Upadacitinib vs IL-17 Inhibitors in PsA, With Jessica Walsh, MD

In a real-world analysis of patients with psoriatic arthritis (PsA) previously treated with a tumor necrosis factor (TNF) inhibitor, initiating upadacitinib (Rinvoq) was associated with greater reductions in swollen and tender joint counts compared with initiating an interleukin-17 (IL-17) inhibitor.¹ The differences were observed at both 3 and 6 months, although the observational design limits conclusions about comparative treatment efficacy.

Jessica Walsh, MD, a rheumatologist at the Salt Lake City Veterans Affairs Health Care System and University of Utah Health, presented the findings at the American College of Rheumatology (ACR) Convergence 2026 in Orlando, Florida. In an interview with HCPLive, Walsh discussed how the analysis may inform treatment sequencing after TNF inhibitor discontinuation and the limitations of real-world comparisons.

Choosing Therapy After TNF Inhibitor Failure in Psoriatic Arthritis

TNF inhibitors are frequently used as an initial advanced therapy for PsA, but some patients discontinue treatment because of inadequate efficacy or intolerance. Selecting a subsequent therapy remains challenging, particularly when patients have no clinical features clearly favoring one drug class over another.

Upadacitinib, an oral Janus kinase (JAK) inhibitor, has demonstrated efficacy in randomized controlled trials involving patients with PsA who had an inadequate response or intolerance to biologic disease-modifying antirheumatic drugs (DMARDs).² However, real-world comparative evidence involving alternative advanced therapies remains limited.

To examine this question, Walsh and colleagues compared changes in joint counts among adults who initiated upadacitinib or an IL-17 inhibitor following prior TNF inhibitor treatment.

OM1 Registry Design and Baseline Patient Characteristics

Investigators analyzed data from the OM1 PremiOM PsA Dataset, which integrates electronic health records and administrative claims data for more than 60,000 US patients with PsA managed by rheumatologists between January 2013 and October 2025.

Eligible patients were aged ≥18 years, had ≥1 prior TNF inhibitor claim, and initiated upadacitinib or an IL-17 inhibitor on or after December 14, 2021. Bimekizumab was included beginning September 23, 2024. Patients were required to have ≥6 months of pre-index data, a baseline outcome measure, and ≥1 follow-up assessment at 3 or 6 months.

The analysis included 601 patients who initiated upadacitinib and 1196 who initiated an IL-17 inhibitor, including 651 receiving secukinumab, 419 receiving ixekizumab, and 126 receiving bimekizumab.

Mean age was 54.2 years in the upadacitinib group and 55.6 years in the IL-17 inhibitor group, with women comprising 72% and 68% of patients, respectively.

Patients initiating upadacitinib had higher baseline joint counts, with a mean 28-joint swollen joint count (SJC28) of 2.9 compared with 2.3 among IL-17 inhibitor initiators. Mean 28-joint tender joint count (TJC28) was 5.8 versus 4.7, respectively.

Walsh also described greater prior medication exposure among patients initiating upadacitinib. Investigators used multivariable models to account for measured baseline differences, although adjustment could not eliminate the possibility of residual confounding.

Greater Joint Count Reductions Observed at 3 and 6 Months

At 3 months, mean SJC28 decreased by 1.18 joints among patients receiving upadacitinib compared with 0.81 among those receiving IL-17 inhibitors (P = .035). At 6 months, the corresponding reductions were 1.49 and 0.70 joints (P = .004).

Tender joint counts followed a similar pattern. Mean TJC28 decreased by 1.95 versus 1.33 joints at 3 months (P = .029) and by 2.71 versus 1.52 joints at 6 months (P = .004), respectively.

Although the between-group differences were statistically significant, their clinical importance requires consideration of the absolute magnitude of change, baseline disease activity, and individual patient characteristics.

“The magnitude of the change of the joint counts is probably meaningful in a lot of patients, but relatively modest when we’re talking about changing about one joint,” Walsh said.

The findings describe differences in joint count outcomes over 6 months and do not establish superiority across other PsA domains, including skin disease, enthesitis, or patient-reported outcomes.

Channeling Bias and Limitations of Real-World Comparisons

Walsh emphasized the potential influence of channeling bias, in which clinical characteristics affect the likelihood of receiving a particular treatment.

Patients with more severe psoriasis, for example, may be more likely to receive an IL-17 inhibitor, while patients with certain cardiovascular risk factors may be less likely to receive a JAK inhibitor such as upadacitinib. These factors were not fully captured in the analysis.

Although multivariable adjustment accounted for measured baseline characteristics, unmeasured differences between treatment groups may have influenced the observed outcomes.

The analysis also focused on patients with available baseline and follow-up joint count measurements, potentially limiting how broadly the findings apply to patients without comparable documentation.

Four of the study's 6 authors were affiliated with AbbVie, the manufacturer of upadacitinib.

Implications for Treatment Selection After TNF Inhibitor Discontinuation

Walsh noted the relatively low baseline joint counts in the registry population compared with those commonly required for enrollment in randomized clinical trials. This difference may make the findings relevant to patients encountered in routine rheumatology practice, although the observational design introduces additional uncertainty.

Treatment selection after TNF inhibitor discontinuation remains dependent on the broader clinical presentation, including skin involvement, inflammatory bowel disease, cardiovascular risk, other comorbidities, and treatment safety considerations.

“Most of our patients are not going to have one of those factors that would help us decide which treatment or category to go into,” Walsh said. “That’s where these data, I think, are particularly helpful, in those patients who don’t have a specific contraindication or reason for going with an IL-17 over upadacitinib.”

The findings provide additional comparative evidence for clinicians considering subsequent therapy after TNF inhibitor use, but do not replace individualized treatment decisions or establish a preferred treatment sequence.

Walsh called for further real-world comparative studies incorporating additional clinical outcomes and improved methods for addressing channeling bias. Continued development of approaches for extracting disease activity measures from large clinical datasets may also expand opportunities for comparative effectiveness research.

Editor's Note: Walsh reports relevant disclosures with Eli Lilly, Janssen, Novartis, Pfizer, Spyre, and UCB. Four study authors were affiliated with AbbVie.

References

1. Walsh J, Ye X, Peng Y, et al. Real World Comparative Effectiveness of Upadacitinib vs. IL-17i in Patients With Psoriatic Arthritis Previously Treated With TNF Inhibitors: Data From the OM1 Registry. Presented at: ACR Convergence 2026; November 6-11, 2026; Orlando, FL. Abstract 1325. https://acrabstracts.org/abstract/real-world-comparative-effectiveness-of-upadacitinib-vs-il-17i-in-patients-with-psoriatic-arthritis-previously-treated-with-tnf-inhibitors-data-from-the-om1-registry/

2. Mease PJ, Lertratanakul A, Anderson JK, et al. Upadacitinib for psoriatic arthritis refractory to biologics: SELECT-PsA 2. Ann Rheum Dis. 2021;80(3):312-320. doi:10.1136/annrheumdis-2020-218870