Cytokine Profiling in Antiphospholipid Syndrome Demonstrates Persistent Immune Dysregulation and a Procoagulant Phenotype

Background: The antiphospholipid syndrome (APS) is an autoimmune thromboinflammatory disease characterized by thrombosis and the persistent detection of antiphospholipid antibodies (aPL). The rate of recurrent thrombosis in APS is between 3 and 24% despite standard of care anticoagulation with a vitamin K antagonist, suggesting a persistent prothrombotic milieu in APS which may be driven by altered cytokine signaling.

Objective: To test the hypothesis that cytokine signaling is dysregulated in APS and contributes to thrombotic risk, we profiled cytokine levels in patients with thrombotic APS and healthy individuals.

Methods: We measured a panel of 46 serum cytokines using the Meso Scale Discovery electrochemiluminescent platform, a validated measurement system that provides linear readouts of analytes over several logs of concentration, in 21 APS patients and 12 healthy individuals. All APS patients met the Revised Sapporo Criteria for thrombotic APS. Measurements were performed in triplicate and the log2 transformed averages were analyzed, setting significance at a FDP-adjusted p value of <0.05, while controlling for age and sex. Cytokine values were correlated with clinical history and antibody isotype profile.

Results: Our cohort was 67% female and 85% identified as white. All APS patients had thrombotic disease: 57% with history of VTE, 19% with ATE, and 24% with both VTE and ATE. 14% had catastrophic APS. 67% were triple-positive for anti-cardiolipin (aCL), anti-β2 glycoprotein 1 (aβ2GP1), and lupus anticoagulant (LAC). 10% had isolated IgM isotypes of aCL or aβ2GP1.

We found significantly elevated levels of TNF-α, IL-6, IL-7, IL-8, IL-10, IL-16, IL-12/IL-23p40, soluble VCAM-1, soluble ICAM-1, PIGF, SAA and CRP and significantly decreased levels of IL-17genB, IL-22 and IL-31 in APS compared with healthy controls (Figure 1). Cytokine levels were measured a median of 5 months (1-240) from the last thrombotic event, demonstrating persistence of the inflammatory response.

To determine if cytokine levels were predictive of thrombotic risk, we correlated the number and pattern of positive aPL isotypes and LAC with cytokine levels. We found that TNF-α, IL-8, IL-10, sVCAM-1, sICAM-1, VEGF, IL-7, IL-12/IL-23p40, IL-4, IL-6, IL-16, PIGF, IL-23 and IL-17A each correlated with aPL positivity, with higher levels in triple-positive APS, and the highest levels observed in patients positive for all five measured antibody isotypes (aCL IgG, IgM, IgA; aβ2GP1 IgG, IgM) plus the functional LAC (Figure 2). IL-31 level inversely correlated to the number of positive aPL tests.

A sub-analysis of cytokines in patients with isolated IgM aCL or aβ2GP1 antibodies with or without LAC showed significantly increased levels of TNF-α compared to healthy controls, and did not differ from APS patients with other patterns of aPL positivity. This suggests that a similar proinflammatory or procoagulant background is present in patients with isolated antiphospholipid IgM antibodies as those with IgG.

Discussion: The sustained differences in cytokine levels measured in APS patients suggests a picture of ongoing immune dysregulation and compensation capable of driving vascular cell activation, and tipping the hemostasis balance into thrombosis. We find upregulation of proinflammatory cytokines, such as TNF-α and IL-6, capable of activating endothelial cells and triggering amplification loops leading to chemokine and acute phase reactant release, cell adhesion molecule expression, recruitment of leukocytes, and accumulation of complement and coagulation proteins, with the net effect of driving inflammation and thrombosis. Decreased levels of IL-17, IL-22 and IL-31 are indicative of dysregulation of the Th17 CD4+ T cell response, which has been implicated in autoimmune pathogenesis. The correlation of cytokine levels with triple or greater aPL positivity suggests a model for advancing risk stratification and has utility as a biomarker of thrombosis. Our study, although limited by size, represents one of the most extensive profiling of cytokines in thrombotic APS.