The use of anticoagulation on paroxysmal nocturnal hemoglobinuria. Free

Background Paroxysmal nocturnal hemoglobinuria (PNH) is a rare acquired hematological disorder with an increased risk of thrombosis. Uncontrolled complement activation plays a role in PNH-related thrombosis but many questions about the mechanism remain. Complement inhibition (CI) with complement protein C5 inhibitors reduce thrombotic complications, but CIs are not available in all countries and not all patients with PNH have an indication for CI. In the UK, treatment criteria with CI include PNH-related thrombosis, hemolysis (with anemia and lactate dehydrogenase > 1.5 x upper limit of normal), pregnancy and complications associated with hemolysis. Patients who do not have an indication for CI receive anticoagulation (AC) as primary thrombosis prevention if the PNH neutrophil or monocyte clone is over 50% and no increased bleeding risk. Patients with a history of thrombosis use AC in conjunction with CI.

Method This is a retrospective study of AC use in patients with a PNH clone at the Leeds PNH center, including patients treated with CI and patients receiving AC as primary prevention of thrombosis. Fatal and severe bleeding events were assessed, with severe bleeding defined as bleeding that is fatal, life-threatening, causes significant organ damage, or requires significant medical intervention, such as blood transfusion. We evaluated the incidence of thrombosis in patients on AC prophylaxis. Records of patients who had died since the national PNH service was commissioned in 2008 were also reviewed.

Results PNH patients were grouped by their latest granulocyte clone size or the clone size closest to: over 50%, 10-50% and less than 10%.

There were 36 patients in the over 50% group, mean age 52, mean granulocyte clone 71.6% with 24 on anticoagulation (heparin (1), warfarin (7), direct oral anticoagulants (DOAC) (16)). Seventy-seven patients were in the 10-50% clone group, mean age 53, mean granulocyte clone 21.4% with 14 on AC (warfarin (4), DOAC (10)). In the <10% group there were 199 patients, mean age 56, mean granulocyte clone 1.9% with 34 on AC (heparin (1), warfarin (14), DOAC (19)). Thirteen of these patients were on AC for non-haematological reasons.

Overall, 7 thromboses occurred (cerebrovascular accident (3), Budd-Chiari (2) cerebral venous (1) renal artery (1)) when AC was used as primary prevention, all on warfarin, over a cumulative duration of 5,708 months (1.47 events per 100 patient years).

The study cohort included 490 patients on CI; in 293/490 CI is ongoing, 106/490 discontinued CI and 91/490 who have died. In those currently on CI, the mean age was 54 and the mean granulocyte clone 80.6% with 140 being male and 153 female. Indication for CI was hemolysis (223), thrombosis (48), pregnancy (12), complications of hemolysis (6) and 4 patients treated as exceptional cases. Of these 293 patients, 69 were also receiving AC (heparin (2), warfarin (21), DOAC (46), with a further 125 having discontinued AC whilst remaining on CI.

Nine patients with a history of thrombosis, initially treated with both AC and CI, discontinued AC. There have been no further thromboses in these patients with a mean follow-up of 8 years 2 months since AC discontinuation.

DOACs were used for anticoagulation in 87 patients for a cumulative of 221.5 patient years with no thrombotic events occuring.

Two fatalities due to bleeding (gastrointestinal (1), cerebral (1)) and 11 severe bleeding episodes (gastrointestinal (3), cerebral (4), ovarian (1), traumatic (1), epistaxis (1), hemarthrosis (1)) occurred implying 0.11 fatal bleeds and 0.72 severe bleeds per 100 patient years.

Discussion The best treatment to prevent the occurrence of thrombosis in PNH is CI. However, CI is not available in many countries and may not be indicated in others unless serious disease complications occur. In our experience, AC used as primary prevention of thrombosis seems effective, with a thrombotic rate of 1.47 events per 100 patient-years which is similar to that observed in patients treated with eculizumab, at 1.07 events per 100 patient-years (Hillmen et al. 2007). Fatal and severe bleeding rates in this analysis were low, supporting the continued use of AC as primary thrombosis prevention. DOACs in this cohort seemed safe and effective.

Nine patients with a history of thrombosis discontinued AC and remained on CI without further thromboses (mean follow-up of 8 years 2 months). Whilst this is encouraging, we continue to recommend AC with CI in this setting.