Key Takeaways:
• A JAAD commentary cautions against cross-trial comparisons of radiographic outcomes among IL-23 inhibitors in psoriatic arthritis due to differences in trial design and patient selection.
• Baseline structural damage explained most of the variance (R² = 0.95; r = 0.97) in placebo-arm radiographic progression across 6 contemporary PsA trials.
• APEX enrolled a biologic-naive, erosion-enriched population, yielding higher placebo progression (1.35 points) than DISCOVER-2 (0.95), FUTURE 5 (0.50), SPIRIT-P1 (0.49), KEEPsAKE-1 (0.32), and BE OPTIMAL (0.31).
• The authors advise weighing radiographic data alongside skin efficacy, joint control, safety, and how closely trial populations match the patient being treated.
A commentary in the Journal of the American Academy of Dermatology argues radiographic outcomes across interleukin-23 inhibitor trials in psoriatic arthritis (PsA), including guselkumab's (Tremfya) phase 3b APEX trial, reflect trial design and patient selection rather than comparative efficacy.¹
The letter, authored by dermatologists David Oberlin, MD, of Forefront Dermatology, and Jesse Veenstra, MD, PhD, of Henry Ford Health and Michigan State University, responds to renewed interest in APEX's radiographic findings for guselkumab. Cross-trial comparisons among interleukin (IL)-23 inhibitors have been complicated by a lack of comparably enriched trial populations, since APEX required baseline erosive disease in a biologic-naive cohort while other IL-23 inhibitor studies did not.² The authors caution against inferring comparative superiority or inferiority from radiographic endpoints collected under different design conditions.
"The most informative trial is not necessarily the one with the largest effect, but the one whose participants most closely resemble the patient being treated," wrote David Oberlin, MD, and Jesse Veenstra, MD, PhD.¹
Guselkumab and Cross-Trial Radiographic Data in Psoriatic Arthritis
APEX enrolled 1020 biologic-naive patients with active PsA and required baseline erosive disease, producing a population enriched for structural risk. Through week 24, the mean placebo-arm baseline radiographic score was 26.8, and the placebo arm progressed by a mean of 1.35 points. Guselkumab-treated arms progressed by only 0.55 points every 4 weeks (Q4W) and 0.54 points every 8 weeks (Q8W).¹
DISCOVER-2 also enrolled a biologic-naive, inflammation-enriched population of 741 patients but did not require baseline erosions. Mean placebo-arm baseline score was 23.8, with placebo progression of 0.95 points versus 0.29 points with guselkumab Q4W and 0.52 points with Q8W, though the Q8W comparison did not reach statistical significance.²
KEEPsAKE-1 enrolled a broader, less structurally damaged population of 964 patients treated with risankizumab (Skyrizi). Mean placebo baseline score was only 13.5, with placebo progression of 0.32 points and a treatment-arm change of 0.23 points, a difference the authors note did not reach statistical significance.³
Baseline Joint Damage and Comparator Biologic Classes in PsA
Across the 6 trials analyzed, baseline structural damage correlated strongly with placebo-arm radiographic progression (R² = 0.95; r = 0.97), the authors reported. Study population and trial duration, rather than treatment class, appeared to drive much of the observed variance in placebo progression.¹
BE OPTIMAL, a bimekizumab (Bimzelx) trial, enrolled 852 patients with a mean baseline score of 12.3 and minimal placebo progression of 0.31 points at week 16.¹ FUTURE 5, a secukinumab (Cosentyx) trial (n = 996), and SPIRIT-P1, an ixekizumab (Taltz) trial (n = 417), showed similarly low placebo progression, 0.50 and 0.49 points respectively, despite biologic-naive enrollment criteria in each.¹
The commentary presents no new safety data, focusing instead on structural-outcome interpretation. The authors note radiographic signals are harder to isolate statistically when placebo-arm progression is inherently low, regardless of an agent's anti-inflammatory activity. They advise clinicians to weigh radiographic evidence alongside skin clearance, joint control, safety, and how closely a trial's population matches the patient under consideration.
The authors observed a similar pattern among IL-17 inhibitors and older tumor necrosis factor (TNF) inhibitor trials, where low absolute placebo-arm progression limited visibility of treatment effects despite established radiographic benefit for these classes.¹ They conclude study design and baseline population, rather than comparative potency alone, should guide interpretation of radiographic data when counseling patients on biologic selection for PsA in dermatology practice.
Editor's Note: This transcript has been edited for grammar and clarity using artificial intelligence tools.
References
Oberlin D, Veenstra J. Interpreting radiographic outcomes across IL-23 inhibitor trials in psoriatic arthritis: the role of trial design. J Am Acad Dermatol. Published online 2026. doi:10.1016/j.jaad.2026.06.135.
Mease PJ, Ritchlin CT, Coates LC, et al. Inhibition of structural damage progression with the selective interleukin-23 inhibitor guselkumab in participants with active psoriatic arthritis: results through week 24 of the phase 3b randomized, double-blind, placebo-controlled APEX study. Ann Rheum Dis. 2025;84:1983-1994.
Mease PJ, Rahman P, Gottlieb AB, et al. Guselkumab in biologic-naive patients with active psoriatic arthritis: a double-blind, randomized, placebo-controlled phase 3 trial. Lancet. 2020;395:1126-1136.
Kristensen LE, Keiserman M, Papp K, et al. Efficacy and safety of risankizumab for active psoriatic arthritis: 24-week results from the randomized, double-blind, phase 3 KEEPsAKE 1 trial. Ann Rheum Dis. 2022;81:225-231.
