Effect of the complement inhibitor eculizumab on thromboembolism in patients with paroxysmal nocturnal hemoglobinuria

Intravascular hemolysis and cell-free plasma hemoglobin have been implicated in the serious clinical sequelae of various hemolytic disorders.¹  Hemolysis is the primary clinical manifestation of the uncommon disease paroxysmal nocturnal hemoglobinuria (PNH) and has been shown to result in chronic disabling morbidities including anemia, severe fatigue, difficulty in functioning, pain, and thrombosis, all of which have a major effect on the patient's quality of life.^1-5  Paroxysmal nocturnal hemoglobinuria is defined by the acquired genetic deficiency of glycosylphosphatidylinositol (GPI)–linked proteins from the surface of blood cells. The absence of GPI-linked complement regulatory proteins on PNH erythrocytes renders them susceptible to terminal complement-mediated hemolysis.

Hemolysis most likely contributes to thromboembolism (TE) in PNH, as patients with larger PNH clones have a higher incidence of TE and events have been temporally associated with increased hemolysis.^6-10  Although the mechanism is not fully understood, hemolysis has been implicated in the initiation of platelet activation and aggregation.¹  Additional in vitro studies have suggested that complement may directly activate platelets from PNH patients.^11,12 

Thromboembolism is the leading cause of mortality in patients with PNH,^2,3,13-17  and an initial thrombotic event increases the relative risk of death in PNH 5- to 10-fold.^15,17  Retrospective studies suggest that, in non-Asian patients, TE accounts for approximately 40% to 67% of deaths with known causes.^13-17  Further, 29% to 44% of patients with PNH have been reported to suffer from at least one TE event in the course of their disease.^13-16 

In a retrospective study of 67 high-risk patients with PNH not taking prophylactic anticoagulants, the TE rate was estimated at 3.7 events per 100 patient-years (19 TE events in 511.5 patient-years), while 0 TE events in 117.8 patient-years was observed in a heterogeneous group of 39 patients with PNH treated with anticoagulants as primary prophylaxis.⁸  However, in other reports, new TE events as well as progression of existing TE have been observed in patients with PNH, despite the use of anticoagulants and/or antiplatelet agents.^7,14  The risk for fatal hemorrhage in patients with PNH is significant, in part due to the frequent occurrence of thrombocytopenia.^8,14 

Eculizumab (Soliris) is a humanized monoclonal antibody that targets complement protein C5, thereby preventing assembly of the terminal complement complex (also called the membrane attack complex or MAC) during complement activation. In clinical studies to date, eculizumab treatment reduced hemolysis and improved anemia, fatigue, and quality of life in patients with PNH.^18,19  In this report, we evaluate the effect of long-term treatment with eculizumab on the prespecified clinical outcome of TE on an intention-to-treat basis in a multinational phase 3 open-label extension study that enrolled patients from 3 independent eculizumab PNH clinical studies (n = 195). We explore the robustness of the effect of eculizumab treatment on TE in subgroup analyses.

The number of TE events and the incidence rates were determined in patients from 3 independent parent clinical studies (the phase 2 pilot study and its extensions,¹⁹  the phase 3 TRIUMPH study,¹⁸  and the phase 3 SHEPHERD study²⁰ ) and the common phase 3 extension study. All of the protocols were approved by the institutional review board at each center, and informed consent was obtained in accordance with the Declaration of Helsinki from all patients.

The open-label 12-week phase 2 pilot study was conducted in 11 patients at 2 study centers in the United Kingdom.¹⁹  Patients who completed the pilot study were eligible to enroll in a one-year extension study followed by a 2-year extension study. Patients 18 years or older who had received at least 4 red cell transfusions in the previous 12 months and had a PNH type III erythrocyte population of at least 10% were eligible to enroll in the pilot study. Patients who were taking stable doses of immunosuppressive drugs (eg, cyclosporin), warfarin, and iron supplements were permitted to continue them during the study. Eculizumab was dosed at 600 mg via intravenous infusion every 7 days for 4 doses, 900 mg 7 days later, and 900 mg every 14 days as a maintenance dose.

The double-blind placebo-controlled 26-week phase 3 efficacy and safety study (TRIUMPH) was conducted in 87 patients in the United States, Europe, Australia, and Canada.¹⁸  Patients 18 years or older, with at least 4 red cell transfusions in the previous 12 months, a PNH type III erythrocyte population of at least 10% or higher, platelet count of at least 100 × 10⁹/L or higher, and lactate dehydrogenase (LDH) level of 1.5 times or more of the upper limit of normal were eligible to enroll. Exclusion criteria were previously described.¹⁸  Patients who were taking stable doses of immunosuppressive drugs, anticoagulants, and iron supplements were permitted to continue them during the study. Patients were centrally randomized (1:1) and stratified according to red cell units transfused in the past year to receive either placebo or eculizumab. Study medication was dosed as follows: eculizumab (600 mg) or placebo via intravenous infusion every 7 (± 2) days for 4 doses; eculizumab (900 mg) or placebo via intravenous infusion 7 (± 2) days later; followed by a maintenance dose of eculizumab (900 mg) or placebo via intravenous infusion every 14 (± 2) days, for a total of 26 weeks of treatment.

The open-label 52-week phase 3 study (SHEPHERD) was conducted in 97 patients in the United States, Europe, Australia, and Canada. Patient characteristics were similar to those of patients treated in TRIUMPH except that the pretreatment transfusion eligibility was adjusted to one or more red cell transfusions in the previous 24 months and the minimum platelet count requirement was lowered to 30 × 10⁹/L; all patients received the identical eculizumab dose as in the TRIUMPH study.

The common open-label 102-week phase 3 extension study is an ongoing study in which a total of 187 patients have enrolled following completion of the parent clinical studies (195 patients were enrolled in parent studies). Placebo-treated TRIUMPH patients entering the phase 3 extension study received the same eculizumab regimen outlined for the TRIUMPH and SHEPHERD studies. Eculizumab-treated patients from the pilot, TRIUMPH, or SHEPHERD studies that entered the phase 3 extension study continued to receive the 900-mg maintenance dose of eculizumab. Doses of immunosuppressive drugs, anticoagulants, and iron supplements could be altered at the discretion of the treating physician.

For patients enrolling in the TRIUMPH, SHEPHERD, and common phase 3 extension studies, each of the respective protocols specified that all adverse events and past medical histories were to be evaluated by the principal investigator to identify clinical TE events as defined by the same major adverse vascular event (MAVE) criteria in all studies. For each study, the protocol specified that the principal investigator record the description, location, method of diagnosis, date of diagnosis, and date resolved for each thrombosis/MAVE event in preidentified case report form fields for events both prior to and during eculizumab treatment. Because of the diverse anatomic locations of thromboses in patients with PNH, multiple diagnostic tests were used. At least 73% (91/124) of pretreatment MAVEs were objectively identified, and the most frequent tests were ultrasound, computed tomography (CT) scan, magnetic resonance imaging (MRI), and angiogram. Events classified as MAVE included thrombophlebitis/deep vein thrombosis, pulmonary embolus, cerebrovascular accident, amputation, myocardial infarction, transient ischemic attack, unstable angina, renal vein thrombosis, mesenteric vein thrombosis, portal vein thrombosis, gangrene, acute peripheral vascular occlusion, sudden death, and a category for other events. All adverse and past medical history events for patients enrolling in the pilot study were also evaluated in the same manner, and the MAVE criteria were retrospectively applied. All MAVEs were independently source verified. The intention-to-treat analysis was based on all 195 patients who enrolled in all 3 parent clinical studies. The pretreatment patient-years included the period extending from the earlier of the date of the diagnosis of PNH or the first thrombotic event to first eculizumab treatment. Pre-eculizumab treatment TE events included all events in all patients prior to enrollment in each of the studies (and during placebo treatment in TRIUMPH). The eculizumab treatment TE events included all events during the period commencing from the first eculizumab dose in the parent studies until the earlier of the follow-up after the last dose of eculizumab in the parent or extension studies or the database lock in November 2006. For the determination of TE events in patients receiving chronic antithrombotic agents, the Anatomical Therapeutic Chemical Classification System was used to identify relevant chronic therapies including anticoagulant and antiplatelet agents.²¹ 

For each type of TE event and for the total group of events collected in the thrombosis record case report form, the incidence rate was tabulated. Rate of thrombosis per patient year for the pre-eculizumab and the eculizumab treatment periods was analyzed for the overall patient population and for each parent clinical study.

TE event rates for pre-eculizumab and eculizumab treatment periods were calculated as the total number of TE events divided by the time in years on a per patient basis. Wilcoxon signed-rank test was used for the prespecified analysis on an intention-to-treat basis and for all subanalyses of the differences between the pre-eculizumab and eculizumab treatment TE events and incidence rates.

The prespecified TE clinical outcome was determined on an intention-to-treat basis in the multinational phase 3 open-label extension study that enrolled patients from 3 independent eculizumab PNH clinical studies (n = 195). Baseline characteristics of patients in each of the parent clinical studies are shown in Table 1. Ninety-six percent of all eligible patients chose to enroll into the common extension study.

Hemolysis is the primary clinical manifestation of PNH and is associated with serious morbidities in PNH including TE.^1,6-9  Inhibition of terminal complement with eculizumab rapidly and consistently reduced hemolysis to near normal levels in patients from all parent clinical studies, while placebo treatment in the TRIUMPH study showed no effect (Figure 1). Subsequent eculizumab treatment of placebo-treated patients also resulted in a rapid and sustained reduction in hemolysis.

Prior to receiving eculizumab, application of the specified MAVE criteria to the evaluation of medical history events identified 124 TE events in the 195 study patients. Pretreatment TE events in the study group occurred in both venous (85%) and arterial (15%) sites (Table 2). The most common sites of venous thrombosis were lower extremity deep veins (18.5%), mesenteric/splenic veins (18.5%), and hepatic/portal veins (16.9%), while the most common site of arterial thrombosis was cerebrovascular accident/transient ischemic attack (13.7%).

To determine the effect of treatment with eculizumab on the incidence of TE events in PNH, the TE event rates in patients with and without eculizumab treatment were compared. Eculizumab treatment resulted in a reduction in the TE event rate in each of the individual clinical studies (Table 3). The prespecified overall TE event rate was reduced from 7.37 events per 100 patient-years (124 total events) before eculizumab treatment to 1.07 events per 100 patient-years (3 total events) with eculizumab treatment in the same patients (P < .001). This represented a relative reduction of 85% and an absolute reduction of 6.3 TE events per 100 patient-years. The TE event rate in placebo-treated patients from the double-blind placebo-controlled TRIUMPH study increased from 2.34 events per 100 patient-years before entering the study to 4.38 events per 100 patient-years during the 6-month period of placebo treatment.

To eliminate a potential bias arising from an unexpected change in the number of TE events immediately preceding trial entry and to adjust for the difference between the number of patient-years accumulated in the pre-eculizumab versus the eculizumab treatment periods, the duration of exposure before and during treatment was equalized for each patient. Compared with the number of TE events before treatment, eculizumab treatment resulted in a significant reduction in the overall number of TE events from 39 to 3 (P < .001; Table 4). This represented a relative reduction of 92%. To determine the pre-eculizumab TE event rate immediately preceding eculizumab treatment, the period 12 months before treatment was examined; the TE event rate during this period in the overall patient population was 17.21 TE events per 100 patient-years (191.8 total patient-years), indicating an increase in rate nearer to initiation of the eculizumab studies.

To evaluate the potential impact of antithrombotic therapy on TE event rates with and without eculizumab treatment, the TE event rate in patients treated with antithrombotic agents was examined. Compared with the pre-eculizumab TE event rate in antithrombotic-treated patients, eculizumab treatment resulted in a reduction in the TE event rate in each of the individual clinical studies and a significant reduction in the overall TE event rate (Table 5). The TE event rate in patients receiving antithrombotics was 10.61 events per 100 patient-years before eculizumab and was reduced to 0.62 events per 100 patient-years in the same patients during eculizumab treatment (P < .001). This represented a relative reduction of 94% and an absolute reduction of 9.99 TE events per 100 patient-years. When the analysis of thromboses in patients receiving antithrombotics was restricted to patients on anticoagulants (n = 91; excludes patients on antiplatelet agents), the TE event rate was reduced from 11.54 events per 100 patient-years before eculizumab to 0.72 events per 100 patient-years during eculizumab treatment (P < .001).

Intravascular hemolysis and the release of red blood cell hemoglobin have been linked to the occurrence of thrombosis and other serious morbidities in various hemolytic disorders.¹  Thromboembolism is a serious and life-threatening complication in PNH and the single most frequent cause of death in patients. Prospective analysis of patient data pooled from 3 clinical studies and a common extension study enabled us to carry out a comparison of the TE event rate in nontreated and in eculizumab-treated patients with PNH. This analysis demonstrates that eculizumab treatment resulted in a dramatic reduction in the TE event rate from 7.37 to 1.07 TE events per 100 patient-years. The reduction in the number of TE events and the TE incidence rates with eculizumab were robust and were not diminished in several clinically relevant sensitivity analyses. Interestingly, TE events in this study occurred in both venous and arterial systems as has been previously described,^7,13,17  confirming that TE is not restricted to the venous system in PNH.

The TE event rate without eculizumab of 7.37 events per 100 patient-years in our patient population, which comprised 195 patients, is more than twice that of the TE event rate published previously for 67 high-risk patients in a single-center descriptive study.⁸  The current TE event assessment was systematic, protocol specified, and performed on a multicenter, multinational, and controlled basis. While most patients in the current study had large populations of PNH granulocytes, TE was also observed before eculizumab in patients who would be expected to have less severe disease with smaller PNH cell populations (< 50%), as has been previously reported.^7,8  The observed variability in the pre-eculizumab TE event rates among the individual studies may be related in part to the different inclusion criteria. For example, the higher pretreatment TE event rate in the SHEPHERD study could reflect inclusion of patients with lower platelet counts that may be due to defective hematopoiesis or platelet consumption as a result of hypersplenism associated with previous thrombosis in the portal system.²²  Despite heterogeneity in the individual prestudy TE event rates, eculizumab treatment consistently resulted in a marked reduction in this rate in each individual study. It is important to note that the TE event rate was higher in the 12-month period immediately preceding eculizumab treatment. The increasing TE event rate of 17.21 events per 100 patient-years observed in the study population immediately prior to trial enrollment may indicate that the overall event rate of 7.37 TE events per 100 patient-years may have been diminished due to the loss of patients who had suffered fatal TE events and did not survive to enroll in the clinical trials or the progression of underlying disease.

The observed reduction of TE events during eculizumab treatment is unlikely to be due to improvements in patient care or more aggressive antithrombotic therapy leading up to and during the clinical studies since the TE event rate did not decrease in the period of time immediately preceding study enrollment or during the study in placebo-treated patients. The pre-eculizumab TE event rate during the 12-month period just prior to study initiation was markedly elevated, reaching a rate of 17.21 events per 100 patient-years. In addition, the TE event rate in placebo-treated patients from the double-blind placebo-controlled TRIUMPH study increased from 2.34 events per 100 patient-years before entering the study to 4.38 events per 100 patient-years during the 6-month evaluation period. Further, initiation of antithrombotic therapy during eculizumab treatment did not contribute to the observed reduction in TE events; only 3 of 195 patients commenced antithrombotic treatment after eculizumab initiation.

Three TE events occurred in patients during eculizumab treatment: a deep vein thrombosis occurred after 1 year of eculizumab therapy in an anticoagulated patient with no history of TE; a pulmonary embolism occurred within 2 months of starting eculizumab therapy in a nonanticoagulated patient with a history of 2 previous TE events (pulmonary embolism and deep vein thrombosis); and a deep vein thrombosis occurred within 6 months of starting eculizumab therapy in a nonanticoagulated patient with a history of 1 TE event (deep vein thrombosis). Hemolysis was effectively controlled in all 3 patients during eculizumab treatment.

Anticoagulation is commonly used in patients with PNH who have a history of thrombosis and has been proposed as a prophylaxis for higher risk patients.^2,8  However, physicians have reported new TE events and advancing TE in patients with PNH while on antithrombotics.^7,8,14  In addition, hemorrhagic adverse events are a significant risk of chronic anticoagulant therapy particularly as some patients are thrombocytopenic, and therefore primary prophylaxis may be contraindicated in this patient population. Chronic anticoagulation was shown to be associated with an overall risk of 7.6 bleeding complications per 100 patient-years with the risk increasing to 11.0 bleeding complications per 100 patient-years during the first 90 days of treatment.²³  The risk for hemorrhage with anticoagulant therapy in patients with PNH has been reported to be approximately 5% or higher, including fatal hemorrhage.^8,14  In the present study, the pre-eculizumab TE event rate remained elevated in patients treated either therapeutically or prophylactically with antithrombotics, such as warfarin, suggesting that such therapies were not adequate to prevent TE in this patient population. By contrast, chronic administration of eculizumab, a therapy that targets the underlying hemolysis in PNH, significantly reduced the overall TE event rate. Whether anticoagulants can be reduced or eliminated in patients with PNH receiving eculizumab is the subject of future investigation. Further, while eculizumab discontinuation in 16 study patients has not been associated with hemolysis above pretreatment levels and concomitant thrombosis, patients should be monitored following eculizumab discontinuation.²⁴ 

The pretreatment TE event rate was elevated, and substantial improvement with eculizumab was observed, even in patients who would be expected to have less severe disease. The TE event rate was reduced from 10.76 events per 100 patient-years before treatment to 2.88 events per 100 patient-years during eculizumab in patients with lower levels of baseline hemolysis (lowest pretreatment LDH quartile). Similarly, the TE event rate was reduced from 4.87 events per 100 patient-years before treatment to 0.00 events per 100 patient-years during eculizumab in patients with minimal pretreatment anemia (0 or 1 transfusion in the year prior to eculizumab treatment). Taken together, these data demonstrate that thrombotic risk is elevated in patients who would be expected to have less severe disease and the benefit of eculizumab in this patient population is maintained.

Hemolysis is characteristic of various hemolytic diseases and has been reported to contribute to serious clinical consequences. In patients with PNH, thromboembolic events have been linked to hemolysis, potentially through the buildup of cell-free plasma hemoglobin.^1,6-9  The effect of free hemoglobin on platelet function and hypercoagulability and the recognized increase in thrombotic tendencies in PNH may be largely due to its ability to scavenge nitric oxide.^25-32  Other potential mechanisms of thrombosis in PNH include the generation of procoagulant platelet microvesicles due to the absence of the terminal complement inhibitor CD59, and the interaction of red cell microvesicles and soluble urokinase plasminogen activator receptor.^11,12,33,34  Terminal complement inhibition with eculizumab would be expected to have a beneficial effect on TE occurring through any of these mechanisms.

The current findings show that long-term administration of eculizumab, in a diverse population of patients, including patients who would be expected to have less severe disease, substantially reduces the risk of thrombosis. In clinical studies to date, eculizumab therapy has controlled hemolysis, improved anemia and quality of life, and appeared to be safe and well tolerated.^18,19,35  Considering that thrombosis has been demonstrated to cause the majority of deaths in PNH, it is reasonable to expect that eculizumab treatment, by decreasing the risk of thrombosis, may increase the life expectancy of these patients.

The studies described in these analyses were funded by Alexion Pharmaceuticals.

We are indebted to Christopher Mojcik, MD, PhD, Henk-Andre Kroon, MD, MBA, Michael Bombara, BS, and Jing Jing Wang, MD, MS (Alexion Pharmaceuticals) for trial oversight; to Jason Chan, PhD (Alexion Pharmaceuticals) for the statistical analysis; and to Kerry Quinn-Senger, PhD (Alexion Pharmaceuticals) for assistance with technical writing.

Additional investigators and institutions who participated in the clinical studies can be found in Document S1 (available on the Blood website; see the Supplemental Materials link at the top of the online article).

Contribution: P.H. designed the study, performed research, collected, analyzed, and interpreted data, and drafted the paper; P.M., L.L., H.S., J. Szer, and R.A.B. performed research, collected, analyzed, and interpreted data, and reviewed the paper; U.D., A.M.R., and A.H. performed research, collected data, and reviewed the paper; J. Schubert, and N.S.Y. designed the study, performed research, collected, analyzed, and interpreted data, and reviewed the paper; G.S. analyzed and interpreted data, and drafted and reviewed the paper; M.B. designed the study, performed research, collected data, and reviewed the paper; S.A.R. designed the study, analyzed and interpreted data, and reviewed the paper; and L.B. and R.P.R. designed the study, analyzed and interpreted data, and drafted and reviewed the paper.

Conflict-of-interest disclosure: P.H., P.M., and G.S. have received grant support from Alexion Pharmaceuticals. P.H., A.M.R., J. Schubert, H.S., R.A.B., G.S., L.L., and A.H. have received lecture fees from Alexion Pharmaceuticals. P.H., J. Schubert, and M.B. have served as consultants for Alexion Pharmaceuticals. P.H., P.M., A.H., J. Schubert, and L.L. have served on Advisory Committees for Alexion Pharmaceuticals. S.A.R., L.B., and R.P.R. are employees of Alexion Pharmaceuticals and have equity ownership in the company. They have assigned to Alexion Pharmaceuticals their inventions made as employees and have received no royalties from the company for these inventions. All other authors declare no competing financial interests.

Correspondence: Peter Hillmen, Department of Haematology, Leeds General Infirmary, Great George Street, Leeds, LS1 3EX, United Kingdom; e-mail: peter.hillmen@nhs.net.