Thrombosis and Hyperinflammation in COVID-19 Acute Phase Are Related to Anti-Phosphatidylserine and Anti-Phosphatidylinositol Antibody Positivity

Abstract

Antiphospholipid antibodies (APLA) are strongly associated with thrombosis seen in patients with antiphospholipid syndrome. In COVID-19, thrombosis has been observed as one of the main comorbidities. In patients hospitalised for COVID-19, we want to check whether APLA positivity is associated with COVID-19-related thrombosis, inflammation, severity of disease, or long COVID-19. We enrolled 92 hospitalised patients with COVID-19 between March and April 2020 who were tested for 18 different APLAs (IgG and IgM) with a single line-immunoassay test. A total of 30 healthy blood donors were used to set the cut-off for each APLA positivity. Of the 92 COVID-19 inpatients, 30 (32.61%; 95% CI [23.41-43.29]) tested positive for APLA, of whom 10 (33.3%; 95% CI [17.94-52.86]) had more than one APLA positivity. Anti-phosphatidylserine IgM positivity was described in 5.4% of inpatients (n = 5) and was associated with the occurrence of COVID-19-related thrombosis (p = 0.046). Anti-cardiolipin IgM positivity was the most prevalent among the inpatients (n = 12, 13.0%) and was associated with a recorded thrombosis in their clinical history (p = 0.044); however, its positivity was not associated with the occurrence of thrombosis during their hospitalisation for COVID-19. Anti-phosphatidylinositol IgM positivity, with a prevalence of 5.4% (n = 5), was associated with higher levels of interleukin (IL)-6 (p = 0.007) and ferritin (p = 0.034). Neither of these APLA positivities was a risk factor for COVID-19 severity or a predictive marker for long COVID-19. In conclusion, almost a third of COVID-19 inpatients tested positive for at least one APLA. Anti-phosphatidylserine positivity in IgM class was associated with thrombosis, and anti-phosphatidylinositol positivity in IgM class was associated with inflammation, as noticed by elevated levels of IL-6. Thus, testing for non-criteria APLA to assess the risk of clinical complications in hospitalised COVID-19 patients might be beneficial. However, they were not related to disease severity or long COVID-19.

Keywords: COVID-19; anti-phosphatidylinositol antibody; anti-phosphatidylserine antibody; antiphospholipid antibodies; disease severity; long COVID-19; thrombosis.

Figure 1

APLA levels in COVID-19 inpatients. Dot plot of levels of APLA in 30 serum samples from HC (blue dots) and 298 serum samples from 92 COVID-19 patients (red dots). A cut-off for each APLA was established based on maximum APLA levels of HC: ≥2 for aA5 IgG, aA5 IgM, ab2 IgG, ab2 IgM, aCL IgM, aPA IgG, aPA IgM, aPE IgG, aPE IgM, aPG IgM, aPI IgM, aPS IgG, aPS IgM, aPT IgG, and aPT IgM; and ≥3 for aCL IgG, aPG IgG, and aPI IgG. APLA that did not reach positivity for any serum sample were aA5 IgG, aPG IgG, aPI IgG, aPS IgG, and aPT IgG. Abbreviations for APLA are: aA5, anti-annexin V; ab2, anti-β2 Glycoprotein I; aCL, anti-cardiolipin; aPA, anti-phosphatidic acid; aPE, anti-phosphatidylethanolamine; aPG, anti-phosphatidylglycerol; aPI, anti-phosphatidylinositol; aPS, anti-phosphatidylserine; aPT, anti-prothrombin; Ig, Immunoglobulin.

Figure 2

APLA levels in serum samples from 92 hospitalised COVID-19 patients. Heatmap presentation of APLA levels from 92 hospitalised COVID-19 patients and 30 healthy controls tested by LIA for 18 different APLA antibodies. APLA levels were assessed according to the template provided by the LIA manufacturer (GA GmBH): ≤1 = none, 2 = low, 3 = medium, and 4 = high. COVID-19 inpatients on the y-axis had multiple blood draws on different days during their hospital stay, and their peak APLA levels are plotted, including the peak APLA levels of the 30 healthy controls (HC on the top row of the heatmap). The APLA positivity was considered when the APLA level achieved by the COVID-19 patient was above the maximum APLA level achieved by the sera of 30 healthy controls. Thus, aCL IgG, aPG IgG, and aPI IgG were considered positive with levels ≥ 3, and for the rest of APLA with levels ≥ 2. Five APLAs did not reach APLA positivity: aA5 IgG, aPG IgG, aPI IgG, aPS IgG, and aPT IgG. Abbreviations for APLA are: aA5, anti-annexin V; ab2, anti-β2 Glycoprotein I; aCL, anti-cardiolipin; aPA, anti-phosphatidic acid; aPE, anti-phosphatidylethanolamine; aPG, anti-phosphatidylglycerol; aPI, anti-phosphatidylinositol; aPS, anti-phosphatidylserine; aPT, anti-prothrombin; Ig, Immunoglobulin Schemes follow the same formatting.

Figure 3

Laboratory data on admission for COVID-19 patients. Heatmap of laboratory data from COVID-19 inpatients on admission (A) and at 48 h of hospitalisation (B). Red, blue colour depict z-scores. Patients with a positive or negative APLA test were grouped as indicated in the top banner. Subsequently, they were grouped by sex (middle banner). Finally, the severity of COVID-19 was indicated in the bottom banner. Laboratory data values are shown in logarithmic form. No clustering of any laboratory parameter was observed except for ferritin, which clusters depending on APLA positivity and sex (see Figure 4A).

Figure 4

Blood markers associated with APLA positivity. Levels of ferritin (A) or procalcitonin (B) on admission (left column) or at 48 h of hospitalisation (right column) of patients who were grouped by whether they had any positive APLA and by their sex (male, red; female, blue). Two-way ANOVA with Tukey’s post-hoc analysis. * p < 0.05; ** p < 0.01; *** p < 0.001.