Clinico-Molecular Risk Factors and Management of Thrombosis in a Large Real-Life Cohort of Patients with Paroxysmal Nocturnal Hemoglobinuria in the New Era of Anti-Complement Inhibitors and Doacs

Abstract

The increased propensity for thrombotic events (TE) is well established in patients with paroxysmal nocturnal hemoglobinuria (PNH) and represents the main cause of disease morbidity/mortality. While the precise pathogenesis of TE remains ill-defined and likely multifactorial, cases with fully blown hemolytic PNH clones show the highest risk for such complications, often occurring in atypical localizations including the hepatic circulation (e.g., Budd-Chiari syndrome, BCS). The introduction of anti-complement therapy greatly abated hemolysis and thrombosis, demonstrating the crucial involvement of the underlying hemolytic process in triggering TE. Despite the current availability of PNH-directed treatment, a scarcity of data exist on anticoagulation and specifically on the efficacy of direct oral anticoagulants (DOACs) in this setting. Here, we took advantage of a large, real-life cohort of patients with classical hemolytic PNH to unravel determinants of thrombosis and decipher anticoagulation strategies in cases developing TE. Among 106 PNH cases (M:F ratio 0.86, median age 39 years, IQR 26-58) followed-up for a median of 68 months (95% CI 50-92), 42% were primary and 58% had antecedent aplastic anemia. Median granulocyte clone size at diagnosis was 71% (46-90) with the majority of cases classified as type III dominant red blood cells (RBCs) clone (81%) by flow cytometric analysis.1In total, 77% of cases harbored PIGA mutations by multi-amplicon deep NGS at a median ploidy-adjusted VAF of 20% (9-33), of whom 58% presenting multiple hits (clonal mosaicism).1 A total of 33 patients (31%) experienced TE, of which 84% occurred as the initial disease manifestation prompting to seek medical attention, 13% during a temporary discontinuation of anti-complement treatment (e.g., preparation to allogeneic bone marrow transplant), and only 3% (N=1) during treatment. Overall, the majority of patients had TE originating in the venous system (91%), arterial cerebrovascular district (6%) or both (3%). TE occurred typically in unusual sites with BCS as the most frequent (34%) followed by deep vein thrombosis (DVT, 20%), DVT complicated by pulmonary embolism (14%), splanchnic (14%), cerebral sinuses and superficial veins (9% each). No differences in disease ontogenesis (primary vs secondary to AA), and baseline variables such as age, PNH clone size, gender, PIGA mutations (presence, type and clonal burden), blood counts, reticulocytes, LDH, creatinine and transaminases were observed. Interestingly, patients experiencing TE were more commonly presenting with type II RBC phenotype (36% vs 13%; p=.027), had median lower albumin (4 vs 4.5 g/dL; p<.001) and elevated D-dimers (1520 vs 325 ng/mL; p=.003) as compared to non-TE cases. Patients with superficial TE were promptly started on anti-complement treatment, while the others improved with additional anticoagulation. One patient with portal vein thrombosis had severe thrombocytopenia and eventually died due to progressive BCS. Anticoagulation strategy consisted of warfarin (50%), DOACs (29%) and low molecular weight heparin (LMWH, 21%) with a median treatment duration of 44 months (17-73). LMWH was preferred in patients with thrombocytopenia (median platelets 45 vs 102x109/L in cases receiving other therapeutics; p=.03). In 1 patient, warfarin was switched to DOACs for prolonged treatment, 2 BCS cases underwent TIPS, and in 1 case with DVT an inferior vena cava filter was positioned because of the concomitant diagnosis of myeloproliferative neoplasm with MPLW515L. Recurrence of TE under anticoagulants was noted in only 2 cases receiving warfarin, whereas 7 (N=5 warfarin/ 2 DOACs) experienced bleeding. Two patients on DOACs had acute rectal hemorrhage with additional bleeding risk factors (clopidogel, ulcerative colitis). No thrombotic recurrence was observed in all 9 patients treated with DOACs at a median observation time of 24 months (12-95). While previously TE prompted lifelong anticoagulation, the control of the underlying disease with anti-complement drugs enabled its successful discontinuation in 42% of patients without any TE recurrence at a median time of 50 months (15-90). This is the largest hemolytic PNH series ever reported detailing clinico-molecular risk factors, characteristics of TE and their management in the era of anti-complement therapy and DOACs, whose use in such a setting warrants attention and further studies.