Switching between complement inhibitors in patients with PNH: A real-world analysis of strategy, efficacy, and safety.

Abstract Complement inhibition (CI) for paroxysmal nocturnal hemoglobinuria (PNH) has transformed patient (pt) care. In 2025, 3 C5 inhibitors (C5i) are approved (eculizumab (ECU), ravulizumab (RAV), crovalimab) & 3 proximal inhibitors (PI) (pegcetacoplan (PEG), ipatacopan (IPTA), danicopan (DAN) plus C5i). Some clinical trial therapies have not continued development (vermicopan (VERM), BCX9930 (BCX), other C5i). PI clinical trials have protocols for changing from terminal to PI, based on drug half-life. Data is not available for changing from PI to alternate approved CI. It's inevitable with increased use of newer CI that some pts will have side effects (SE), inadequate response or clinical status changes that require CI change. We present the first, largest cohort of PNH pts switching across different CI & provide preliminary consensus about switching. Method: PNH pts requiring CI which has included a PI (either to PI or C5i) were included. Anonymised data included demographics, treatment indications and regimen, process of changing CI, hemolysis complications within 14 days of starting the next CI (LDH> 2 x ULN with hemoglobin (Hb) drop; severe: >30g/l, non-severe <30g/l), transfusion requirements & patient outcomes. Results: Sixty-two pts from 8 countries were included with mean age at diagnosis 38.4 years (range 16-79) & mean Hb 87.8 g/L (missing data, n=17). Where reported indications for CI were hemolysis (49/62), hemolysis and thrombosis (5/62), thrombosis (3/62). Mean time on CI was 103.6 months (range 23-276; missing n=3) & mean granulocyte clone 86% (range 31-99; missing n=9). First-line CI were ECU/RAV (50/62), VERM (9/62) & investigational C5i (3/62). Initial treatment response (per Risitano et al. 2019): complete (6/62), good (26/62), partial (19/62), or no response (3/62); 8/62 undefined. There were 129 CI changes in 62 pts, classified as terminal-to-proximal, proximal-to-proximal & proximal-to-terminal. Sixty-four terminal-to-proximal changes were reported, pts switching as per clinical trial guidance. Indications were extravascular hemolysis (55/64), hemolysis (5/64), pt choice (2/64), SE (1/64) & trial termination (1/64). Seven hemolytic events occurred (11%), 5 required transfusion. Proximal-to-proximal switches (35): PEG to IPTA (19); BCX to PEG (1) or IPTA (7); 2 IPTA to PEG; 2 DAN+C5i to IPTA (1) or PEG (2); 1 PEG to DAN+C5i; and 2 VERM to DAN+C5i (1) or IPTA (1). Switch indications: hemolysis (11/35), trial termination (10/35), SE (6/35), pt choice (5/35) & other (3/35). DAN+C5i pts had DAN tapered as per protocol, then commenced the next CI. The VERM to DAN+C5i pt had a C5i dose prior to taper, DAN starting the day after taper. One BCX to PEG pt had a 4-week overlap & one had a 72-hour CI gap (BCX to IPTA). Remaining pts had no CI interruption or overlap. There were 3 hemolytic events (3/35; 8.5%); 2 BCX to IPTA, & 1 IPTA to PEG. Two events were severe, all 3 required transfusion. Twenty six proximal-to-terminal switches occured: 9 VERM to C5i; 9 PEG to C5i; 4 IPTA/ACH228 to C5i; 3 BCX to C5i; and 1 DAN+C5i to C5i alone. Switch indications: trial termination (10/26), hemolysis (8/26), SE (3/26), pt choice (2/26) & other (3/26). Factor D inhibitor pts had treatment taper as per protocol. One pt on PEG had treatment overlap. Excluding those on VERM, remaining pts had no interruption or overlap of CI. Thirteen hemolysis events occurred (13/26; 50%). Ten pts had severe hemolysis, all requiring transfusion. Of the 9 VERM pts, 5 had treatment taper and CI treatment gap of 10-163 days; all 5 hemolysed rapidly off CI. Three were tapered & started C5i on the last day of taper, with 2 hemolysing; 1 pt who started C5i before VERM taper also experienced hemolysis. For the PEG to C5i pts, 3 without overlap had hemolysis. The ACH228 pt (tapered before C5i started), also experienced hemolysis. For the DAN+C5i to C5i alone pt, no taper was done & moderate hemolysis occurred.Discussion: PNH treatment with CI is usually lifelong. With the development of newer CI, 'switching’ treatments is increasingly common. No guidance for changing from one PI to another CI is available. This cohort would suggest pts changing from one CI to another CI should not have CI interrupted due to high risk of hemolysis. Overlapping inhibitors is not required (PI to PI; PI to CI) and tapering of factor D inhibitors is not effective. Pts switching from PI to terminal inhibition are also at risk of hemolysis and clinicians should vigilant in this regard.