Major surgery management in patients with haemophilia A and inhibitors on emicizumab prophylaxis without global coagulation monitoring

Abstract

Patients with haemophilia and high-titre inhibitors were often denied surgery due to the high risk of perioperative bleeding. However, data on bypassing agents for the management of major orthopaedic surgery showed efficacy values between 81% and 85% for activated recombinant human factor VII (rFVIIa) (Novoseven®, Novo Nordisk, Bagsværd, Denmark) and between 81% and 91% for plasma-derived activated prothrombin complex concentrates (aPCC) (Giangrande et al., 2018). Recently, emicizumab (Hemlibra®, Roche, Basel, Switzerland), a humanised monoclonal antibody that bridges activated factor IX (FIXa) and FX to restore activated FVIII (i.e. a non-factor-based therapy), was approved in the USA, Australia and Europe for routine prophylaxis in adults and children with haemophilia A and inhibitors. Although in the HAVEN-1 and HAVEN-2 studies on emicizumab prophylaxis, patients with planned surgery were excluded, emergency surgeries (n = 29 in 22 patients) were performed (Kruse-Jarres et al., 2017), and two were considered as major (i.e. right knee arthroscopy with synovectomy, debridement of arthrofibrosis and chondroplasty; and appendicectomy). A few case reports (i.e. two hip replacements and one knee replacement) also have been published with different strategies and outcomes (Kizilocak et al., 2019; Santagostino et al., 2019; Seaman & Ragni, 2019) (Table 1). Severe adverse events, including thrombotic microangiopathy (TMA), thrombosis and skin necrosis, were observed when emicizumab was co-administered with aPCC (but not rFVIIa or FVIII concentrates) at cumulative doses >100 iu/kg/day and for >1 day (Oldeburg et al., 2017). Therefore, this risk is restricted to aPCC use in surgical settings (Collins et al., 2018). Here, we describe the management of major elective orthopaedic surgery in two patients with severe haemophilia A and high-responding inhibitors treated with emicizumab. Day 1: thigh haematoma with red blood cell transfusions, resolution with pdFVIII. Day 7: inhibitor at 80 BU/ml; thrombin generation not able to prevent bleeding Patient 1 (60-year-old man, historical peak titre: 223 Bethesda Units [BU]/ml, titre before surgery: 1 BU/ml) and patient 2 (60-year-old man, historical peak inhibitor titre: 96 BU/ml, inhibitor titre before surgery: 43 BU/ml) had severe bleed phenotype and advanced arthropathy. They previously received prophylaxis or on-demand treatments with standard bypassing therapies (rFVIIa and aPCC). After 6 months of emicizumab prophylaxis (3 mg/kg subcutaneously once/week for 4 weeks, followed by 1·5 mg/kg/week), both patients experienced no bleed and reported increased activities. Then, patient 1 underwent total right hip replacement, and patient 2 right knee prosthesis followed by concomitant bilateral ankle replacement after 4 months. To our knowledge, this is the first report of a simultaneous bilateral total ankle replacement in a patient with severe haemophilia A and high-responding inhibitor. Emicizumab prophylaxis was continued throughout the peri- and postoperative period (dose administered in the morning of surgery that was planned in function of the prophylaxis schedule). In patient 1 with low titre inhibitor (<5 BU/ml), plasma-derived FVIII (pdFVIII, Factane®, LFB Biomédicaments, Lille, France) was administered (100 iu/kg) before surgery and then every 8 h for 5 days. Surgery was uneventful. FVIII levels and anti-FVIII inhibitors were monitored peri- and postoperatively by chromogenic assay (FVIII chromogenic assay, Siemens, Healthcare Diagnostics, Marburg, Germany, CS 5100 Systex SAS, Villepinte Roissy, France) (Table 2). In patient 2, rFVIIa was given because of persistent high-titre inhibitor and the recommendation against co-administering aPCC with emicizumab (Collins et al., 2018). For both surgeries, a bolus injection of rFVIIa (90 µg/kg) was given immediately before the skin incision, followed by a bolus injection every 2 h for 48 h. Then, the dosing interval was increased to 3 h, but had to be reduced to 2 h because of ankle swelling that decreased rapidly without need of any invasive procedure. The dosing interval was increased to 3 h. No coagulation test was performed. Both patients did not show any sign of TMA. No antifibrinolytic was used. Both patients were discharged at day 7 after surgery. Emicizumab was continued alone during the rehabilitation period that was uneventful. No bleeding occurred, no blood transfusion was necessary, and intraoperative blood losses were similar to patients without blood disorders. There was no evidence of thrombosis, and no thromboprophylaxis was given. No adverse event was recorded. When rFVIIa is used alone, different dosing regimens (35–220 µg/kg) have been recommended (Mingot-Castellano et al., 2016). In the first prospective randomised study, the dose of 90 μg/kg every 2 h from skin incision/intubation and for the following 48 h and every 2–6 h for an additional 3 days resulted in successful haemostasis in all patients (60% of success with 35 µg/kg). This dose was recommended as a first-line choice for surgical procedures (Abshire & Kenet, 2008), but was gradually increased to 120 and 180 μg/kg following other studies. Moreover, the incidence of thrombosis associated with 90 µg/kg rFVIIa was about 4/100 000 infusions (Abshire & Kenet, 2008). Here, we used 90 µg/kg for co-administration with emicizumab (patient 2). For patient 1, we chose pdFVIII. The anamnestic response is usually observed after 4–10 days. In the case of early anamnestic response with bleeding, treatment should be switched to rFVIIa. To date, no routine coagulation test is recommended, although monitoring such drugs seems important during the peri- and postoperative period, mainly to avoid complications (Mancuso & Fasulo, 2016a). Global coagulation assays, such as thrombin generation assays (TGA) and rotational thromboelastometry, have been proposed to support treatment choice for patients with high-responding inhibitor but are controversial (Mancuso et al., 2016b). Recently, Dargaud et al. used TGA to limit the adverse events in a patient with perirenal haematoma treated with emicizumab associated with other haemostatic drugs (Darguaud et al., 2018). Kizilocak et al. (2019 ) demonstrated the utility of preoperative TGA for tailoring bypassing therapy in a 25-year-old man who underwent knee arthroplasty before the recommendation of not using emicizumab with aPCC. Conversely, TGA could not predict the occurrence of bleeding complications in a patient who underwent surgery and was treated with emicizumab and rFVIIa (Santagostino et al., 2019). Morover, TGA does not reflect the fibrin clot stability, which is essential for complete haemostasis. Whole blood thromboelastography that records coagulation and fibrinolysis would be more suitable, but lacks standardisation (Mancuso & Fasulo, 2016a; Mancosa et al., 2016b). In conclusion, emicizumab represents an interesting strategy for coagulation management during major surgery in patients with high-responding inhibitor. The value of routine monitoring of bypassing agents and of new drugs in patients with inhibitors needs to be better assessed. The authors thank very much Mrs Elisabetta Andermacher for her assistance with English revision. They report no grants or funds. C. Biron-Andréani collected, assembled the data, analysed and interpreted the results, wrote the manuscript and approved the final version; I. Diaz-Cau and A. Ranc analysed and interpreted the biological results and approved the final version; C. Leonardi, M. Dischino, R. Guy, R. Navarro, C. Garcia-Gournay analysed and interpreted the results and approved the final version. E. Santagostino and J.F. Schved provided critical revision of the article and approved the final version.

Topics & Concepts