Committee decided to prematurely stop the clinical trial, based on both,
slow recruitment rate (in part related with the vaccination campaign) and futility.
Details of the statistical analysis applied are shown in [Supplementary Appendix] (available in the online version).
Between October 2020 and May 2021, 72 patients were enrolled. Six patients were excluded
due to consent withdrawal or not meeting eligibility criteria ([Supplementary Fig. S1], available in the online version). Of the remaining patients, 33 were allocated
to standard thromboprophylaxis and 33 to therapeutic-dose bemiparin. A patient in
the therapeutic-dose arm did not start the assigned treatment due to ICU transfer
before its first administration, and was excluded from primary analysis. Baseline
characteristics are shown in [Table 1]. Overall, there was a good balance between both study arms. All patients received
the study drug in accordance to the study protocol. After completion of the study
treatment period, two-thirds of patients continued extended prophylaxis for a median
of 10 additional days. No patient was lost during follow-up.
Table 1
Characteristics of patients
|
Bemiparin
3,500 IU
|
Bemiparin
115 IU/kg
|
|
N
|
33
|
32
|
|
Age (y); mean ± SD
|
62.3 ± 12.2
|
63.0 ± 13.7
|
|
Sex (male/female); n (%)/n (%)
|
24 (72.7)/9 (27.3)
|
17 (53.1)/15 (46.9)
|
|
BMI (kg/m2); median (IQR)
BMI > 30; n (%)
|
26.1 (24.1–28.8)
4 (12.1)
|
25.8 (24.0–29.4)
5 (15.6)
|
|
Comorbidities
Hypertension; n (%)
Diabetes mellitus; n (%)
Chronic pulmonary disease; n (%)
Cardiopathy; n (%)
Previous arterial or venous thrombosis; n (%)
Current or former smoking habit; n (%)
Cancer; n (%)
|
12 (36.3)
3 (9.1)
6 (18.2)
1 (3.0)
0
10 (30.3)
1 (3.0)
|
10 (31.2)
2 (6.3)
5 (15.6)
3 (9.3)
0
16 (50.0)
1 (3.1)
|
|
Days since COVID-19 diagnosis; median (IQR)
|
6 (3–8)
|
5 (2–8)
|
|
Days since symptoms onset; median (IQR)
|
8 (6–10)
|
8 (7–10)
|
|
Status at inclusion
Oxygen requirement; n (%)
D-dimer; median (IQR)
Ferritin; median (IQR)
IL-6; median (IQR)
Brescia COVID-19 score ≥2; n (%)
SIC score ≥ 4; n (%)
|
18 (54.5)
770 (590–1,030)
1,093 (514–1,751)
24.8 (5.1–57.9)
3 (9.1)
1 (3.0)
|
20 (62.5)
780 (600–1,125)
518 (287–1,248)
34.1 (15.7–77.7)
1 (3.1)
0
|
|
COVID-19 therapy
Steroids; n (%)
Statins; n (%)
Remdesivir; n (%)
Tocilizumab; n (%)
|
30 (90.9)
20 (60.6)
5 (15.2)
8 (24.2)
|
32 (100)
23 (71.9)
4 (12.5)
7 (21.9)
|
|
Extended prophylaxis
After end of study treatment; n (%)
Duration (d); median (IQR)
|
21 (63.6)
10 (7–14)
|
23 (71.9)
10 (8–14)
|
Abbreviations: BMI, body mass index; IL-6, interleukin-6; IQR, interquartile range;
SD, standard deviation; SIC, sepsis-induced coagulopathy.
Note: Absence of statistically significant differences between groups for all variables.
Study outcomes are shown in [Table 2]. The primary efficacy outcome was observed in 6/33 (18%) of patients receiving prophylactic-dose
bemiparin and in 7/32 (22%) of patients treated with therapeutic dose. No major or
NMCRB events during the study treatment period were registered. No serious adverse
event, unrelated with evolution of COVID-19, was recorded either.
Table 2
Primary and secondary outcomes
|
Outcome
|
Control (3,500 IU)
N (%)
|
Experimental (115 IU/kg)
N (%)
|
Absolute difference (95% CI)
|
Odds ratio (95% CI)
|
p-Value
|
|
Primary efficacy outcome (day 10)[a]
|
6 (18.2)
|
7 (21.9)
|
0.04 (−0.16 to 0.24)
|
1.26 (0.37 to 4.26)
|
0.95
|
|
Secondary outcomes
Death (day 10)
Death (day 30)
Need of ICU (day 10)
Need of ICU (day 30)
ATE/VTE (day 10)
ATE/VTE (day 30)
Discharge in first 10 days
Negative PCR[b] at day 10
|
0
1 (3.0)
4 (12.1)
4 (12.1)
1 (3.0)
2 (6.1)
26 (78.8)
15/25 (60.0)
|
0
2 (6.3)
4 (12.5)
5 (15.6)
0
0
21 (65.6)
18/27 (66.7)
|
–
0.03 (−0.07 to 0.13)
0.004 (−0.16 to 0.16)
0.04 (−0.13 to 0.20)
–
–
–0.13 (−0.35 to 0.08)
0.07 (−0.20 to 0.33)
|
–
2.13 (0.18 to 24.76)
1.03 (0.24 to 4.55)
1.34 (0.33 to 5.53)
–
–
0.51 (0.17 to 1.56)
1.33 (0.43 to 4.13)
|
–
0.61
1.0
0.73
–
–
0.36
0.83
|
|
Major or clinically relevant bleeding (day 10)
|
0
|
0
|
–
|
–
|
–
|
Abbreviations: ATE, arterial thromboembolism; CI, confidence interval; ICU, intensive
care unit; PCR, polymerase chain reaction; VTE, venous thromboembolism.
a The primary efficacy outcome was a composite of death, admission at ICU, need of
mechanical ventilation support, development of moderate/severe acute respiratory distress
syndrome, and venous or arterial thrombosis within 10 days of enrollment.
b Negative or cycle threshold (Ct) value > 30.
In an exploratory analysis, a significantly larger decrease in ferritin levels was
found in patients receiving prophylactic-dose bemiparin, compared with those treated
with therapeutic dose. In contrast, similar reductions of D-dimer and interleukin-6
levels were observed ([Supplementary Tables S1] and [S2], available in the online version).
Several RCTs addressing the optimal intensity of anticoagulants in hospitalized patients
with COVID-19 have been initiated.[12]
[13] In severe COVID-19 increasing the dose of heparin appears insufficient to cool down
the intense underlying inflammatory and thrombotic stimuli, as recently suggested
by the INSPIRATION and the multiplatform REMAP-CAP, ACTIV-4a, and ATTAC studies.[14]
[15]
Focusing on nonsevere COVID-19 hospitalized patients, in the ACTION trial the use
of therapeutic rivaroxaban did not improve survival or duration of hospitalization
compared with standard thromboprophylaxis.[16] Similarly, in the RAPID trial no significant differences between therapeutic or
prophylactic heparin for the composite outcome of death, need of mechanical ventilation,
or ICU admission were found.[17] On the contrary, in 2,219 noncritically ill patients included in the aforementioned
multiplatform RCT the use of therapeutic anticoagulation was associated with a 4.0%
increased probability of survival to hospital discharge without need of organ support,
although in-hospital mortality was similar (7.3% vs. 8.2%).[18] The multiplatform design favors evaluation of larger number of patients but the
nonconcurrent nature of experimental and control groups is a potential source of bias.[19]
Some factors such as the time gap between admission and treatment onset, heterogeneity
of study drugs, or other concomitant therapies could influence the results. Of note,
the use of steroids was higher in our study. We selected a relatively lower-risk population;
all patients received the same anticoagulant molecule, changing only the dose, and
the maximum time gap between admission and randomization was 2 days, shorter than
other RCTs. A short duration of the study treatment period was chosen since a high
rate of early discharge was anticipated.
Safety data seem much more uniform, instead. The risk of bleeding increases with the
use of higher-intensity anticoagulation in most studies, either RCT or large cohorts.[8]
[14]
[15]
[16]
[17]
[18]
[20] However, in the BEMICOP study no major bleeding event was recorded, in part due
to the short duration of the study treatment period.
We acknowledge some limitations. First, the open-label design and the lack of an independent
adjudication committee, although the primary outcome included a combination of objective
variables. Second, some differences between participating sites in the management
of COVID-19 patients could exist, although the distribution of concomitant therapies
was similar in both study groups. Third, the relatively short duration of study treatment
limits the evaluation of long-term impact. Finally, the limited number of patients
implies a reduction of the statistical power. However, given the low absolute number
of events in both arms, it seems unlikely that significant differences could be reached
after completion of the initially planned recruitment.
In conclusion, in COVID-19 patients hospitalized with nonsevere pneumonia but elevated
D-dimer, the use of a 10-day course of therapeutic-dose bemiparin does not seem to
improve clinical outcomes compared with prophylactic doses. Further research is needed
to evaluate other therapeutic strategies and identify subgroups of COVID-19 patients
who benefit most of them.
