SARS-COV-2 INFECTION CAUSED THROMBOSIS in the LUPUS MODEL with ANTIPHOSPHOLIPID ANTIBODY, whereas COVID-19 ASSOCIATED THROMBOSIS WAS IRRELEVANT in PATIENTS with POSITIVE ANTIPHOSPHOLIPID ANTIBODY

ABSTRACT

Background: Thrombosis is a unique complication in coronavirus disease 2019 (COVID-19). We have reported that elevated ferritin and D-dimer on admission were the risk factors of thromboses by analyzing the patients sequentially admitted to our hospital due to COVID-19 (1). However, we have not analyzed throm-botic complications in the view of the antiphospholipid antibodies (aPLs), which are frequently detected in the COVID-19 patients. Objectives: To elucidate the thrombogenic effects of aPLs in COVID-19, we analyzed the development of thrombosis in three lupus models after SARS-CoV-2 infection. Additionally, we evaluated the association of thrombotic events and the serum profile of aPLs in Japanese patients with COVID-19. Methods: Three animal models of lupus (MRL-lpr/lpr, NZBxNZW F1 and NZW×BXSB F1) were evaluated in this study. NZW×BXSB F1 was also considered as a model of antiphospholipid syndrome (APS) since aPLs were detected with a high titer (2). Experimental SARS-CoV-2 infection was induced using mouse-passaged virus strain (3). The incidence of thromboses in the lungs and kidneys were identifed by evaluating H&E staining and PTAH staining of paraf-fn-embedded sections. We have experienced 44 thrombotic events in 34 out of 594 patients admitted to our institute. As a non-thrombotic COVID-19, 68 patients were selected to make a 1 to 2 matched-pair based on the propensity score. In total 102 patients, seven types of aPLs (anti-cardiolipin (CL) IgG/IgM, anti-β2GP1 IgG/IgA/IgM, and anti-phosphatidyl serine/prothrombin complex (PS/PT) IgG/IgM) were measured using specifc ELISA kits. The patients' clinical characteristics and serological profile of aPLs were further evaluated. Results: We identifed the development of thromboses in the lungs or kidneys in 6 out of 12 (50%) NZW×BXSB F1 mice after the SARS-CoV-2 infection, whereas no thrombosis was observed in non-infected mice. Further, there was no thrombosis in the other lupus models (0%) after the infection. These fndings might suggest the pathogenic role of aPLs under the SARS-CoV-2 infection. Among our COVID-19 patients, 39 out of 102 (38%) were tested positive for one or more aPLs. The positive ratios of any aPLs were statistically indifferent between the patients with or without thrombosis;anti-CL IgG (8.8% vs 5.9%)/IgM (0% vs 5.9%), anti-β2GP1 IgG (21% vs 12%)/IgA (8.8% vs 15%)/IgM (0% vs 1.5%), and anti-PS/PT IgG (0% vs 2.9%)/IgM (12% vs 13%), respectively. In addition, their titers were relatively lower than those observed in APS patients. The patients' characteristics and the prognosis of COVID-19 were comparable regardless of the detection of any aPLs. These fndings suggested that COVID-19 associated aPLs were irrelevant to thrombotic complications. Conclusion: Thromboses were induced after the infection of SARS-CoV-2 only in the APS model. However, aPLs detected in COVID-19 patients have little impact on the development of thrombosis. SARS-CoV-2 infection might have a high risk of thrombosis, especially in APS patients, as shown in the case report (4). The discrepancy of its thrombogenic effects of aPLs might be explained by the low titer of the antibody or the diversity of antibody epitope. Further analyses are required to clarify the mechanisms of aPLs production and the development of thrombosis in COVID-19.