One-Year Comparative Effectiveness and Safety of Upadacitinib Versus Risankizumab for CD

Two newer treatment options for patients with Crohn’s disease (CD) are the selective Janus kinase 1 (JAK1) inhibitor upadacitinib and the interleukin-23 (IL-23) p19 monoclonal antibody risankizumab. Both agents have independently demonstrated efficacy in the treatment of CD.^1-3 However, their relative efficacy and safety are unknown owing to lack of head-to-head trials.

At DDW 2025, Rahul S. Dalal, MD, MPH, and colleagues presented results of a real-world retrospective analysis comparing the effectiveness and safety of upadacitinib and risankizumab in patients with CD receiving care in a large urban academic health system (Table 1).⁴ Investigators used statistical tools, including inverse probability of treatment-weighted Cox and logistic regression with a priori–selected covariates, to account for differences in baseline disease severity between groups.

Table 1.

1-Year Comparative Effectiveness and Safety of Upadacitinib vs Risankizumab for CD

Unadjusted outcomes	Proportion of participants
	Upadacitinib	Risankizumab
Clinical responsea
12 weeks	71% (46/65)	68% (104/152)
52 weeks	67% (37/55)	79% (107/136)
Steroid-free clinical remission
12 weeks	37% (24/65)	32% (49/152)
52 weeks	42% (23/55)	50% (68/136)
Endoscopic responseb within 52 weeks	45% (9/20)	65% (37/57)
Radiologic response within 52 weeks	18% (4/22)	29% (14/49)
Treatment discontinuation within 52 weeks	34% (23/67)	15% (23/152)
Surgery within 52 weeks	9% (6/67)	3% (4/152)
Adverse event within 52 weeks	38.8% (26/67)	19.7% (30/152)

CD, Crohn’s disease; HBI, Harvey-Bradshaw Index; SES-CD, Simple Endoscopic Score for Crohn’s Disease.

^aClinical response: reduction of HBI by ≥3 points.

^bEndoscopic response: SES-CD reduction by 50%.

Adapted from Dalal RS et al. Presented at: DDW 2025; May 3-6, 2025; San Diego, California, USA. Abstract 269.4

The cohort included 219 patients who started either upadacitinib (n=67) or risankizumab (n=152) in 2023. Most baseline characteristics were similar between arms, although patients receiving upadacitinib were younger than those receiving risankizumab, with a median age of 34 years and 42 years, respectively (P<.01). Sex, median disease duration, race/ethnicity, median body mass index, smoking status, and most CD characteristics were similar between groups.

There was a trend toward a higher proportion of ileal CD in the risankizumab cohort vs the upadacitinib cohort (21.7% vs 9.0%), although this did not reach statistical significance. Patients receiving upadacitinib were more likely than patients receiving risankizumab to have received prior anti–tumor necrosis factor (TNF) therapies (P<.01), prior ustekinumab (P<.01), and prior risankizumab (P<.01). Baseline corticosteroid use was similar between groups.

Rates of active endoscopic inflammation were similar between groups, as were imaging findings within the past year. In the subset of patients with data available, median C-reactive protein (CRP) was higher in the upadacitinib group vs the risankizumab group (6.2 vs 3.5 mg/L; P=.01), as was the median Harvey-Bradshaw Index (HBI) (9 vs 6; P=.02).

In unadjusted analyses, no significant differences between groups were noted for rates of clinical response at 12 weeks and 52 weeks, steroid-free clinical remission (SFCR; defined as HBI <5 with no use of oral corticosteroids at follow-up or physician global assessment at 52 ± 4 weeks) at 12 weeks and 52 weeks, and rates of clinical, endoscopic, and radiologic response at 52 weeks.

Investigators reported a significant difference between groups in time to treatment discontinuation within 52 weeks, with 34% of patients discontinuing upadacitinib vs 15% of patients discontinuing risankizumab (P<.01). Upadacitinib was also associated with higher rates of surgery within 52 weeks (9% vs 3%; P<.01) and adverse events (AEs) within 52 weeks (38.8% vs 19.7%; P<.01).

Treatment discontinuations through 52 weeks remained more frequent with upadacitinib vs risankizumab after adjusting for competing events; these included discontinuations owing to any cause (weighted hazard ratio [HR], 3.2; 95% CI, 1.7-6.0) and those specifically owing to nonresponse (P<.01). Exploratory Cox and logistic regression analyses found no significant difference between groups in SFCR rates at 52 weeks or in treatment discontinuation rates after excluding TNF-naive patients. Adding baseline CRP and HBI scores also did not change the findings.

AE profiles were as expected with these agents, with upadacitinib yielding higher rates of several AEs, including rash.

Investigators concluded that risankizumab and upadacitinib appeared to yield similar SFCR rates. Risankizumab may provide a more durable therapy, but residual confounding factors and higher disease severity among patients receiving upadacitinib could have contributed to this difference. Thus additional study is warranted to further compare the effectiveness of these agents.

Comparative effectiveness and safety, especially of 2 newer highly effective therapies, is an important clinical inquiry. This single-center retrospective study comparing upadacitinib and risankizumab in 219 patients with CD found higher treatment failure with upadacitinib, but similar steroid-free remission rates. Despite offering early real-world data, its interpretation is limited because of imbalances in baseline characteristics, small sample size (of upadacitinib patients), underreporting of objective markers, and single-center design. Larger prospective multicenter trials are needed to clarify their comparative effectiveness, especially in treatment-refractory patient populations.

—Remo Panaccione, MD, FRCPC