Two methods of identifying patients with axPsA, including an artificial intelligence approach, reported nearly identical outcomes with positive improvement of symptoms of axPsA.
Philip Mease, MD, explains the results of his recent American College of Rheumatology Convergence 2022 presentation, “Efficacy of Guselkumab in Three Cohorts of Biologic-Naïve PsA Patients with Axial Involvement Defined Based on Imaging and Machine-Learning Criteria: Pooled Analysis of 2 Phase 3 Studies,” in which patients treated with guselkumab reported significant and clinically meaningful improvements across a variety of domains as early as week 8. Mease is a rheumatologist and Director of Rheumatology Research at the Swedish Medical Center/Providence-St Joseph Health.
“About 40%, more or less, of patients with psoriatic arthritis (PsA) have evidence of inflammatory spine disease,” Mease stated. “When you look at the back and the sacroiliac joints, it's very similar to patients with ankylosing spondylitis.”
In the post-hoc analysis, efficacy was assessed using a variety of measures, including Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), spinal pain, morning stiffness, and the Ankylosing Spondylitis Disease Activity Score (ASDAS).
“They also, interestingly, had another group in which they applied machine learning technology, artificial intelligence technology, to the phase 3 datasets,” Mease noted. “In that, they found a cluster of patients who all had back pain, who had back involvement, or spine involvement and other characteristics.”
Both groups reported nearly identical improvements in BASDAI, ASDAS, spine pain, and spine involvement.
“The purpose of this abstract was simply to show that by 2 different methods of identifying patients with axPsA, including this artificial intelligence approach, they came out with basically the same outcomes with positive improvement of symptoms of axPsA,” Mease emphasized.