Abstract
Paroxysmal Nocturnal Hemoglobinuria (PNH) is characterised by chronic complement-mediated hemolysis. The alternative pathway (AP) of the complement cascade is structured around C3 convertase C3bBb which cleaves C3 into C3b. C3b binds to C3bBb to form C5 convertase which cleaves C5 and releases C5b. C5b initiates the assembly of membrane attack complex (MAC) by recruiting C6-9 to the target cell membrane causing cell lysis. In PNH, red blood cells (RBCs) lack surface-bound complement regulators CD55 and CD59 which protect against complement amplification and MAC formation on host cells. The use of complement inhibitors (CI) has improved the morbidity and mortality in patients with PNH but intravascular hemolysis (IVH) may return with underdosing or complement amplifying events (breakthrough hemolysis (BTH)).
Complement Factor H (CFH) is the main fluid phase regulatory protein of the AP. It accelerates the irreversible decay of C3bBb through its binding to C3b, causing the release of Factor Bb from C3 convertase. In binding to C3b it competes with Factor B to prevent the formation of new C3bBb, and acts as a cofactor for a second fluid phase protein, Complement Factor I (CFI), which cleaves C3b into inactive products.
Some patients appear to be more prone to BTH than others. Given the role of CFH and CFI in protecting against complement amplification, it is of interest to investigate how levels vary between patients with PNH at baseline and during episodes of BTH.
EDTA blood samples were collected from 61 patients with PNH (PNH granulocyte and monocyte clone >10%) during routine clinic visits. Patients were grouped according to CI at sampling; untreated [n=16], eculizumab [n=10], ravulizumab [n=15], pegcetacoplan [n=16], iptacopan [n=15]). Some patients [n=11] were sampled at two timepoints following a change in treatment group. Those on CI had an LDH <1.5x the upper limit of normal (ULN).
EDTA blood samples were also collected during 14 BTH events in 10 patients. 9 patients had baseline clinic samples available for paired comparison; 1 did not. BTH was defined as new signs/symptoms of IVH with an LDH rise >1.5x ULN. Samples were obtained within 5 days of symptoms.
Plasma was separated by centrifugation and stored at –80°C. CFH and CFI levels were measured using in-house sandwich ELISAs.
All patients were consented to the PNH Research Tissue Bank.
CFH levels measured during BTH (mean 503.8 µg/mL) were significantly higher than those measured in untreated patients (398.1 µg/mL; p=0.0052) and those on CI at baseline (LDH<1.5x ULN; 339.7 µg/mL; p<0.0001). All 9 individuals with paired clinic and BTH samples exhibited an increase in CFH during BTH, with a mean absolute increase of 111.2 µg/mL, a 32% rise.
CFI levels did not differ significantly between untreated patients, those on CI at baseline (LDH<1.5x ULN), and BTH samples (mean 42.49, 39.05, and 39.50 µg/mL, respectively). In patients with paired samples, no change in CFI levels was seen during BTH (mean 38.63 vs 39.32 µg/mL).
CFI levels in baseline samples were significantly lower in patients on CI with a history of BTH compared to those without (36.46 vs 41.29 µg/mL; p=0.0173). There was no significant difference in CFH levels (p=0.67) between these groups.
There was no significant difference in CFH levels between treatment groups at baseline. CFI levels differed significantly between two groups; patients receiving pegcetacoplan had significantly lower baseline CFI levels compared to those receiving ravulizumab (mean difference: 10.56 µg/mL; p<0.05). No other pairwise comparisons reached statistical significance.
In our cohort, CFH levels rose during episodes of BTH whilst CFI levels remained unchanged. Baseline CFI levels were lower in patients with a history of BTH, suggesting that reduced CFI levels could confer a predisposition.
The rise in CFH during BTH may reflect a homeostatic response aimed at limiting AP amplification. CFH is primarily made in the liver but is also produced by immune cells recruited to sites of complement amplification. Increased complement activity during BTH likely drives an upregulation in CFH levels to restore regulatory control. CFI is dependent on the presence of cofactors, such as CFH, to regulate complement which may explain why upstream CFH levels and not CFI levels rise in BTH. These findings support the concept that dynamic changes in CFH may play an active role in the acute regulation of complement during BTH.
