Venous Thromboembolism in Hospitalized Patients With COVID-19 Receiving Prophylactic Anticoagulation

To the Editor:

Venous thromboembolism (VTE) has been reported in mechanically ventilated patients with severe acute respiratory distress syndrome resulting from SARS-CoV-2 infection (COVID-19). Two of the first 31 non-mechanically ventilated patients with moderate severity COVID-19 admitted to our hospital developed VTE while receiving uninterrupted prophylactic anticoagulation.

Patient A is a 54-year-old man with history of deep venous thrombosis of the right lower extremity following patellar realignment surgery in his mid-20s. He presented with 7 days of rhinorrhea, cough, fatigue, and fever. He did not take medications and had never smoked. Before presentation, he had received cefuroxime and azithromycin without improvement. His temperature was 39.2°C, pulse 109 beats/min, blood pressure 147/81 mm Hg, respiratory rate 22 breaths/min, and oxygen saturation 92% breathing ambient air. He was alert and speaking in full sentences. He exhibited accessory muscle use while breathing. Posterior lung auscultation revealed inferolateral crackles bilaterally. Chest x-ray showed bilateral linear and patchy parenchymal opacities with basilar predominance. Electrocardiogram showed sinus tachycardia. Laboratory tests were significant for white blood count 6930 cells/μL, absolute lymphocyte count 720 cells/μL, platelet count 238,000 cells/μL, C-reactive protein 16.9 mg/dL, D-dimer 0.5 mg/L, and ferritin 1386 ng/mL. Tests for influenza and respiratory syncytial virus were negative. Polymerase chain reaction assay for SARS-CoV-2 was positive.

He was admitted to a medical floor where he received hydroxychloroquine. His Padua prediction and modified International Medical Prevention Registry on Venous Thromboembolism (IMPROVE) risk assessment model (RAM) scores were 4 and 3, respectively.^{1 , 2} Starting on hospital day 2, he received subcutaneous heparin 5000 units every 8 hours. Between hospital days 2 and 4, the patient’s oxygen requirement rapidly increased to 7 L/min face mask oxygen, but he did not require intubation. With supportive care and twice daily proning, his condition improved such that oxygen saturation was greater than 94% on 3 L/min nasal cannula oxygen by hospital day 4. At that time, enoxaparin 40 mg daily was prescribed for VTE prophylaxis in place of subcutaneous heparin. On hospital day 5, the patient reported swelling and right calf pain while ambulating. Repeat laboratory tests revealed C-reactive protein 19.2 mg/dL, D-dimer 1.64 mg/L, and ferritin 2482 ng/mL. Venous Doppler ultrasound of the right leg showed a right popliteal and peroneal vein deep venous thrombosis. Rivaroxaban was prescribed for therapeutic anticoagulation.

Patient B is a 39-year-old man with no significant medical history who presented with 7 days of fever, nasal congestion, nonproductive cough, and myalgia. He took over-the-counter vitamins and had never smoked. Before presentation, he had received amoxicillin without improvement. His temperature was 38.3°C, pulse 114 beats/min, blood pressure 119/70 mm Hg, respiratory rate 18 breaths/min, and oxygen saturation 98% on room air. He was alert and speaking in full sentences. Lung auscultation was normal without adventitious sounds. Electrocardiogram showed sinus tachycardia. Chest x-ray showed bilateral hazy opacities in the basilar lung parenchyma. Laboratory tests were significant for white blood count 5270 cells/μL, absolute lymphocyte count 810 cells/μL, platelet count 126,000 cells/μL, C-reactive protein 2.8 mg/dL, erythrocyte sedimentation rate 25 mm/h, D-dimer 0.41 mg/L, ferritin 289 ng/mL, and troponin less than 0.04 ng/mL. Tests for influenza and respiratory syncytial virus were negative. Polymerase chain reaction assay for SARS-CoV-2 was positive.

He was admitted to a medical unit where he received hydroxychloroquine, azithromycin, and ceftriaxone. His Padua prediction and modified IMPROVE VTE RAM scores were 1 and 0 points, respectively. He received enoxaparin 40 mg daily starting on hospital day 2. Between hospital days 2 and 5, he experienced fever, dyspnea, diarrhea, and tachycardia, but he maintained oxygen saturations greater than 94% breathing ambient air. Laboratory tests obtained on hospital day 5 were notable for white blood count 8800 cells/μL, sodium 138 mmol/L, creatinine 1.0 mg/dL, C-reactive protein 20.9 mg/dL, erythrocyte sedimentation rate 100 mm/h, D-dimer level 0.77 mg/L, and ferritin 915 ng/mL. Blood cultures and stool testing for Clostridium difficile were negative. The patient’s fever, dyspnea, and diarrhea resolved by hospital day 7, but he continued to experience tachycardia. Electrocardiogram showed sinus tachycardia with ventricular rates ranging between 90 and 130 beats/min at rest. On hospital day 11, he underwent computed tomography angiogram of the chest, which showed lobar and segmental pulmonary emboli in the right lower lobe. Apixaban was prescribed for therapeutic anticoagulation.

Both non-mechanically ventilated patients hospitalized with COVID-19 developed symptomatic VTE while receiving uninterrupted standard prophylactic dose unfractionated or low molecular weight heparin. Reports linking severe COVID-19 and increased risk for VTE have emerged, including a high incidence of abnormal D-dimer levels among patients with COVID-19,³ an association between COVID-19–associated disturbances in the coagulation pathway and increased mortality,⁴ and lower 28-day mortality among patients with COVID-19 who received prophylactic heparin if they had D-dimer greater than 3.0 mg/L or sepsis-induced coagulopathy score greater than or equal to 4.⁵ Experts also have published interim summaries and guidance for managing VTE risk in patients with COVID-19, although the optimal approach for inpatient VTE risk stratification and prophylaxis remains uncertain.^{6 , 7} Our observations suggest that moderate severity COVID-19 might also predispose hospitalized patients to higher VTE risk than standard RAMs would predict, and might even precipitate acute VTE despite use of standard prophylactic-dose anticoagulant medications. It is instructive that standard VTE RAMs would have recommended thromboprophylaxis for only one of the two patients, and that neither patient met the D-dimer cut-off reported by Tang et al.⁵ Although our experience is too preliminary to draw firm conclusions, it brings into question whether current RAMs are adequate for guiding VTE thromboprophylaxis decisions in patients with COVID-19. This issue clearly warrants larger studies. In the interim, we propose that all patients hospitalized with COVID-19 receive pharmacologic VTE prophylaxis unless a contraindication to anticoagulation exists. We also suggest that the COVID-19 research agenda include studies to examine whether empiric therapeutic-strength anticoagulation improves outcomes compared with prophylactic anticoagulation, without leading to an untoward increased risk of major bleeding events.