Prediction of Long Term Joint Damage in Placebo Patients with Active Psoriatic Arthritis

• Activity Number: 77 - Contributed Poster Presentations: Biopharmaceutical Section
• Type: Contributed
• Date/Time: Monday, August 3, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Yan Liu* and Prasheen Agarwal and Xiwu Lin and Bei Zhou
• Companies: Johnson and Johnson and Johnson and Johnson and Janssen Research and Development, LLC and Johnson and Johnson
• Keywords: Psoriatic arthritis ; long term joint damage; prediction; random forest; placebo

Abstract

Psoriatic arthritis (PsA) is a chronic inflammatory disease of arthropathy of the peripheral and axial joints associated with psoriasis, classified within the group of spondyloarthritis. If left untreated, patients with PsA can have persistent inflammation, progressive joint damage, physical limitations and disability. The measurement of structural damage is considered an important domain when assessing the efficacy of therapies for PsA where such structural damage is quantified by objective evaluation of radiographic images. Due to a short placebo-controlled period (typically 24 weeks), the evaluation of long term structural damage in patients receiving placebo is very challenging. A random forest model was developed to predict structural damage at Week 52 by using radiographic imaging data along with patients’ demographic and disease characteristics from 5 historical PsA clinical trial. This method has two components- the first being the development of a reasonable prediction model for patients in placebo arm based on data at Week 0 and Week 24, and then utilizing this model to estimate long term joint damage of placebo patients after one year hypothetical exposure to placebo.