PRO65 INDIRECT COMPARISON BETWEEN HUMAN C1-INHIBITOR (C1-INH) SUBCUTANEOUS TREATMENT AND LANADELUMAB FOR ROUTINE PREVENTION OF HEREDITARY ANGIOEDEMA (HAE) ATTACKS

Abstract

For prophylaxis of HAE attacks, approved therapies include C1 Esterase Inhibitor Subcutaneous (Human) [C1-INH(SC)] replacement and lanadelumab, a monoclonal antibody. In the absence of a head-to-head comparative study of the two treatments, an indirect comparison of data from two independent clinical trials was undertaken. Two phase 3 multicenter, randomized, double-blind, placebo-controlled, trials were identified which evaluated either lanadelumab (the HELP study; NCT02586805; 26 weeks; parallel-group) or C1-INH(SC) (the COMPACT study; NCT01912456; 2x16-week periods; crossover) for routine prevention of HAE attacks. To standardize differences in design and methodology, COMPACT individual patient data were re-analyzed using Poisson regression as in HELP. Estimates (95% CI) for the monthly attack rate reduction over placebo were compared to HELP published results. Treatment group, period, sequence and normalized run-in period attack rates were used as model fixed effects. The original COMPACT model-based mean (95% CI) HAE attack rate for the recommended C1-INH(SC) 60 IU/kg twice weekly and the placebo sequences were 0.52 (0.00, 1.04) and 4.03 (3.51, 4.55), respectively. The HELP model-based mean HAE attack rates for the lanadelumab treatment arms (150mg q4weeks, 300mg q4weeks, 300mg q2weeks) and the placebo arm were 0.48 (0.31, 0.73), 0.53 (0.36, 0.77), 0.26 (0.14, 0.46) and 1.97 (1.64, 2.36), respectively. The HELP least-squares mean reduction in attack rates over placebo were 1.49 (1.08, 1.90); 1.44 (1.04, 1.84), and 1.71 (1.33, 2.09), respectively, compared to the HELP-model-based (Poisson regression) COMPACT reduction of 3.29 (2.65, 3.98) attacks per month. HELP-model- based COMPACT results confirm that the strong efficacy of C1-INH (SC) is invariant to the statistical method used.