Identification of Plasma Cytokine Levels Correlating with FVIII Inhibitor Titer during ITI

Introduction: Polymorphisms in genes coding for IL-10 and TNF-α have been identified as genetic risk factors for the development of inhibitory FVII-specific antibodies (inhibitors) during FVIII replacement therapy. Very limited data on cytokine levels of haemophilia A patients revealed differences in patients according to inhibitor status. Here, we longitudinally monitored cytokine profiles of patients undergoing immune tolerance induction (ITI).

Methods: A total of 107 plasma samples originating from 18 patients participating in the RES.I.S.T trial were examined, which included patients with a poor prognosis for ITI success. A Human High Sensitivity T cell Discovery Array 14-Plex (Eve Technologies Corp, Calgary,Canada) was used for quantification of 14 cytokines. FVIII inhibitor titers, Bethesda Units (BU), of the samples were available for analysis. In addition, FVIII-specific antibodies were quantified in ELISA. Bethesda titers (BU^pos) between 0.6 - < 5.0 were considered as low titer (BU^low), and BU ≥5 as high titer (BU^high). Statistical analyses were performed using GraphPad Prism 7. Mann-Whitney U tests and Spearman correlation tests were considered significant for P values < 0.05.

Results: Correlations between inhibitor titers and plasma levels of five cytokines were identified. Positive correlations were found for levels of TNFa (p = 0,014) and IL-8 (p = 0,048). Plasma levels of IL-10 (p = 0,041), IL-12 (p = 0,038), and IL-1B (p = 0,026) negatively correlated with inhibitor titers. Significantly higher plasma levels of TNFa (p = 0,016) and lower levels of IL-12 (p = 0,047) and IL-23 (p = 0,049) were measured in samples with detectable FVIII inhibitor (BU^pos) when compared with BU^neg samples. When inhibitor titers were subdivided into BU^high and BU^low, additional correlations were identified for IL-1B (p = 0,023), IL-2 (p = 0,043) and IL-17 (p = 0,036). Plasma levels of these cytokines were significantly lower in BU^high samples compared with BU^neg. Furthermore, signal intensity of FVIII-specific IgG in ELISA negatively correlated with the plasma level of IL-10 (p = 0.045).

Conclusion: The first longitudinal cytokine measurements in plasma samples of patients undergoing ITI were conducted. Correlation with inhibitor titers or the levels of FVIII-specific antibodies were found with IL-10 and TNFa - of note, gene polymorphisms of these cytokines are identified as risk factor for inhibitor development. Patients showed an increase in plasma levels of IL-12, IL-17, and IL-23 with the disappearance of their inhibitor. Interestingly, the IL-17/IL-23 immune axis has been identified to play an important role in autoimmunity, while an increase in these cytokines seems to be beneficial for inhibitor-positive haemophilia A patients. These findings may help gain a more detailed knowledge about the immune response toward FVIII and to identify prognostic markers for ITI success.