Severely Ill COVID-19 Patients May Exhibit Hypercoagulability Despite Escalated Anticoagulation

Abstract

Introduction: Severely ill COVID-19 patients receiving prophylactic-dose anticoagulation exhibit high rates of thrombosis and mortality. The escalation of anticoagulation also does not reduce mortality and has an uncertain impact on thrombosis rates. The reasons why escalated doses fail to outperform prophylactic doses in reducing risks of thrombosis and death in severely ill COVID-19 patients remain unclear. We hypothesized that escalated anticoagulation would not effectively prevent hypercoagulability and, consequently, would not reduce the risk of thrombosis and death in some severely ill patients. Methods: We conducted a prospective multicenter study that enrolled 3860 COVID-19 patients, including 1654 severely ill. They received different doses of low-molecular-weight or unfractionated heparin, and their blood coagulation was monitored with activated partial thromboplastin time, D-dimer, and Thrombodynamics. A primary outcome was hypercoagulability detected by Thrombodynamics. Blood samples were collected at the trough level of anticoagulation. Results: We found that escalated anticoagulation did not prevent hypercoagulability in 28.3% of severely ill patients at the trough level of the pharmacological activity. Severely ill patients with such hypercoagulability had higher levels of inflammation markers and better creatinine clearance compared to severely ill patients without it. Hypercoagulability detected by Thrombodynamics was associated with a 1.68-fold higher hazard rate for death and a 3.19-fold higher hazard rate for thrombosis. Elevated D-dimer levels were also associated with higher hazard rates for thrombosis and death, while shortened APTTs were not. The simultaneous use of Thrombodynamics and D-dimer data enhanced the accuracy for predicting thrombotic events and fatal outcomes in severely ill patients. Conclusions: Thrombodynamics reliably detects hypercoagulability in COVID-19 patients and can be used in conjunction with D-dimer to assess the risk of thrombosis and death in severely ill patients. The pharmacological effect of LMWH at the trough level might be too low to prevent thrombosis in some severely ill patients with severe inflammation and better creatinine clearance, even if escalated doses are used.

Keywords: LMWH; Thrombodynamics; UFH; thromboprophylaxis.

Figure 1

Distribution of hospitalized patients by admission route, hospitals, and severity of illness. The study was conducted in seven hospitals. The majority of patients were admitted from home via emergency medical services (ambulance, n = 3819), while another group of patients was transferred from other hospitals (n = 213). For 172 patients, COVID-19 was not confirmed, and they were excluded from the study. Ambul.—ambulance, Transf.—transferred.

Figure 2

Algorithm for the use of LMWHs and UFH in the treatment of COVID-19 in adult patients in an inpatient setting. Anticoagulation de-escalation was recommended for patients with kidney injury. It could also be temporarily lowered or discontinued in cases of bleeding, intubation, or surgical intervention. A table below the flowchart shows the doses of heparins used during the study. LMWH—low-molecular-weight heparin, UFH—unfractionated heparin, DVT—deep vein thrombosis, PE—pulmonary embolism, s/c—subcutaneously, TID—three time a day, BID—two times a day.

Figure 3

TDX-V distributions of mildly ill and severely ill patients receiving different doses of heparins. LMWH—low-molecular-weight heparin, UFH—unfractionated heparin. Thrombodynamics clot growth rates (TDX-V). The distributions show TDX-V values obtained from the second day of hospitalization (since TDX-V values at admission were obtained before a patient was given LMWH or UFH). Patients received prophylactic, intermediate, or therapeutic doses of heparins. Samples were collected at the trough level of heparins’ pharmacological activity (excluding UFH infusion, which was collected at least 6 h after the bolus or the dose adjustment). Boxes under the distributions show their medians and quartiles. Green areas show the normal range (20–29 µm/min). Statistical significance was assessed with the Mann–Whitney U test; levels of significance are shown to the left of the boxes. **—p-value < 0.01, ***—p-value < 0.001. (A,B) Distributions in mildly ill and severely ill patients receiving different doses of LMWH. (A) Prophylactic—458 patients with a total of 1035 measurements, intermediate—558 patients with a total of 1526 measurements, therapeutic—259 patients with a total of 593 measurements. (B) Prophylactic—144 patients with a total of 327 measurements, intermediate—452 patients with a total of 1281 measurements, therapeutic—882 patients with a total of 3939 measurements. (C) Distributions in severely ill patients receiving different doses of UFH. Distributions of mildly ill patients receiving UFH are not shown due to a small number of cases. Prophylactic—42 patients with a total of 79 measurements, intermediate—122 patients with a total of 352 measurements, therapeutic—245 patients with a total of 1151 measurements.

Figure 4

TDX-V of severely ill patients who experienced intense hypercoagulability. (A) Examples of TDX-V time courses. Five severely ill (#1–5) and one mildly ill patient (#6) are shown. Red boxes show the moments at which thrombotic complications were confirmed by instrumental methods. DVT—deep vein thrombosis, PE—pulmonary embolism. Dashed horizontal lines show the lower limit of intense hypercoagulability. Patient #5 had gastrointestinal bleeding on day 11. (B) TDX-V averages of severely ill patients who had periods of intense hypercoagulability (grey) and those who did not have them (red). Statistical significance was assessed with the Mann–Whitney U test, n = 312 (grey box) and 492 (red box). ***—p-value < 0.001.

Figure 5

Laboratory averages of severely ill patients who had intense hypercoagulability and those without it. The values of severely ill patients receiving therapeutic doses of heparins were analyzed. Black boxes show the averages of severely ill patients who had intense hypercoagulability, red boxes show the averages of severely ill patients without intense hypercoagulability. Green areas show the normal ranges. Statistical significance was assessed with the Mann–Whitney U test. **—p-value < 0.01, ***—p-value < 0.001. (A) D-dimer level, the logarithmic scale is used, n = 219 (grey box) and 308 (red box). (B) SpO₂ to FiO₂ ratio, n = 308 (grey box) and 580 (red box). (C) Lactate dehydrogenase activity, n = 212 (grey box) and 340 (red box). (D) White blood cell count, n = 339 (grey box) and 629 (red box). (E) C-reactive protein level, n = 339 (grey box) and 625 (red box). (F) Chronic kidney disease–epidemiology collaborative group equation, n = 298 (grey box) and 620 (red box).

Figure 6

Intense hypercoagulability increases the risks of death and thrombosis in severely ill patients. Kaplan–Meier curves are shown for survival (A) and thrombosis (B) probabilities in severely ill patients with different durations of intense hypercoagulability (<25% of the time before the event (death in (A) and thrombosis in (B))—red line, ≥25%—blue, and ≥50%—black).