In all, 137 of 145 patients with PNH (94.5%) who completed a pegcetacoplan trial entered the 307-OLE. Four of four patients (100%) who completed the PHAROAH parent study [27], four of four (100%) who completed PALOMINO [28], 15 of 20 (75.0%) who completed PADDOCK [28], 64 of 67 (95.5%) who completed PEGASUS [25], and 50 of 50 (100%) who completed PRINCE [26] entered the 307-OLE. All 137 patients had received at least one dose of pegcetacoplan (ITT population) at data cutoff and were included in this analysis (Table 1). At data cutoff, eight patients had withdrawn from the 307-OLE (Table 1). The reasons for study withdrawal were adverse event (three patients from PEGASUS), physician decision (two patients from PEGASUS), loss to follow-up (one patient from PEGASUS), sponsor request (in a joint decision made with the investigator; one patient from PADDOCK), and death (one patient from PRINCE).

At 307-OLE entry, patients in the total population (N = 137) received pegcetacoplan 1080 mg twice weekly (122 patients, 89.1%), once every 3 days (14 patients, 10.2%), or three times per week (One patient, 0.7%). As of the data cutoff, the maximum dosing frequency was twice weekly for 104 patients (75.9%), once every 3 days for 25 (18.2%), and three times per week for eight (5.8%). Nearly all patients (134 patients, 97.8%) completed all injections. As of the data cutoff, 107 patients (78.1%) had completed 48 weeks of additional treatment with pegcetacoplan in the 307-OLE. The median (range) treatment duration was 48.1 (7.9–50.1) weeks.

Demographic and clinical characteristics and laboratory measurements at 307-OLE entry are shown in Table 2. In the total population, the mean (SD) patient age was 46.5 (14.6) years, and slightly more than half of the patients were female. The most common race was Asian (46.0%), followed by White (35.0%), Other (5.8%), American Indian or Alaska Native (3.6%), and Black or African American (2.2%). Most patients (81.0%) were not of Hispanic or Latino ethnicity.

Patients in the total population had a mean (SD) body mass index of 25.6 (5.3) kg/m2 at 307-OLE entry. The median (IQR) time since PNH diagnosis was 6.9 (0.61–39.0) years. The median durations of PNH were shorter in studies of C5 inhibitor-naive patients (with the exception of PADDOCK) and longer in studies of patients with low hemoglobin despite eculizumab treatment.

At 307-OLE study entry, mean hemoglobin concentration was 11.6 g/dL for the total population, and hemoglobin concentrations were comparable across parent study populations. The median LDH concentration (184.0 U/L) was within the normal range (113–226 U/L) in the total population and in all parent study populations except PHAROAH, in which it was somewhat elevated (302.3 U/L). Across the parent study and total populations, the mean absolute reticulocyte counts were within normal limits (10–140 × 109 cells/L for male patients; 10–120 × 109 cells/L for female patients), as were indirect bilirubin concentrations (normal range 1.7–15.4 μmol/L). The mean FACIT-Fatigue scores were at or near the general population norm of 43.6 [29] in all populations.

Mean hemoglobin concentrations were below 12 g/dL (the LLN for female patients) at baseline of the parent studies in the total population and the four parent study populations analyzed (PALOMINO, PADDOCK, PEGASUS, PRINCE) (Fig. 2). Mean hemoglobin concentration for the total study population at baseline of the parent studies was 8.9 g/dL. By 307-OLE entry, hemoglobin concentrations approached 12 g/dL in the total population and all analyzed parent study populations except PALOMINO, which had a higher mean hemoglobin concentration of 13.3 g/dL. Concentrations remained stable through week 48 in all parent study populations, aside from week 4 in patients from PHAROAH, which only included two patients.

Concentrations of hemoglobin at baseline of the PALOMINO, PADDOCK, PEGASUS, and PRINCE parent studies and from 307-OLE entry to week 48 in the total and parent study populations of patients with PNH. LLN lower limit of normal, NA not available, OLE open-label extension, PNH paroxysmal nocturnal hemoglobinuria, SD standard deviation. aHemoglobin LLN for female patients: 12 g/dL. bPatients in PALOMINO, PADDOCK, and PRINCE were complement inhibitor-naive at baseline. cIncludes PHAROAH patients (N = 4), unless noted. dDoes not include PHAROAH patients

At parent study baseline, median LDH concentrations were markedly elevated in all analyzed parent study populations except PEGASUS, in which patients previously treated with eculizumab had concentrations near the ULN (Fig. 3). The median (IQR) LDH values at parent study baseline were as follows: 2585 (2098–3000) U/L (PALOMINO); 2734 (1579–3315) U/L (PADDOCK); 223 (194–271) U/L (PEGASUS); and 2038 (1410–2658) U/L (PRINCE). At 307-OLE entry, median LDH concentrations were within normal limits in the total population and all parent study populations except PHAROAH, which was < 1.5 × the ULN (i.e., 339 U/L). Median LDH concentrations were within the normal range from weeks 12 through 48 of the 307-OLE in all parent study populations.

Concentrations of LDH at baseline of the PALOMINO, PADDOCK, PEGASUS, and PRINCE parent studies and from 307-OLE entry to week 48 in the total and parent study populations of patients with PNH. IQR interquartile range, LDH lactate dehydrogenase, LLN lower limit of normal, NA not available, OLE open-label extension, PNH paroxysmal nocturnal hemoglobinuria, ULN upper limit of normal. aError bars for parent study baseline values were not included to maintain the scale. Patients in PALOMINO, PADDOCK, and PRINCE were complement inhibitor-naive at baseline. bIncludes PHAROAH patients (N = 4), unless noted. cDoes not include PHAROAH patients

At parent study baseline, all analyzed groups had mean FACIT-Fatigue scores below the general population norm of 43.6 [29] (Fig. 4), with a mean score of 34.1 in the total population. In patients from PEGASUS and PRINCE, mean FACIT-Fatigue scores were 31.5 and 36.9 at baseline, respectively. At 307-OLE entry, mean FACIT-Fatigue scores were approaching or above the general population norm in all populations. These scores were maintained through the 48-week cutoff.

Scores for FACIT-Fatigue at baseline of the PALOMINO, PADDOCK, PEGASUS, and PRINCE parent studies and from 307-OLE entry to week 48 in the total and parent study populations of patients with PNH. FACIT-Fatigue Functional Assessment of Chronic Illness Therapy–Fatigue, NA not available, OLE open-label extension, PNH paroxysmal nocturnal hemoglobinuria, SD standard deviation. aDefined by Cella and colleagues [29]. A FACIT-Fatigue score increase ≥ 5 points is considered a clinically meaningful improvement [31]. bPatients in PALOMINO, PADDOCK, and PRINCE were complement inhibitor-naive at baseline. cIncludes PHAROAH patients (N = 4), unless noted. dDoes not include PHAROAH patients

A hemoglobin concentration > 12 g/dL was reached by 40.2% of evaluable patients in the total population at data cutoff (Table 3). In patients from PEGASUS and PRINCE, 36.8% and 46.4% had hemoglobin concentrations > 12 g/dL, respectively. Sex-specific hemoglobin normalization (i.e., ≥ 13.6 g/dL, male patients; ≥ 12 g/dL, female patients) occurred in 31.8% of the total population and in 31.6% of patients from PEGASUS and 28.6% from PRINCE. Normalization of LDH (i.e., ≤ 226 U/L [ULN]) was reached by 67.0% of evaluable patients in the total population. In patients from PEGASUS and PRINCE, 70.2% and 63.0% of patients had LDH normalization at the data cutoff in the 307-OLE, respectively.

In the total population, 83.2% of patients did not require a transfusion during the 307-OLE treatment period up to data cutoff (Fig. 5). Transfusion avoidance occurred during the 307-OLE in 76.6% of patients from PEGASUS and 90.0% of patients from PRINCE.

Percentage of patients with transfusion avoidancea from 307-OLE entry to data cutoff in the total and parent study populations of patients with PNH. OLE open-label extension, PNH paroxysmal nocturnal hemoglobinuria. aPatients who avoided transfusion did not require a transfusion during the treatment period to the data cutoff (i.e., the earliest of week 48 of the 307-OLE or August 27, 2021)

Most patients (73.7%) in the total population (N = 137) reported AEs in the 307-OLE through data cutoff (Table 4). In the total population, 22 patients (16.1%) had AEs considered related to pegcetacoplan by the investigator. Fifteen patients (10.9%) reported AEs related to injection site reactions (ISRs). The most common TEAEs were hemolysis (n = 23, 16.8%), followed by fatigue (n = 8, 5.8%), and then upper respiratory tract infection, urinary tract infection, nasopharyngitis, headache, anemia, and injection site erythema (n = 7, 5.1% for all). No thrombotic events or meningococcal infections were reported.

Serious AEs were reported in 27 patients (19.7%) in the total population (none related to pegcetacoplan) (Table 4). The serious AE of hemolysis, the only serious AE reported by at least 5% of patients in the total population, occurred in 11 patients (8.0%). Three patients, two in PEGASUS and one in PRINCE, discontinued the study because of AEs; all three were due to hemolysis. One patient in PRINCE had a serious AE that was of severe intensity and led to death (sudden cardiac event on day 69 of the 307-OLE); this was deemed unrelated to pegcetacoplan.

In the 307-OLE, 23 patients (16.8%) a TEAE with the PT hemolysis (Table 4). One of these patients, who had high PNH disease activity and several other poorly controlled comorbidities, was considered an anomaly and was not included here. In the 22 patients with hemolysis who were analyzed, there were 25 events of hemolysis; 16 events were associated with a potential CAC, and for 9 events a potential CAC was not identified. Ten hemolysis events required transfusions. Patients had recovered per investigator judgment from 22 of the events at data cutoff. In all, 15 events were treated with an increase in pegcetacoplan dosage; one patient with a dosage increase had subsequent dosage reduction after a second hemolysis event. Three patients with hemolysis received ravulizumab or eculizumab during a hemolysis event, in addition to treatments for hemolysis, CACs, PNH, and/or comorbid conditions. Three of 22 patients with hemolysis events discontinued the study. Additional details are provided in the Supplementary Table.

At 307-OLE entry, more than 80% of patients in the total population had D-dimer normalization [30]. The percentage of patients with normalization remained steady through the 48-week data cutoff (Supplementary Figure).