Case report of a fatal rectal haemorrhage in a person with severe haemophilia A receiving emicizumab and high‐dose bypassing agents in the HAVEN 1 study

Dear Editor, Prophylaxis of haemophilia A (HA) requires replacement of missing activated factor VIII (FVIII) to prevent bleeding episodes. Approximately 30% of persons with HA (PwHA) who take standard treatment with FVIII concentrates develop neutralizing antibodies (known as FVIII inhibitors), which render replacement FVIII therapy ineffective.1 As such, bleeding episodes in PwHA with FVIII inhibitors require treatment with bypassing agents (BPAs), such as activated prothrombin complex concentrates (aPCC) or recombinant human factor VIIa (rFVIIa).2-5 In addition, morbidity and mortality rates tend to be higher in PwHA with FVIII inhibitors compared with those without FVIII inhibitors.6, 7 Emicizumab (HEMLIBRA®), a bispecific, humanized monoclonal antibody, bridges activated factor IX (FIXa) and FX, replacing the function of missing activated FVIII, with resultant downstream thrombin generation and activation of the coagulation cascade.8 In the Phase III, multi-centre, HAVEN 1 study of PwHA with FVIII inhibitors (NCT02622321), emicizumab prophylaxis was associated with statistically significant and clinically meaningful reductions in bleeding versus no prophylaxis.9 Due to their mode of action, all procoagulants have the potential to contribute to the development of thrombosis and thromboembolic events (TEs). Co-administration of emicizumab and aPCC has been associated with thrombotic microangiopathy (TMA) and TEs; thus, the emicizumab label describes a known risk of TE/TMA associated with on average a cumulative dose of aPCC > 100 U/kg/d for ≥ 24 hours.10 In HAVEN 1, five cases of TMA and TE were reported in people on emicizumab prophylaxis and who received on average > 100 U/kg/d of aPCC for ≥ 24 hours before the onset of event.9 Here, we report additional details to those previously published,9 of a case of a 41-year-old man with severe HA complicated by high-titre FVIII inhibitors (peak titre 4968 Bethesda units/mL) who died from a rectal haemorrhage while receiving emicizumab prophylaxis and BPA in the HAVEN 1 study. The person was diagnosed with HA (intron 9 intervening sequence T > C mutation) as a child, and there was no family history of haemophilia. Treatment was complicated by the development of FVIII inhibitors at six years of age. Previous treatment included failed immune tolerance induction and rituximab. In early June 2016, he was enrolled in the HAVEN 1 study where he received loading doses of subcutaneous emicizumab 3 mg/kg/wk for four weeks, followed by 1.5 mg/kg/wk. At time of enrolment, his treatment consisted of aPCC prophylaxis 4500 IU (48.9 IU/kg) three times per week and his annualized bleeding rate was ~ 24. His other comorbidities included hepatitis C (genotype 2b), hypertension and haemophilic arthropathy. Significantly, in 2007 (9 years prior to enrolment in the HAVEN 1 study), he also had a total colectomy, ileostomy and rectal stump formation for bowel perforation due to severe constipation from opioid medication usage. On the study, he achieved steady-state levels of emicizumab, with no anti-drug antibodies. During the study, he experienced six minor bleeds, none of which required treatment with BPA. In January 2017 (study Day 238 since commencing emicizumab prophylaxis), the patient was admitted to a hospital close to home, which was not part of the clinical trial, with a one day history of bleeding into his colostomy bag and bright red bleeding from his rectal stump. His symptoms included postural dizziness, tachycardia (heart rate 120 beats per minute) and exertional dyspnoea. On admission, his weight was 92 kg and he presented with a mild anaemia (haemoglobin [Hb] 118 g/L), normal white cell count 9.8 × 109/L and platelets which were reported as clumped, but plentiful; his creatinine was mildly elevated at 142 umol/L (estimated glomerular filtration rate [eGFR] 51 mL/min/1.73 m2), with normal lactate dehydrogenase (LDH) levels (Table 1). He self-adminstered 1.5 mg/kg emicizumab and 8 mg of rFVIIa (87 μg/kg) on the day of admission prior to coming to hospital. His last dose of aPCC prior to hospital admission had been on the 03 June 2016. His concomitant medications on presentation were as follows: methadone 80 mg daily, paracetamol 1 g four times a day, celecoxib 200 mg daily as needed, amitriptyline hydrochloride 75 mg daily and perindopril 10 mg/amlodipine 5 mg daily. Examination revealed no fissure, but there was a fresh clot in the posterior anal verge with no rectal lesion seen. Hemospray® was applied to the distal rectum. The patient, however, continued to have large volume rectal bleeding despite receiving further doses of rFVIIa for a total of 174 μg/kg on his first day of admission. There were multiple attempts to identify a bleeding site, and even with ongoing 435 μg/kg/24 hours rFVIIa treatment on the second day, the patient continued to bleed, with a drop in Hb to 79 g/L. Further examination revealed friable mucosa, with generalized ooze, with no identifiable bleeding blood vessel. The patient refused administration of any blood products due to his religious beliefs. Despite being aware of the dosing guidance (provided in October 2016) for BPA in conjunction with emicizumab, the clinical decision at the treating hospital was to commence aPCC on day three of admission (study Day 240). This was due to ongoing catastrophic bleeding while on regular rFVIIa, as the patient's haemoglobin had further dropped to 65 g/L and the patient had experienced a syncopal episode with a period of loss of consciousness. In addition, the patient had previously had a good response to aPCC. The patient underwent angiography with embolization of several rectal arteries via interventional radiology with the commencement of the aPCC. The total BPA received on day three of admission was 21 000 IU aPCC (228 IU/kg/24 hours: 9000 IU initial dose, then two doses of 6000 IU) and 261 μg/kg/24 hours rFVIIa. There was a temporary cessation of rectal bleeding. Other medication administered during this time included tranexamic acid 1 g three times per day, darbepoetin alfa 40 gμ single dose and iron polymaltose 200 mg. The patient was monitored in intensive care. Over the subsequent three days (study Days 241-243), the patient continued to receive aPCC, at a dose of 6000 IU three times per day (65 IU/kg per dose; total daily dose of 195 IU/kg/24 hours). Six days since initial presentation and ~ 48-72 hours since commencing aPCC, the patient experienced new onset abdominal pain. Of concern, the patient developed clinical and biochemical evidence of TMA, which included new onset confusion, worsening thrombocytopenia (platelet count 33 × 109/L) and anaemia (Hb 47 g/L), acute renal failure (creatinine 490 μmol/L, eGFR 12 mL/min/1.73 m2), rising LDH of 2746 U/L, and a blood film that showed the presence of schistocytes and red cell fragmentation. The patient continued to receive aPCC. Other causes of microangiopathy were excluded with a normal ADAMTS13 activity of 101% (normal range 40%-130%) and normal haptoglobin 0.6 g/L (normal range 0.3-2.0 g/L). The concern of a TMA process was raised on study Day 243, and aPCC was immediately ceased. The patient had received a total of four consecutive days of aPCC treatment (total cumulative aPCC 93,000 IU, with an average dose of 200 IU/kg/24 hours). The patient remained in anuric renal failure; plasmapheresis with 4% albumin was attempted to try and improve the symptoms of TMA on study Days 244 and 245. The choice of albumin was due to the patient's refusal of any blood products. Within 36 hours of cessation of aPCC, there was some improvement in the TMA, with the platelet count rising to 114 x109/L, improving LDH of 775 U/L, and D-dimer (XDP 0.94 mg/L). The full blood count and biochemistry results in relation to timing of presentation to hospital and administration of BPA are summarized in Table 1. Despite intensive care at this time, the patient developed a further catastrophic rectal bleed. A computed tomography angiogram showed ongoing bleeding with contrast extravasation into the rectum and devitalized rectal stump. As there was no surgical or radiological intervention possible, the patient was palliated and passed away nine days after his initial presentation. The authors believe that the cause of death was ongoing rectal haemorrhage in a patient who refused potentially life-saving blood transfusions. Although the patient did develop TMA due to the combination of emicizumab and prolonged (four consecutive days), high-dose aPCC (approximately twice the recommended dose while on combined treatment with emicizumab), the TMA was improving with the cessation of aPCC. The patient continued to refuse blood and blood products and passed away from further blood loss. The clinical and laboratory characteristics of this case of TMA are consistent with those observed in other previously reported cases in HAVEN 1.9 This case highlights the ongoing challenges in treating PwHA complicated by FVIII inhibitors. While studies have demonstrated significant reductions in bleeding episodes with emicizumab, this case highlights that life-threatening bleeds still can occur. Although the death was considered unrelated to emicizumab, careful observation of patients when receiving concomitant prophylaxis with emicizumab and BPAs is warranted. There is a need to improve awareness of the potential serious complications when combining emicizumab with aPCC, especially as emicizumab is now more widely used and available outside the clinical trial setting. As part of the post-marketing safety plan, an extended period of emicizumab monitoring is ongoing. The HAVEN 1 study was sponsored by F. Hoffmann-La Roche Ltd. Editorial assistance for this manuscript was provided by Alex Coulthard, BSc, of Gardiner-Caldwell Communications, and was funded by F. Hoffmann-La Roche Ltd. LK is an employee of NSW Health Pathology, has received honoraria from BioMarin, Takeda, and Sanofi, is on the speaker bureau of Takeda, F. Hoffmann-La Roche Ltd and Sanofi, and has received reimbursement for travel/accommodation/expenses from F. Hoffmann-La Roche Ltd; SM has participated in advisory boards for BioMarin and F. Hoffmann-La Roche Ltd; GK has received reimbursement for travel/accommodation/expenses from CSL Behring; MP is an employee of ACT Health and has received reimbursement for travel/accommodation/expenses from F. Hoffmann-La Roche Ltd; AP and AK are employees of F. Hoffmann-La Roche Ltd and hold stock/stock options with F. Hoffmann-La Roche Ltd; and TC is an employee of Genentech, Inc/F. Hoffmann-La Roche Ltd and holds stock/stock options with Genentech, Inc/F. Hoffmann-La Roche Ltd.