Effect of vitamin D3 on antiphospholipid antibodies in hospitalized patients with moderate to severe COVID-19

Highlights • The impact of vitamin D3 supplementation on autoimmunity remains a subject of debate.• A single dose of 200,000 IU of vitamin D3 was not able to modulate autoantibodies in COVID-19 patients.• aPL antibody positivity was not associated with thrombotic events despite vitamin D3 supplementation.• aPL antibodies associated with the virus seem to be transient in critical patients.


Introduction
Since the beginning of the Coronavirus Disease 2019 (COVID-19) pandemic, caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), a possible relationship between the infection and autoimmune reactions has been hypothesized. 13][4] One of the explanations is the fact that SARS-CoV-2 infection affects regulatory T-cell levels, aggravating inflammatory processes that may lead to autoimmunity. 5n this context, Zhang et al. 6 described three cases of thrombosis associated with both Antiphospholipid (aPL) antibodies, Anticardiolipin (aCL), and anti−β2-Glycoprotein-I (aβ2-GP).These antibodies target phospholipid-binding proteins and phospholipids in cell membranes, leading to a hypercoagulable state through interference with the coagulation cascade, activation of endothelial cells and platelets, and inhibition of natural anticoagulant pathways. 7ascolini et al. 8 showed that the frequency of autoantibody positivity in the COVID-19 group was significantly higher than in the control group (45 % vs. 12 %; p = 0.03).The subgroup of patients testing positive for autoantibodies exhibited higher lactate levels and a poorer prognosis compared to the subgroup testing negative.The mortality rate due to COVID-19 complications was significantly higher in the autoantibody-positive subgroup compared to the negative subgroup (40 % vs. 5.5 %; p = 0.03).
The role of vitamin D 3 supplementation in autoimmunity has been extensively investigated.In addition to influencing viral neutralization, recruitment of immune system cells and dendritic cell maturation, 9,10 vitamin D 3 acts on lymphocytes T and B proliferation, immune cells that participate in antibody production. 11,12For instance, Najafipoor et al. 13 demonstrated that patients who received 50,000 IU of vitamin D 3 per week for 6 months, in addition to interferon therapy, presented lower increases in IgG levels against Epstein-Barr virus 1 and viral capsid antigen compared to the control group, who only received diseasemodifying interferon therapy alone.These findings align with previous evidence on the importance of adequate serum levels of vitamin D to reduce disease activity and remission, and to enhance responsiveness to rheumatoid arthritis treatment. 14][17] Although numerous clinical trials worldwide have investigated the impact of vitamin D 3 on COVID-19 outcomes, [17][18][19][20][21] the role of vitamin D 3 on aPL antibodies remains poorly understood.To fill this gap, the present exploratory study aimed to assess the effect of a single oral dose of 200,000 IU of vitamin D 3 on aPL antibodies in hospitalized patients with moderate to severe COVID-19.

Study design
This is a post-hoc, exploratory analysis from a double-blind, placebocontrolled, randomized clinical trial performed in two centers in Sao Paulo, Brazil, and registered in ClinicalTrials.gov,NCT04449718.This study was conducted in accordance with the Declaration of Helsinki and approved by the ethics committee of both centers: Hospital das Clinicas da Faculdade de Medicina da Universidade de São Paulo and Hospital de Campo do Ibirapuera (Ethics Committee Approval Number 30959620.4.1001.0068).Before being admitted to the study, all patients provided written informed consent.This manuscript was reported according to the CONSORT guidelines.Additional information about the trial concept and design, patient recruitment, supplementation protocol and blindness, procedures, and endpoints has been previously published. 22rticipants Patients were recruited from both hospitals from June 2, 2020 to August 27, 2020.The end of follow-up was on October 7, 2020.Inclusion criteria were age ≥ 18 years; confirmed diagnosis of COVID-19 by polymerase chain reaction testing for SARS-CoV-2 from nasopharyngeal swabs, or computed tomography scan findings consistent with those found in COVID-19 (bilateral multifocal ground-glass opacities ≥ 50 %); and flu syndrome with institutional criteria for hospitalization on hospital admission (respiratory rate > 24 breaths/minute, oxygen saturation < 93 % on room air, or the presence of risk factors for complications (e.g., obesity, diabetes, systemic arterial hypertension, neoplasms, immunosuppression, heart disease, pulmonary tuberculosis), followed by COVID-19 confirmation.Patients who met these criteria were considered to have moderate to severe COVID-19 according to the criteria from the NIH-COVID-19 Treatment Guidelines. 23atients were excluded if they were unable to read and sign the written informed consent form; were already under invasive mechanical ventilation; received prior vitamin D 3 supplementation (> 1000 IU/d or weekly equivalent); had kidney failure requiring dialysis or creatinine > 2.0 mg/dL or hypercalcemia (calcium > 10.5 mg/dL); were pregnant or lactating women; were expecting to be discharged within 24 h.Absence of fever in the previous 72 h, no need for supplemental oxygen in the previous 48 h, and oxygen saturation greater than 93 % on room air without respiratory distress were used as criteria for hospital discharge.

Randomization and masking
Patients were allocated in a 1:1 ratio into the vitamin D 3 or placebo groups as previously described. 22Patients enrolled in the vitamin D 3 group received a single oral dose of 200,000 IU of vitamin D 3 diluted in vehicle (10 mL of a peanut oil solution) immediately after randomization, while those in the placebo group received only vehicle.Both solutions were prepared by the same unit (pharmacy of Hospital das Clínicas) and were identical in appearance, color, smell, and taste.This selected dose is within the recommended dose range for effectively promoting vitamin D sufficiency. 24

Procedures
Baseline demographic, self-reported anthropometric (weight and height), and clinical characteristics (coexisting chronic diseases, acute COVID-19 symptoms, patients' medications throughout hospitalization, oxygen supplementation requirement, and imaging features) were collected upon hospital admission.Serum levels of 25-hydroxyvitamin D were assessed by chemiluminescent immunoassay (ARCHITECT 25-OH Vitamin D 5P02; Abbott Diagnostics).The frequency of thrombotic events during hospitalization was assessed through electronic medical records.The presence and titers of aCL antibodies (IgG, IgM and IgA) were analyzed through commercial fluoro immunoenzymatic assay (Thermo Scien-tific™/Phadia™ 250 Immunoassay Analyzers).The aβ2-GP antibodies (IgG, IgM and IgA) were detected by ELISA using a commercial kit (Quanta Lite®, Inova Diagnostics Ins., San Diego, CA, USA).Only patients who had blood samples collected on the day of randomization and on hospital discharge were included in this study.Patients who died throughout followup were not included due to the absence of blood samples.

Outcome measures
Herein, it was reported on the following post hoc exploratory outcomes: presence and titers of aPL antibodies: aCL (IgG, IgM and IgA) and aβ2-GP (IgG, IgM and IgA).

Statistical analysis
The sample size was chosen based on feasibility and resources, as previously described. 22Independent t-test and Mann-Whitney U test were used for continuous variables.Proportions were analyzed by chisquare or Fisher's exact test.Generalized Estimating Equations (GEE) for repeated measures were used for testing possible differences in serum presence and titers of autoantibodies assuming group and time as fixed factors, with marginal and binomial distributions, and a first-order autoregressive correlation matrix to test the main and interaction effects.Bonferroni's adjustment was performed in GEE analyses to maintain a family-wise two-sided significance threshold of 0.05, considering 6 pairwise comparisons for all outcomes.All analyses were performed by a per-protocol approach with no imputation for missing, data using IBM-SPSS software, version 20.0.The significance level was set at a twosided p-value ≤ 0.05.

Results
A total of 1240 patients were screened for eligibility; however, only 240 were randomized during the acute-phase of SARS-CoV-2 infection, with 120 being assigned to each group.Three patients withdrew their informed consent.Of the 119 patients who were randomized to the vitamin D3 group, 13 (11 %) were excluded due to insufficient blood samples, and 9 (7 %) died throughout the follow-up.From 118 patients who were randomized to the placebo group, 15 (13 %) were excluded due to insufficient blood samples, and 6 (5 %) died throughout the follow-up (Fig. 1).
One hundred nine (66.5 %) patients tested positive for at least one type of aPL antibody.There was a significant group by time interaction (p = 0.046) for the frequency of aCL (IgG), with values increasing from baseline to discharge in the placebo group [from 13 (13.4%) to 25 (25.8 %), p = 0.004], while frequency of aCL (IgG) in vitamin D 3 group remained similar [from 25 (25.8 %) to 29 (29.9 %), p = 1.00], after Bonferroni's adjustment.However, the frequency of aCL (IgG) did not change between the groups on discharge.No significant differences between vitamin D 3 and placebo groups were found for titers and frequency of aCL (IgM and IgA) and aβ2-GP (IgG, IgM and IgA) antibodies (Table 2).

Discussion
To the best of our knowledge, this is the first double-blind, placebocontrolled, randomized clinical trial investigating the influence of 200,000 IU of vitamin D 3 on aPL antibodies among hospitalized patients with moderate to severe COVID-19.Overall, no significant differences between vitamin D 3 and placebo groups were found for any autoantibodies upon hospital discharge.
Antiphospholipid antibodies, such as aCL, aβ2-GP, and lupus anticoagulant, belong to a heterogeneous group of antibodies associated with Antiphospholipid Antibodies Syndrome (APS), a thrombosis-related systemic autoimmune disease affecting arteries, veins, and small blood vessels. 25][32][33] Similarly, herein the authors found that 66.5 % (129/194) of the patients showed at least one type of aPL antibody.
The immunomodulatory role of vitamin D appears to involve antigen-presenting cells, such as dendritic cells and macrophages, expressing a nuclear Vitamin D Receptor (VDR), a member of the nuclear receptor superfamily of transcriptional regulators involved in 1-alfa,25-dihidroxicolecalciferol signaling. 34However, the mechanisms underlying the inhibitory effects of vitamin D on antibody production are not fully understood.In part, this event could be explained by the regulatory effect of vitamin D on B-cells related to immune tolerance.Vitamin D sufficiency would act in the proliferation and apoptosis of activated Bcells, which are closely related to antibody production, [35][36][37] while vitamin D deficiency would play the opposite effect, increasing the production of autoantibodies through the activation of B-cells. 37,38iancho-Zarrabeitia et al. 39 demonstrated an association between circulating vitamin D insufficiency (between 10 and 30 ng/mL) and the presence of lupus anticoagulant, an aPL antibody.When evaluating only patients with aPL antibodies syndrome, they showed an association between vitamin D insufficiency and higher frequency of anti-β2-glycoprotein-I, but not aCL. 39nother study assessed the effect of 50,000 IU of vitamin D in patients with Hashimoto's thyroiditis, a condition also characterized by autoimmune inflammation.Participants received the supplement weekly for three months and had a significant reduction in serum thyroid autoantibodies levels compared to baseline. 40Accordingly, an in vitro study 41 showed that the 1,25-dihydroxyvitamin-D modestly reduced autoantibody production in peripheral blood mononuclear cells of disease-active systemic lupus erythematosus.
In patients with COVID-19, the presence of aPL antibodies has been suggested as one of the physiopathologic mechanisms for the cause of hypercoagulation and thrombotic events. 6,42In addition to promoting platelet aggregation and activation, aPL antibodies promote the upregulation of pro-inflammatory cytokines, cell adhesion molecules, and endothelial nitric oxide synthase that induce a pro-inflammatory and pro-coagulant endothelial phenotype. 43otwithstanding, controversial results exist.In a prospective singlecenter observational study, patients with positive aPL antibodies did not have an increased risk of thrombosis risk during intensive care unit hospitalization. 32Additionally, Borghi et al. 44 did not observe a significant correlation between the presence of aβ2-GP IgG and thrombosis, suggesting that aPL antibodies in COVID-19 may be different from those detectable in APS. 44Accordingly, the patients with aPL antibodies did not experience a high number of thrombotic events during hospitalization (4/133; p = 0.680), with no differences noted between the groups (p = 0.700).It is important to note that the presence of aCL in the present study was not significantly different between groups at discharge.
These conflicting results might be due to the fact that most of aPL antibodies associated with the virus are thought to be transient in patients with critical illness, 45,46 so it is unclear whether either they represent a simple epiphenomenon or are actually involved in COVID-19associated coagulopathy. 47In a study in which aPL testing was repeated 1 month after COVID-19 aPL-positive patients were admitted to the Data expressed as n (% within group) and median (IQR).Data were analyzed by Generalized Estimating Equations (GEE).aCL, Anticardiolipin; aβ2-GP, Anti-β2-Glycoprotein-I.No significant differences were observed between groups at baseline; p-value derives from unadjusted group by time interaction from F-test.
a p < 0.05 for main effect of time.
b Significantly different (p < 0.05) from discharge in both groups.
intensive care, aPL antibodies were primarily a transient phenomenon that occurred during the acute phase. 48Another confounded factor is that most patients were under anticoagulant medication, a drug that is related to the prevention of thrombotic events.This study has some strengths, such as the comprehensive assessment of antibodies related to autoimmune diseases, the randomized, controlled, double-blinded design, and the enrollment of hospitalized patients with moderate to severe COVID-19.Also, these findings increase the knowledge on the role of vitamin D 3 on aPL antibodies and better inform health professionals on more efficient healthcare delivery and resource management.
However, there are several limitations.First, the relatively small sample size could increase the chances of type 2 error.Second, the heterogeneity in pre-existing diseases and treatments of the patients, although no significant differences were observed between groups at baseline.Third, the results may have been influenced by the high rate of patients using anticoagulants (∼90 %).Finally, the authors were unable to measure other aPL antibodies, such as lupus anticoagulant antibodies.
In conclusion, these findings do not support the use of a single oral dose of 200,000 IU of vitamin D 3 to modulate autoantibodies in hospitalized patients with moderate to severe COVID-19.

a
Pardo is the exact term used in Brazilian Portuguese, meaning "mixed ethnicity", according to the Brazilian Institute of Geography and Statistics.b BMI data was missing for 7.7 % of patients (n = 7 in the vitamin D 3 group and n = 8 in the placebo group).c Anticoagulant, antibiotic, glucocorticoid, antihypertensive, proton pump inhibitor, antiemetic, hypoglycemic, hypolipidemic, thyroid and antiviral data were missing for 0.5 % of patients (n = 1 in the vitamin D 3 group).d Analgesic data was missing for 1.0 % of patients (n = 1 in the vitamin D 3 group and n = 1 in the placebo group).e Computed tomography findings data were missing for 14.4 % of patients (n = 12 in the vitamin D 3 group and n = 16 in the placebo group).f Reference values: Platelet (150-400/10 3 /µL); Mean platelet volume (9.4-12.4Fl); ESR (≤ 15 mm/h); C-reactive protein (< 5 mg/L); D-dimer (< 500 ng/mL).

Table 1
Baseline demographic and clinical characteristics.

Table 2
Effect of vitamin D 3 on autoantibodies in patients with moderate to severe COVID-19.