Registration of an intravenous secukinumab regimen not tested in clinical trial – Analyses used to support this Model-Informed Drug Development approach
Thomas Dumortier1, Gerard Bruin2, Didier Renard1, Jelena Mijatovic1, Anshu Marathe4, Ruvie Martin3
1Novartis Pharma AG, 2Novartis Biomedical Research, 3Novartis Pharmaceuticals Corporation, 4Novartis Biomedical Research
INTRODUCTION COSENTYX® (secukinumab) was originally approved in Spondyloarthritis (SpA) with two subcutaneous (SC) regimens, with maintenance doses of 150 and 300 mg every 4 weeks (q4w). Based on feedback from healthcare providers, an intravenous (IV) dosing (weight-based) is desirable to add to the available therapeutic armamentarium for adult patients with SpA. In December 2021, a Model-Informed Drug Development (MIDD) approach was discussed with the US Food and Drug Administration (FDA) in the context of the Agency’s MIDD pilot program to select an IV regimen for registration. Specifically, it was agreed to use “PK bridging”, i.e., (1) to select, by popPK prediction, an IV regimen that approximates the steady-state exposure of the two approved SC regimens; (2) to assess the efficacy and safety from this IV regimen by extrapolation from the approved SC regimens, following the key pharmacological assumption that same exposure should lead to same response regardless of the route of administration (“same exposure-response relationship” assumption). While this approach obviates the need of conducting an efficacy/safety study, the FDA was still hinting at the need of a PK study to check that this IV regimen achieves the predicted exposure levels. Having to conduct such a study would significantly delay the IV program. OBJECTIVES -Present the MIDD approach, including the interactions with the FDA -Show how pharmacometrics models were used to identify the IV regimen -Discuss the models (relative) shortcomings, the sensitivity analyses used to assess their impact, and more generally to support the credibility of our MIDD approach, ultimately allowing to avoid a PK study METHOD & RESULTS Based on an available popPK model developed using IV and SC data from 14 studies, we identified that an IV regimen with maintenance doses of 1.75 mg/kg q4w, never tested in clinical trial, was expected to best approximate the steady-state exposure of the two approved SC regimens. External validation of the popPK model, performed using data of an IV study at a higher dose (3 mg/kg), substantiated the model capacity at predicting IV exposure. However, internal validations showed discrepancies of moderate magnitude (+/- 15%) in some of the studies used for estimation that could not be fully explained, even after inclusion of between-study variance component in the popPK model. We kept a potential 15% prediction error on PK in perspective by evaluating its consequences efficacy- and safety-wise. This was done by means of exposure-response analyses for 18 important efficacy and safety endpoints, estimated using data from 9 Phase 3 studies from the SC clinical program. The analyses showed that the risk of meaningfully affecting the safety was low, as indicated by the absence of exposure effect of relevant magnitude on the risk of experiencing adverse events (AEs) such as general infections and infestations, and Candida infections. The risk of meaningfully affecting the efficacy appeared also low, in particular for primary efficacy endpoints (ACR20 and ASAS40) for which the IV regimens were predicted to deliver exposures at the plateau of the exposure-response relationship. Finally, the “same exposure-response relationship” assumption was supported by showing that the efficacy and safety summary results in patients treated with 3 mg/kg IV q4w were consistent with the exposure-response relationships estimated based on SC data only. CONCLUSIONS The IV regimen with maintenance doses of 1.75 mg/kg q4w, selected following a PK bridging approach and never tested in clinical trial, was submitted for registration in December 2022. As the outcome of those analyses was to be used for regulatory approval decision making, it was essential to establish the credibility of the involved models and assumptions [2,3]. The credibility analyses showed that (1) the popPK model was fit for purpose, i.e., adequate to accurately select the candidates IV regimens, (2) the consequence of exposure misprediction, if any, would be minor, i.e., would not meaningfully affect the efficacy and safety of the IV regimens, and (3) supported the assumption of “same exposure-response relationship” regardless of the route of administration. The popPK and the credibility analyses formed the basis of the dossier supporting registration of the 1.75 mg/kg q4w IV Cosentyx regimen in SpA. The regimen was approved in October 2023, without having conducted any further clinical study.
Video link:
https://www.youtube.com/watch?v=68kxPG-KcGg
