B78-14 Establishing a Reference Interval for Functional Alpha-1 Antitrypsin in the Serum of Healthy Pi*MM Donors Using an Anti-neutrophil Elastase Capacity Assay

Rationale This study determined a reference interval (RI) and lower limit of normal (LLN) for functional alpha-1 antitrypsin (fAAT) in serum from Pi*MM phenotype-verified individuals using the World Health Organization AAT reference standard. Antigenic AAT levels determined by methods such as nephelometry represent total protein levels, but the functional component is most relevant physiologically. Methods Serum from 237 self-reported healthy donors were collected in Massachusetts for setting the RI. Samples were analyzed twice for fAAT levels at BioAgilytix using a validated anti-neutrophil elastase capacity (ANEC) assay and phenotyped at the Mayo Clinic using the AAT proteotype S/Z, LC-MS/MS serum (A1ALC) test. Mean results from donors having non-disease phenotype Pi*MM (N = 211) were used to calculate an RI for fAAT in human serum. Afterward, to demonstrate validity of the RI, serum samples from three additional cohorts (FL1, FL2, CA3) representing geographically distinct locations (Florida [FL]; California [CA]) were collected, phenotyped, and analyzed once for fAAT to verify the RI and LLN. Results A total of 417 donor samples were collected, with 374 phenotyping as Pi*MM and therefore included in the analyses. Among Pi*MM donors across the Massachusetts, FL1, FL2, and CA3 (n = 374) cohorts, the mean age was 49, with 53% identifying as female. Race data were not collected in FL1. Baseline race characteristics across the Massachusetts, FL2, and CA3 cohorts (n = 317) were Caucasian, 63%; Black, 22%; other, 10%; Asian, 4%. The fAAT RI (2.5th to 97.5th percentile) for Pi*MM phenotype donors from the Massachusetts cohort was determined to be 23.8μM (assigned LLN) to 42.4μM, with a median of 30.9μM (Table). Verification donor sets FL2 and CA3 were then independently shown to have similar data distributions with no statistically significant differences from the RI set (FL2: 22.4μM‒43.7μM, median 30.4μM; CA3: 23.7μM‒48.4μM, median 32.2μM); however, samples from FL1 demonstrated a significantly lower distribution (19.0μM‒38.9μM, median 29.7μM) than the RI cohort. The FL1 dataset was subsequently rejected after an investigation of suspected sample degradation. Among the FL2 and CA3 verification cohorts, the percentage of donors that fell below the LLN was 7.6% (4/52) and 1.9% (1/54), respectively. Conclusions A normal reference range for functional AAT, which is deemed more physiologically relevant than AAT total protein levels, was established and confirmed using healthy, Pi*MM individuals. This can be used going forward to assess the efficacy of current and future AATD therapies in bringing fAAT levels into the normal range. This abstract is funded by: Sanofi