Phase 1 Study of the Safety and Pharmacokinetics of CSL889 (Hemopexin) in Adults with SCD

Abstract Presentation Date: 6/8/2024 Presentation Start Time: 9:00:00 AM Background Vaso-occlusive crisis (VOC) is the primary reason for hospitalization among patients with sickle cell disease (SCD). Free heme from ongoing hemolysis contributes to VOC by activating endothelium and blood cells adhesiveness. The resulting blockage of blood circulation produces the characteristic pain of VOC. Hemopexin (Hx) is an endogenous plasma glycoprotein that binds extracellular heme. In people living with SCD, Hx is consumed during hemolysis, leading to its depletion. Administration of Hx relieves experimental vaso-occlusion induced by free heme, hemoglobin or hypoxia-reoxygenation in Townes SCD mice (Gentinetta et al. 2022). We are investigating the potential for plasma-derived human hemopexin (CSL889) to neutralize heme toxicity and normalize blood circulation. We report results of a phase 1, first-in-human study evaluating the safety, tolerability and pharmacokinetics (PK) of CSL889 in adults with SCD [NCT04285827]. Methods This 2-part, multicenter, open-label cohort study was conducted in the US, UK, and the Netherlands. The primary objective was to determine the safety and tolerability of single intravenous (IV) doses of CSL889 in adults with SCD of any genotype. Secondary and exploratory objectives included pharmacokinetics, immunogenicity, and biomarkers of target engagement and mechanism of action. In Part A, 6 ascending single dose levels (3, 10, 30, 60, 120, and 200 mg/kg) were tested in subjects not in VOC (n = 4 per cohort). In Part B, 4 subjects hospitalized for management of VOC with IV opioids received a single IV dose of 60 mg/kg CSL889 within 36 hours of admission. All subjects were followed for approximately 32 days post-infusion of CSL889. The serum assay of Hx measures total Hx and cannot distinguish CSL889 from endogenous Hx, nor heme-bound from free Hx. Noncompartmental analysis of PK was completed using total and pre-dose baseline-adjusted Hx values. Results All 28 subjects self-reported as Black or African American, 15/28 (54%) were female, and age ranged from 20 to 57 years. Most (86%) had HbSS genotype. Half of the subjects in Part A and all in Part B reported concomitant hydroxyurea use. In Part B, 3 subjects were receiving voxelotor and/or crizanlizumab. All Part B subjects had histories of ≥ 3 VOC within the prior 12 months, while most Part A subjects (18/24) had ≤ 2. Seventy treatment emergent adverse events (TEAEs) were reported in 22 subjects (51 mild, 18 moderate and 1 severe). Number or severity of TEAEs did not increase with dose or VOC status. The most common TEAEs were sickle cell anemia with crisis (29% of subjects, all ³1 week after CSL889) and headache (17%). Four TEAEs were considered related to CSL889, all nonserious, of mild severity and resolved: 2 TEAEs of transient dizziness in 2 subjects; and TEAEs of transient increase in fibrin D-dimer and decrease in blood fibrinogen level in 1 subject. Two serious adverse events (SAEs) of moderate severity were reported in 1 subject in Part A, neither related to CSL889 (simultaneous COVID-19 and VOC, both resolved). One SAE of acute chest syndrome occurred in Part B, in the setting of an intercurrent diarrheal illness > 8 days after CSL889; considered not related to CSL889 and all resolved. No ECG changes or other notable lab findings were observed. No subjects developed treatment-emergent anti-CSL889 antibody. Baseline levels of endogenous Hx were low but highly variable (range 7.2 to 499 µg/mL; normal value 2000 µg/mL). Total serum heme was also variable over time, within and across cohorts. CSL889 Tmax was ≤ 2.33 hours and T1/2 ranged from 0.375 to 3.52 days across the dose levels. Mean total Hx Cmax at 200 mg/kg was approximately 2-fold higher than the average healthy human level (Santiago et al 2018). Mean Cmax was comparable within variability based on standard deviation in the subjects with and without VOC that received the same dose (60 mg/kg). Biomarker analysis confirmed target engagement, but no trends in pharmacodynamic biomarker concentrations were observed. Conclusions CSL889 had an excellent safety and tolerability profile when administered as a single dose up to 200 mg/kg in subjects with stable SCD and at 60 mg/kg in subjects with VOC. The observed half-life may support dosing once daily or every other day. These results provide a strong foundation for future trials to evaluate potential efficacy.