Evidence of systemic endothelial injury and microthrombosis in hospitalized COVID-19 patients at different stages of the disease

Abstract

Highlights Schistocytes are fragments of red blood cells which may be encountered in the peripheral blood smear of patients suffering from a variety of microangiopathic diseases. In the hospitalized COVID-19 patients at different stages of disease severity, a schistocyte count ≥ 1% was documented in approximately 70% of patients. Evidence of myocardial injury was observed in 87.5% of all who had a count of schistocytes ≥ 1%. Schistocytes may serve as a simple and inexpensive biomarker to identify a high-risk subpopulation with a latent systemic microvascular damage irrespective of respiratory symptoms. Severe endothelial injury and widespread microthrombosis have been recently described in postmortem examinations of coronavirus disease-2019 (COVID-19) patients [1–5]. Whether a systemic microangiopathy is present at different stages of the disease irrespective of the extent of pulmonary involvement, it has not been confirmed. Yet, a wide range of extrapulmonary clinical manifestations (e.g. thromboembolism, myocardial infarction with normal coronary arteries, kidney function impairment) has been reported in a significant number of COVID-19 patients [6]. Such manifestations usually accompany the most common respiratory symptoms, even if they appear to be unrelated to the severity of lung involvement, and may manifest at any stage of the disease, therefore suggesting the presence of an underlying systemic vascular disorder with hypercoagulability as a common pathophysiological substrate of the disease. Schistocytes are fragments of red blood cells which may be encountered in the peripheral blood smear (PBS) of patients suffering from a variety of microangiopathic diseases [e.g. disseminated intravascular coagulation (DIC), thrombocytopenic purpura]. They are the result of a mechanical damage to erythrocytes which are sheared by fibrin strands of microthrombi in the peripheral circulation. The underlying pro-thrombotic state is linked to a damage of the endothelium which promotes thrombus formation and microvascular dysfunction. In this context, the presence of schistocytes may serve as a surrogate biomarker for in-vivo assessment of a diffuse endothelial damage with formation of fibrin thrombi. We aimed at documenting the presence of schistocytes in the PBS of hospitalized COVID-19 patients at different stages of disease severity. Fourteen consecutive patients with severe acute respiratory syndrome coronavirus-2 infection confirmed by reverse-transcriptase-polymerase-chain-reaction were included in this study. The median age was 70 years (IQR: 59–76) and 85.7% were males. Baseline characteristics are reported in Table 1. Table 1 Baseline characteristics, laboratory results, drug therapy, and outcomes of the overall population (Panel A), and characteristics of 7 patients requiring mechanical ventilation (Panel B) and of 7 patients under noninvasive ventilation or high-flow nasal cannula (Panel C) Characteristic Overall (n = 14) Invasive Mechanical Ventilation (n = 7) Noninvasive Ventilation/Nasal Cannula (n = 7) Panel A  Age, median (IQR) [yr] 70 (59–76) 70 (58–76) 70 (60–75)  Male Sex, n (%) 12 (85.7) 7 (100) 5 (71.4)  Medical History   Hypertension, n (%) 10 (71.4) 5 (71.4) 5 (71.4)   Diabetes, n (%) 8 (57.1 5 (71.4) 3 (42.9)  Symptoms at Onset   Fever, n (%) 13 (92.9) 7 (100) 6 (85.7)   Cough, n (%) 11 (78.6) 5 (71.4) 6 (85.7)   Diarrhea, n (%) 3 (21.4) 1 (14.3) 2 (28.6)  Imaging Features   Ground-Glass Opacity, n (%) 13 (92.9) 7 (100) 6 (85.7)   Bilateral Pulmonary Infiltrates, n (%) 14 (100) 7 (100) 7 (100)  Laboratory Findings   White-Cell Count/mm3, median (IQR) 6,740 (5,880–10,900) 6,870 (5,790–12,600) 6,200 (5,950–9,050)   White-Cell Count/mm3 > 10,000, n (%) 4 (28.6) 3 (42.9) 1 (14.3)   Lymphocytes Count/mm3, median (IQR) 750 (667–1,052) 790 (685–1,140) 700 (545–950)   Lymphocytes Count/mm3 < 1,000, n (%) 9 (64.3) 4 (57.1) 5 (71.4)   Platelet Count/mm3, median (IQR) 196,000 (150,000–245,000) 163,000 (106,000–196,000) 229,000 (193,000–341,00)   Platelet Count/mm3 < 100,000, n (%) 2 (14.3) 2 (28.6) 0 (0.0)   LDH, median (IQR) [U/L] 370 (284–531) 295 (278–519) 398 (353–423)   LDH > 280 U/L, n (%) 10 (71.4) 4 (57.1) 6 (85.7)   Creatinine, median (IQR) [µg/L] 97 (83–135) 84 (82–103) 117 (97–146)   PT, median (IQR) [sec] 12.8 (12.3–16.8) 12.4 (11.9–13.8) 15.0 (12.7–29.0)   aPTT, median (IQR) [sec] 33.3 (31.5–35.8) 34.1 (32.3–48.9) 32.0 (29.9–35.4)   Fibrinogen, median (IQR) [g/L] 4.6 (2.8–7.26) 2.9 (1.6–5.1) 6.3 (4.6–7.9)   Fibrinogen < 1 g/L, n (%) 0 (0.0) 0 (0.0) 0 (0.0)   D-dimer, median (IQR) [mg/L] 1.91 (1.26–4.47) 2.21 (1.81–4.47) 1.21 (0.91–3.21)   D-dimer > 1 mg/L, n (%) 11 (78.6) 7 (100) 4 (57.1)  Peripheral Blood Smear   Schistocyte > 1%, n (%) 10 (71.4) 5 (71.4) 5 (71.4)  Treatment   Antibiotic Agent, n (%) 13 (92.9) 6 (85.7) 7 (100)   Antiviral Agent, n (%) 1 (7.1) 1 (14.3) 0 (0.0)   Hydroxychloroquine, n (%) 13 (92.9) 6 (85.7) 7 (100)   Corticosteroids, n (%) 4 (28.6) 1 (14.3) 3 (42.9)  Outcomes   Death, n (%) 3 (21.4) 2 (28.6) 1 (14.3) Characteristic Patient 1 Patient 2 Patient 3 Patient 4 Patient 5 Patient 6 Patient 7 Panel B  Age, yr 72 82 70 54 38 81 62  Sex Male Male Male Male Male Male Male Initial findings   Medical history Hypertension Diabetes Hyperthyroidism Hypertension Diabetes Alzheimer Hypertension Hypertension Diabetes Asthma Hypertension Diabetes Diabetes   Symptoms at disease onset Fever Dyspnea Cough Diarrhea Fever Dyspnea Syncope Fever Dyspnea Cough Fever Dyspnea Cough Fever Dyspnea Cough Fever Dyspnea Cough Fewer Dyspnea Asthenia   Imaging features Ground-glass opacity, Bilateral pulmonary infiltrates Ground-glass opacity, Bilateral pulmonary infiltrates Ground-glass opacity, Bilateral pulmonary infiltrates Ground-glass opacity, Bilateral pulmonary infiltrates Ground-glass opacity, Bilateral pulmonary infiltrates Ground-glass opacity, Bilateral pulmonary infiltrates Ground-glass opacity, Bilateral pulmonary infiltrates   Treatment Hydroxychloroquine Antibiotics Hydroxychloroquine Lopinavir-Ritonavir Hydroxychloroquine Immune Globulin Antibiotics Hydroxychloroquine Antibiotics Immune Globulin Antibiotics Hydroxychloroquine Immune Globulin Antibiotics Hydroxychloroquine Immune Globulin Antibiotics Laboratory findings   White cell count (per mm3) 6,610 23,290 4,970 6,870 13,740 11,470 3,920   Total Neutrophils (per mm3) 5,370 21,800 3,740 4,750 11,910 7,210 3,070   Total Lymphocytes (per mm3) 710 790 660 1,210 1,070 3,670 330   Total monocytes (per mm3) 270 260 400 320 540 340 360   Platelet count (per mm3) 217,000 361,000 163,000 140,000 65,000 175,000 75,000   Hemoglobin (g/L) 88 103 137 88 112 89 83   Albumin (g/L) 25 26 34 36 30 26 27   Alanine aminotransferase (U/L) 23 25 22 14 72 62 22   Aspartate aminotransferase (U/L) 20 40 35 12 5 25 9   Lactate dehydrogenase (U/L) 249 295 280 383 1,433 654 276   Creatinine (mg/dL) 1.55 1.10 0.81 0.71 0.84 0.95 0.83   High-sensitivity cardiac troponin T (µg/L) 0.042 0.073 0.010 0.009 0.033 0.009 0.046   Prothrombin time (sec) 11.07 12.41 12.34 11.49 14.93 12.75 26.62   Activated partial thromboplastin time (sec) 34.1 32.1 71.0 35.6 29.6 32.5 62.2   Fibrinogen (g/L) 8.39 6.67 1.96 3.61 1.18 2.88 1.12   D-Dimer (mg/L) 4.473 2.217 1.886 1.430 4.474 4.502 1.732   Procalcitonin (ng/mL) 0.04 0.88 0.05 0.18 0.16 0.22 0.023   High-sensitivity C-reactive protein (mg/L) 3.3 318 1.4 27.5 3.2 10.9 46.4 Peripheral blood smear   Schistocyte (%) 1–2 1–2 1–2 1–2 0.5 0.5 1–2 Characteristic Patient 8 Patient 9 Patient 10 Patient 11 Patient 12 Patient 13 Patient 14 Panel C Age, yr 75 82 70 76 42 63 58 Sex Male Male Female Male Male Female Male Initial findings   Medical history Hypertension Hypertension COPD Hypertension Diabetes Hypertension Diabetes – Hypertension Diabetes –   Symptoms at disease onset Fever Dyspnea Cough Diarrhea Fever Dyspnea Cough Diarrhea Fever Dyspnea Cough Dyspnea Cough Fever Dyspnea Cough Fever Dyspnea Cough Fever Dyspnea   Imaging features Ground-glass opacity, bilateral pulmonary infiltrates Ground-glass opacity, bilateral pulmonary infiltrates Focal airspace disease in the peripheral right midlung, discoid atelectasis at the left lung base Ground-glass opacity, bilateral pulmonary infiltrates Ground-glass opacity, bilateral pulmonary infiltrates Ground-glass opacity, bilateral pulmonary infiltrates Ground-glass opacity, bilateral pulmonary infiltrates   Treatment Hydroxychloroquine immune globulin antibiotics Hydroxychloroquine immune globulin antibiotics Hydroxychloroquine antibiotics Hydroxychloroquine antibiotics Hydroxychloroquine antibiotics Hydroxychloroquine antibiotics Hydroxychloroquine antibiotics Laboratory findings   White cell count (per mm3) 2,750 5,800 8,830 6,250 11,670 6,120 9,380   Total neutrophils (per mm3) 1,470 5,010 7,400 7,500 2,000 – 8,200   Total lymphocytes (per mm3) 690 350 900 400 700 1,000 1,500   Total monocytes (per mm3) 410 270 300 500 400 200 200   Platelet count (per mm3) 146,000 165,000 430,000 229,000 221,000 251,000 531,000   Hemoglobin (g/L) 112 127 98 128 169 96 137   Albumin (g/L) 30 28 31 36 47 30 33   Alanine aminotransferase (U/L) 110 23 10 24 48 23 144   Aspartate aminotransferase (U/L) 151 12 20 59 75 47 36   Lactate dehydrogenase (U/liter) 357 398 241 348 506 677 539   Creatinine (mg/dL) 0.72 0.98 1.61 0.97 1.35 1.50 0.67   High-sensitivity cardiac troponin T (µg/L) 0.009 0.012 0.1 0.076 0.037 0.011 0.023   Prothrombin time (sec) 12.55 10.76 35.9 13.2 – 16.8 33.1   Activated partial thromboplastin time (sec) 29.9 32 – 36.2 35.4 – 13.2   Fibrinogen (g/L) 2.80 4.26 7.46 6.27 4.97 8.39 8.59   D-Dimer (mg/L) 0.909 4.481 1.940 0.920 1.210 4.930 0.810   Procalcitonin (ng/mL) 0.08 0.08 0.5 0.1 0.09 0.3 0.01   High-sensitivity C-reactive protein (mg/L) 4.1 76.9 1.4 27.5 3.2 10.9 46.4 Peripheral blood smear   Schistocyte (%) 0.5–1 0 1–2 1–2 1–2 1–2 1–2 aPTT Activated partial thromboplastin time, IQR interquartile range, LDH lactate dehydrogenase, PT prothrombin PBSs were taken after a median of 3 days from admission (range 1–5 d) and examined by two experienced pathologists who were blinded to disease severity. The presence of schistocytes (abnormal cut-off value  ≥  1%) was microscopically evaluated following the International Council for Standardization in Hematology recommendations [7]. Patients with mechanical cardiac valvular prostheses, chronic kidney disease stage 4–5, diabetic microangiopathy, or other causes of schistocyte formation were excluded. None had required hemodialysis or extracorporeal membrane oxygenation during hospitalization. Symptoms and signs at presentation included: fever (93%), cough (79%), and diarrhea (21%). At the time PBS was performed, patients were hospitalized and had different degrees of COVID-19 severity: 7 (50%) patients had severe lung injury requiring invasive mechanical ventilation, 2 (14.3%) noninvasive ventilation, and 5 (35.7%) high-flow nasal cannula. A schistocyte count ≥ 1% was documented in 10 (71.4%) patients; one (7.1%) patient had 0.8% and 3 (21.4%) had ≤ 0.5%. The median platelet count was 196,000/mm3 (IQR: 150,000-245,000) and all but 2 (14.3%) patients had > 100,000 platelets/mm3. None had a fibrinogen level < 1 g/L and fulfilled the diagnostic criteria for overt DIC. Evidence of myocardial injury, as demonstrated by elevated levels of high-sensitive troponin T (> 0.014 µg/L), was observed in 8 (57.1%) patients, 7 of whom (87.5%) had a count of schistocytes  ≥ 1% and no preexisting history of cardiovascular disease. All 8 patients had normal left ventricular ejection fraction (EF) but one (patient 2) with reduced EF and regional wall motion abnormalities. All patients were prescribed with systemic anticoagulation, 4 (28.6%) received low-dose corticosteroids. During hospitalization, one patient had pulmonary thromboembolism (patient 10) and 3 (21.4%) died of multiorgan failure. All four patients had a count of schistocytes > 1%. At the time of discharge, PBS was repeated in other 4 patients with a previous abnormal schistocyte value [after a median of 22 days (range 16–28)], revealing a normal count in all. Hereby, we report a high prevalence (71.4%) of an abnormal count of schistocytes in the PBS of COVID-19 patients. Schistocytes were observed at any stage of disease severity, irrespective of lung involvement. Additionally, increased high-sensitive troponin T was observed in the majority of patients with schistocytes (7 out of 10; 70.0%), compared to those without. Since none of the patients fulfilled the diagnostic criteria for overt DIC and other causes of schistocyte formation were excluded, the presence of these fragments of red blood cells may imply a subclinical impairment of the endothelial cell layer of the microvasculature with formation of microthrombi in the coronary and peripheral circulation. These findings are consistent with recent studies which described endotheliitis and a systemic microthrombotic disease in patients who died from COVID-19 [1, 3]. However, our study for the first time extends the observations of post-mortem studies by correlating a similar pathophysiological substrate also to milder forms of the disease. This pattern of endothelial injury and hypercoagulability may explain the variety of clinical manifestations (e.g. kidney failure, myocardial infarction with normal coronary arteries, neurological manifestations, purpura) that has been described so far in COVID-19 patients [6]. As such, a therapeutic approach targeting the underlying endothelial dysfunction and prothrombotic state (e.g. early systemic anticoagulation, immunomodulators) may be justified at any stage of the disease to prevent clinical progression and multi-organ involvement. Additionally, schistocytes may serve as a simple and inexpensive biomarker to identify a high-risk subpopulation with a latent systemic microvascular damage irrespective of respiratory symptoms.