Affinities of FVIII-specific IgG1 and IgG4 during ITI

Introduction: Elimination of inhibitory anti-FVIII antibodies (inhibitors) by immune tolerance induction (ITI) is only successful in ~70% of patients with haemophilia A. The development of inhibitors and the mechanisms underlying ITI are still poorly understood. Prediction models for ITI success are still missing. The FVIII-specific immune response including affinities is therefore analyzed, to find prognostic markers.

Methods: 107 plasma samples from 18 patients were collected during the RES.I.S.T trial, including patients with a poor prognosis for ITI success. Inhibitor titers were partly known for analyzed samples. In addition, we measured levels of FVIII-specific IgG1 and IgG4. Based on these results, the affinity of IgG1 and IgG4 in early and late plasma samples from each patient were analyzed by the method described by Friguet and modified by Bobrovnik allowing the identification of two antibody populations with differing affinities. For some samples, plasma levels of 14 cytokines were measured using a 14-Plex Bead Array.

Findings: We measured K_d values in a range from 1.1x10⁻⁸ M to 2.9x10⁻¹³ M for FVIII-specific antibodies of the subclasses IgG1 and IgG4. For some patients two different antibody populations with different affinities were measured within one sample. All patients had high affinity antibodies as soon as antibody levels allowed for the measurement of affinities.

Two plasma samples of one patient at enrolment (10.4 BU/ml) and two months during treatment (< 0.6 BU/ml) contained two FVIII-specific IgG4 populations with different affinities. The antibody population with an affinity of 5.6x10⁻¹¹ M did not change significantly, while a reduction in affinity from 3.9x10⁻¹⁰ M to 3.1x10⁻⁸ M was observed for the antibody population with lower affinity. Another patient showed fluctuant K_d values in the course of ITI in a range of 4.6x10⁻¹⁰ M to 8.1x10⁻¹¹ M. All other patients did not show differences in affinities while being BU positive.

So far, affinities of 8 plasma samples originating from 6 patients were correlated with cytokine plasma levels. Interestingly, we found a correlation (p = 0,0046) between the affinity of FVIII-specific IgG4 and the GM-CSF plasma level. No other correlations were observed.

Conclusion: Our analyses of the affinity of FVIII-specific antibodies have identified affinities in the high nanomolar range during early treatment. In the course of treatment these affinities can fluctuate, increase, or decrease and high affinity antibody populations do not necessarily disappear with the loss of the inhibitor. Correlation of further parameters including cytokine levels needs further investigation and is supported by the role of GM-CSF in B cell maturation in autoimmune diseases. Analysis of further samples will show if the determination of FVIII-specific affinities can result in the identification of prognosis markers for ITI success.