AUTHORS
Anna Schwartz, BS, Brian Boylan2, Brandi Dupervil, DHSC, MPH3, Bhavya S Doshi, MD4
AFFILIATIONS
1 Emory College of Arts & Sciences, Atlanta, Georgia, United States
2 Centers for Disease Control & Prevention, Atlanta, Georgia, United States
3 Centers for Disease Control & Prevention, Atlanta, Georgia, United States
4 Aflac Cancer and Blood Disorders Center, Children's Healthcare of Atlanta, Atlanta, Georgia, United States, Division of Hematology/Oncology, Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia, United States
ABSTRACT
Background: Inhibitors to factor VIII (FVIII) remain the most significant complication of treatment in hemophilia A. Emicizumab is widely used for hemostatic prophylaxis in severe hemophilia A patients with inhibitors due to demonstrable decrease in bleeding rates and ease of administration compared to conventional bypassing agent (BPA) therapies. The improvement in bleeding rates on emicizumab has raised debate about the pursuit of immune tolerance induction (ITI) in all inhibitor patients. Whether and how emicizumab use in place of ITI will affect the ongoing immune response to FVIII and outcomes for patients are unknown. Understanding the change in anti-FVIII IgG subclasses over time in these patients could inform their risk of anamnesis, refine tolerance definitions, and guide ITI utilization.
Aim: To conduct a longitudinal analysis of anti-FVIII IgG subclasses in severe hemophilia A patients with and without FVIII inhibitors receiving emicizumab stratified by FVIII exposures and immune tolerance status to determine the risk of anamnesis to FVIII.
Methods: A retrospective cohort study was performed using laboratory and associated clinical data from Community Counts, including Inhibitor Lab Results, Specimen Forms, and Initial and Annual Visit Forms collected between 2014 and 2023. Severe hemophilia A patients enrolled in Community Counts before initiation of emicizumab were included. Those with ongoing ITI or unclear inhibitor history were excluded. Participants were stratified by inhibitor and ITI statuses into Group A (no inhibitor), Group B (inhibitor and history of ITI; B1: tolerized, B2: non-tolerized), or Group C (inhibitor but no history of ITI; C1: positive Bethesda titer, C2: negative Bethesda titer at emicizumab initiation). FVIII IgG subclasses were measured by median fluorescence intensity (MFI) via fluorescence immunoassay and normalized to negative control. Inhibitor titers were measured by chromogenic Nijmegen-modified Bethesda assay in chromogenic Bethesda units (CBU)/mL. Descriptive statistics by group were calculated and compared by Kruskal-Wallis with Dunn’s correction. Baseline and final FVIII IgG1 and IgG4 were compared by Wilcoxon signed-rank tests. Fisher’s exact test was used to compare proportions between tolerized (B1) and non-tolerized (B2, C1, and C2) groups.
Results: We identified 1004 participants categorized as: 661 (65.8%) Group A, 142 (14.1%) B1, 147 (14.6%) B2, 33 (3.3%) C1, and 21 (2.1%) C2. The median age at emicizumab initiation was higher in group A (16.2, 10.0-29.0 years) compared to group B2 (11.9, 7.5-21.3 years), p=0.001. Inhibitor titer (mean ± SD) at emicizumab initiation was highest in B2 (21.8 ± 78.4) followed by C1 (5.0 ± 5.2) and negative in A (0.1 ± 0.2), B1 (0.1 ± 0.1), and C2 (01 ± 0.2), p<0.001. As expected, baseline IgG1 and IgG4 for group A (non-inhibitor) and B1 (tolerized patients) were lower than those with inhibitors in groups B2 and C1 (p<0.001). IgG4 was higher in group C2 vs A (p<0.001) and B1 (p<0.05), suggesting that IgG4 is still detectable even without a positive Bethesda titer in patients who have not undergone ITI.
Group B2 IgG1 and IgG4 decreased from baseline at emicizumab start to last sample (p=0.004, p<0.001, respectively). In groups A and B1, there was no significant difference from baseline to follow-up IgG1 by FVIII exposure status. In the non-tolerized groups, a higher proportion had increase in IgG4 with FVIII exposure (60 vs 12%, p<0.001) correlating with a smaller overall decrease in IgG4 levels (-116 vs -267 fold of cutoff MFI, p=0.05) and Bethesda titer (-4.6 vs -12.2 CBU/mL, p=0.08). Among FVIII exposed patients, the rate of increased IgG4 was higher in non-tolerized patients compared to tolerized patients (60 vs 23.5%, p<0.05). Of the 12 non-tolerized patients with increased IgG4 following FVIII exposure, 4 (33%) had an increase in Bethesda titer.
Conclusions: Our data support that IgG1 and IgG4 correlate with inhibitor and tolerance status in patients with hemophilia A. While there was an overall decrease in IgG1 and IgG4 in non-tolerized patients on emicizumab therapy, a higher proportion of these patients had increase in IgG4 following FVIII therapy and 1/3 had a rise in Bethesda titer. These data suggest an ongoing risk of anamnestic response in non-tolerized hemophilia A patients.
DISCLOSURES
Anna Schwartz, BS: No relevant conflicts of interest to declare.