Using FXIa-triggered thrombin generation to assess the risk of breakthrough bleeding in patients receiving emicizumab for acquired Hemophilia A

Abstract The treatment of patients with acquired hemophilia A (AHA) is currently changing. Immunosuppressive therapy (IST), the standard of care to suppress autoantibody formation against factor VIII (FVIII), has repeatedly been associated with a high risk of adverse events, including fatal infections. As a result, IST is now used more cautiously or even postponed until the patient has recovered from the initial illness. Instead, emicizumab is used for bleed prophylaxis until patients achieve remission of AHA. In the GTH-AHA-EMI study, emicizumab was confirmed to prevent new bleeding episodes in most patients, but 30% still had one or two bleeding events during the 12-week prophylaxis period. Clinical baseline characteristics did not predict the risk of breakthrough bleeding in these patients. The objective of the current study was to identify predictors of bleeding in AHA patients receiving emicizumab prophylaxis. We used plasma collected during the GTH-AHA-EMI clinical trial to assess thrombin generation assay (TGA) as a potential predictor of bleeding in these patients. The semi-automated calibrated automated thrombogram (CAT) was used with minute amounts of activated factor XI (FXIa) or tissue factor (TF) as trigger reagents. Analysis was focused on peak thrombin generation. Other parameters, such as endogenous thrombin generation potential (ETP) and lag time, were also recorded. Furthermore, we assessed FVIII activity and levels of other coagulation factors (FIX, FX, FXI) to investigate determinants of thrombin generation and potential predictors of bleeding. Twenty-eight patients from the GTH-AHA-EMI study provided informed consent for biobanking and had samples available one to four weeks after emicizumab loading. Baseline characteristics of this subgroup were similar to the overall GTH-AHA-EMI population, with a mean age of 75 years, 43% female, 57% male, a baseline FVIII activity of 2 IU/dl (median), and an inhibitor concentration of 14 BU/ml. The sensitivity of FXIa-triggered TGA to low levels of emicizumab and residual FVIII was higher compared to TF-triggered TGA. After emicizumab loading, the mean (range) emicizumab levels were 22.2 µg/ml (<10-49.2 µg/ml). The mean (range) TF-TGA peak was 19.5 (2.5-57.7), clearly below the reference interval (61-320 nM) determined in an age- and sex-matched control group. FXIa-TGA peak was 212.6 nM (119.5-377.5 nM, reference interval 318-624 nM). Both emicizumab levels and residual FVIII activity were significant determinants of the FXIa-TGA peak. Higher FXIa-TGA peak levels were related with a lower rate of breakthrough bleeding as indicated by an incidence rate ratio for risk of bleeding below 1 (IRR 0.40, 95% confidence interval 0.17-0.84, p<0.05). TF-TGA peak, emicizumab concentration, FVIII activity, inhibitor level, or concentrations of other coagulation factors were not significantly related to the bleeding rate. In conclusion, FXIa-TGA may serve as a surrogate marker for the risk of breakthrough bleeding in AHA patients receiving emicizumab. FXIa-TGA peak is influenced by both emicizumab level and residual FVIII activity. We suggest a simple nomogram to estimate FXIa-TGA peak from emicizumab and FVIII activity. Patients with impaired FXIa-triggered thrombin generation may be at increased risk of bleeding. Identifying this risk could support more personalized treatment decisions.