Systemic thrombosis in a large cohort of COVID-19 patients despite thromboprophylaxis: A retrospective study

Abstract

Background: Incidence of thrombotic events associated to Coronavirus disease-2019 (COVID-19) is difficult to assess and reported rates differ significantly. Optimal thromboprophylaxis is unclear.

Objectives: We aimed to analyze the characteristics of patients with a confirmed thrombotic complication including inflammatory and hemostatic parameters, compare patients affected by arterial vs venous events and examine differences between survivors and non-survivors. We reviewed compliance with thromboprophylaxis and explored how the implementation of a severity-adjusted protocol could have influenced outcome.

Methods: Single-cohort retrospective study of COVID-19 patients admitted, from March 3 to May 3 2020, to the Infanta Leonor University Hospital in Madrid, epicenter of the Spanish outbreak.

Results: Among 1127 patients, 80 thrombotic events were diagnosed in 69 patients (6.1% of the entire cohort). Forty-three patients (62%) suffered venous thromboembolism, 18 (26%) arterial episodes and 6 (9%) concurrent venous and arterial thrombosis. Most patients (90%) with a confirmed thrombotic complication where under low-molecular-weight heparin treatment. Overt disseminated intravascular coagulation (DIC) was rare. Initial ISTH DIC score and pre-event CRP were significantly higher among non-survivors. In multivariate analysis, arterial localization was an independent predictor of mortality (OR = 18, 95% CI: 2.4-142, p < .05).

Conclusions: Despite quasi-universal thromboprophylaxis, COVID-19 lead to a myriad of arterial and venous thrombotic events. Considering the subgroup of patients with thrombotic episodes, arterial events appeared earlier in the course of disease and conferred very poor prognosis, and an ISTH DIC score ≥ 3 at presentation was identified as a potential predictor of mortality. Severity-adjusted thromboprophylaxis seemed to decrease the number of events and could have influenced mortality. Randomized controlled trials are eagerly awaited.

Keywords: Coronavirus infections; Disseminated intravascular coagulation; Heparin, low-molecular-weight; Thrombophilia; Thrombosis.

Fig. 1

Human picture depicting thrombotic events in a 1127-patient COVID-19 cohort.

Fig. 2

Representative radiological images of the different thrombotic events. A. Floating thrombi in the aortic arch. CT pulmonary angiography (CTPA) shows two round filling defects (arrowheads). B. Pulmonary embolism, CTPA, Coronal Multiplanar Reconstruction (MPR). Partial filling defects in the right interlobar/basal trunk, right superior lobar artery and left superior lobar artery (arrowheads). Peripheral bilateral consolidations caused by COVID-19 viral pneumonia. C. Basilar artery thrombosis. Circle of Willis CT angiography, Coronal Volume Rendering Technique (VRT). Non-visualization of the basilar artery (arrowhead) due to basilar artery occlusion. D. Portal vein non-occlusive thrombosis. Single-phase contrast enhanced abdominal CT. Partial filling defect in the posterior branch of the right portal vein (arrowhead). No hepatic infarction was demonstrated. E. Splenic artery thrombosis. Single-phase contrast-enhanced abdominal CT shows a large splenic infarction (arrow) secondary to a large filling defect in the splenic artery (arrowheads). F. Renal artery thrombosis. Single-phase contrast-enhanced abdominal CT, Coronal MPR. Subtotal infarction of the left kidney (arrow) caused by a large thrombus in the left renal artery (arrowhead).

Fig. 3

Acro-ischemic lesions. A. Acro-ischemic lesions located mainly in the distal side of the right index finger. The first, fourth and fifth fingers were also involved. B. Ischemic changes in acral skin with confluent epidermal necrosis, subepidermal edema, vascular ectasia, and no relevant dermis inflammation (H&E stain, 10×). C. Mild perieccrine inflammatory infiltrate and necrosis of the eccrine glands. (H&E stain, 20×). D. Focal thrombosis (arrow) in papillary dermis capillaries (H&E stain, 40×). E. The small thrombi (arrows) were highlighted by immunostaining to anti-von Willebrand's factor (anti F VIII stain, 40×). H&E: Hematoxylin and eosin.