Antiphospholipid antibodies and neurological manifestations in acute COVID-19: A single-centre cross-sectional study

BACKGROUND: A high prevalence of antiphospholipid antibodies has been reported in case series of patients with neurological manifestations and COVID-19; however, the pathogenicity of antiphospholipid antibodies in COVID-19 neurology remains unclear. METHODS: This single-centre cross-sectional study included 106 adult patients: 30 hospitalised COVID-neurological cases, 47 non-neurological COVID-hospitalised controls, and 29 COVID-non-hospitalised controls, recruited between March and July 2020. We evaluated nine antiphospholipid antibodies: anticardiolipin antibodies [aCL] IgA, IgM, IgG; anti-beta-2 glycoprotein-1 [aβ(2)GPI] IgA, IgM, IgG; anti-phosphatidylserine/prothrombin [aPS/PT] IgM, IgG; and anti-domain I β(2)GPI (aD1β2GPI) IgG. FINDINGS: There was a high prevalence of antiphospholipid antibodies in the COVID-neurological (73.3%) and non-neurological COVID-hospitalised controls (76.6%) in contrast to the COVID-non-hospitalised controls (48.2%). aPS/PT IgG titres were significantly higher in the COVID-neurological group compared to both control groups (p < 0.001). Moderate-high titre of aPS/PT IgG was found in 2 out of 3 (67%) patients with acute disseminated encephalomyelitis [ADEM]. aPS/PT IgG titres negatively correlated with oxygen requirement (FiO(2)R=-0.15 p = 0.040) and was associated with venous thromboembolism (p = 0.043). In contrast, aCL IgA (p < 0.001) and IgG (p < 0.001) was associated with non-neurological COVID-hospitalised controls compared to the other groups and correlated positively with d-dimer and creatinine but negatively with FiO(2). INTERPRETATION: Our findings show that aPS/PT IgG is associated with COVID-19-associated ADEM. In contrast, aCL IgA and IgG are seen much more frequently in non-neurological hospitalised patients with COVID-19. Characterisation of antiphospholipid antibody persistence and potential longitudinal clinical impact are required to guide appropriate management. FUNDING: This work is supported by UCL Queen Square Biomedical Research Centre (BRC) and Moorfields BRC grants (#560441 and #557595). LB is supported by a Wellcome Trust Fellowship (222102/Z/20/Z). RWP is supported by an Alzheimer's Association Clinician Scientist Fellowship (AACSF-20-685780) and the UK Dementia Research Institute. KB is supported by the Swedish Research Council (#2017-00915) and the Swedish state under the agreement between the Swedish government and the County Councils, the ALF-agreement (#ALFGBG-715986). HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532), the European Research Council (#681712), Swedish State Support for Clinical Research (#ALFGBG-720931), the Alzheimer Drug Discovery Foundation (ADDF), USA (#201809-2016862), and theUK Dementia Research Institute at UCL. BDM is supported by grants from the MRC/UKRI (MR/V007181/1), MRC (MR/T028750/1) and Wellcome (ISSF201902/3). MSZ, MH and RS are supported by the UCL/UCLH NIHR Biomedical Research Centre and MSZ is supported by Queen Square National Brain Appeal.