IL-6 Autoantibodies Predict Lower Platelet Counts and Altered Plasma Cytokine Profiles in Healthy Blood Donors: Results From the Danish Blood Donor Study

Abstract

Cytokine-specific autoantibodies (c-aAb) represent a novel type of immune dysfunction. Though they have been detected in both patient cohorts and healthy individuals, and have immunomodulatory properties, the full extent of their influence remains unknown. Based on the critical role of several cytokines in thrombopoiesis, we investigated if there is an association between c-aAb and platelet variables in healthy individuals, with a specific focus on c-aAb against a known thrombopoietic cytokine, IL-6. Using platelet count and mean platelet volume in 3,569 healthy participants of the Danish Blood Donor Study as dependent variables, we performed a series of multivariate regression analyses using five cytokine autoantibodies, including IL-6 c-aAb, as independent variables. In men, high titers of IL-6 c-aAb were negatively associated with platelet counts (β = -24 ^*10⁹/l (95% confidence interval -43 to -6), p = 0.008) and positively associated with mean platelet volume (β = 0.4 fL (95% confidence interval 0.0-0.7) p = 0.043). These associations were exacerbated when adjusting for undetectable C-reactive protein levels, which we used as a proxy for c-aAb mediated IL-6 inhibition in vivo. Furthermore, in a smaller subgroup, individuals with high vs. low titer IL-6 c-aAb had different profiles of plasma IL-6, IL-10, TNFα and TPO, further suggesting a functional inhibition of IL-6 by high titers of circulating IL-6 c-aAb. We therefore speculate that in addition to their immunomodulatory potential IL-6 c-aAb may interfere with thrombopoiesis - directly or indirectly - under normal physiological conditions. This study is the first to suggest an influence of c-aAb on platelets in healthy individuals, beyond their apparent effects on immune competence.

Keywords: IL-6; autoimmunity; blood donors; cytokine autoantibodies; cytokines; epidemiology; platelets; thrombopoiesis.

Figure 1

Association between IL-6 c-aAb and CRP levels with platelet count and MPV. Multivariate linear regression models were used to investigate levels of IL-6 c-aAb and CRP as predictors of (A) platelet count and (B) MPV. Analyses were performed for all combinations of c-aAb and CRP levels as independent variables, with individuals with low titer c-aAb and detectable CRP used as the baseline. High titers of c-aAb were defined as MFI above the 99th percentile, intermediary c-aAb titers were defined as MFI values > negative control + 4SD and < the 99th percentile, and low titers of c-aAb were defined as below the intermediate titers. Separate c-aAb titers were calculated for women and men. Undetectable CRP was defined as having CRP < 0.1 mg/L, and detectable CRP as having CRP ≥0.1 mg/L. Other independent variables include titers of c-aAb, age, smoking, BMI, oral contraceptives, 1-year prescription history of antimicrobials, 1-year donation history and comorbidity. For binary variables usage of contraceptives and having had at least one prescription a year before c-aAb measurement were coded as 1. Data are presented as regression coefficients with 95% Confidence Interval. * denotes a p value of ≤ 0.05, ***p ≤ 0.001.

Figure 2

Association of IL-6 c-aAb and CRP levels with having low or high platelet variables. Multivariate logistic regression models were used to investigate levels of IL-6 c-aAb and CRP as predictors of the odds of having (A) low platelet count and (B) high MPV. Analyses were performed for all combinations of c-aAb and CRP levels as independent variables, with individuals with low titer c-aAb and detectable CRP used as the baseline. Low or high levels of platelet count and MPV were defined as scores below or above the 10th and 90th percentiles, respectively. High titers of c-aAb were defined as MFI above the 99th percentile, intermediary c-aAb titers were defined as MFI values > negative control + 4SD and < the 99th percentile, and low titers of c-aAb were defined as below the intermediate titer. Separate c-aAb titers were calculated for women and men. Undetectable CRP was defined as having CRP <0.1 mg/L, and detectable CRP as having CRP ≥0.1 mg/L. Other independent variables include titer of c-aAb, age, smoking, BMI, oral contraceptives, 1-year prescription history of antimicrobials, 1-year donation history and comorbidity. For binary variables usage of contraceptives and having had at least one prescription per year before c-aAb measurement were coded as 1. Data are presented as OR with 95% Confidence Interval (CI). ** denotes a p-value of ≤ 0.01.