Key takeaways:

• Researchers developed models with a 3-year and 15-year horizon to identify predictors of PsA in psoriasis.
• The models demonstrated net benefits for decisions on referral and monitoring.

Pain, systemic inflammation and HLA-B27 are the “strongest” predictors for psoriatic arthritis among patients with psoriasis, according to data published in The Lancet Rheumatology.

“Psoriasis is typically straightforward to diagnose, with affected individuals usually seen by general practitioners (GPs) or dermatologists,” Axel Svedbom, PhD, of the division of dermatology and venereology at the Karolinska Institute, in Stockholm, and colleagues wrote. “In contrast, psoriatic arthritis can be challenging to diagnose due to its subtle, nonspecific manifestations and the absence of definitive diagnostic criteria or biomarkers, and diagnosis of the disease is ideally conducted by a rheumatologist.

Data derived from Svedbom A, et al. Lancet Rheumatol. 2026;doi:10.1016/S2665-9913(26)00115-3.

“Therefore, a key objective for GPs and dermatologists in the process of psoriatic arthritis diagnosis lies in identifying participants who would benefit from rheumatology referral,” they added. “For rheumatologists, identifying participants with subclinical psoriatic arthritis who are likely to develop psoriatic arthritis is an important challenge.”

Svedbom and colleagues set out to develop and internally validate two sets of prediction models for onset of PsA among patients with psoriasis. The first set was designed to identify patients with new onset-psoriasis who may be candidates for a referral to rheumatology. The purpose of the second set was to identify patients with subclinical PsA who may be at high risk for developing clinical disease.

In drafting their models, the researchers examined data from the Stockholm Psoriasis Cohort, which enrolled patients with psoriasis within 1 year of their first lesion “on non-hairy skin” between January 2001 and December 2005. Data were assessed for predictors of PsA at baseline on 3- and 15-year horizons.

Data for 628 participants who reported no PsA at enrollment were included in the final data set. The cohort was 55% women and 45% men, with a median age of 40.9 years (interquartile range, 30.1-56). Concomitant PsA was observed in 13% of the cohort.

The researchers stratified patients into four groups based on factors such as arthralgia, fatigue, psoriasis phenotype, high-sensitivity C-reactive protein and psoriasis disease activity. Risk for concomitant PsA ranged from 1% to 62%, the researchers wrote.

According to the researchers, all models demonstrated “good discrimination” with an optimism-adjusted area under the curve of 0.76-0.84. Additionally, although the models showed “reasonable calibration” and “positive net benefit” for referral and monitoring, they demonstrated “limited value” for preventive treatment, the researchers wrote.

The most important predictors of PsA development were pain, HLA-B27 and systemic inflammation, according to the results.

“Predictors of psoriatic arthritis are already present at psoriasis onset,” Svedbom and colleagues wrote. “The models developed here could support clinical decision making, but they require external validation before implementation.”