Interleukin 12/23 and interleukin 23 inhibitors for moderate-to-severe ulcerative colitis: a systematic review and network meta-analysis

Introduction

Ulcerative colitis (UC) is a chronic inflammatory illness that presents with mucosal inflammation in the colon and rectum, with standard symptoms of rectal bleeding, diarrhea and urgency [1]. It affects about 1.5 million people and has a prevalence of 0.4% in North America [1]. UC symptoms are associated with a lower quality of life, decreased social and psychological function, and higher healthcare costs [1,2]. Medical therapy aims to reduce symptoms by controlling mucosal inflammation and, in the long term, to avoid disability, colectomy and colon cancer. [3]. Tumor necrosis factor inhibitors (TNF), such as infliximab and adalimumab, are first-line treatments for UC; however, about one third of patients fail to respond to the initial therapy with a TNF inhibitor [4].

Interleukin (IL)-23 is linked to intestinal inflammation and UC pathophysiology [5]. IL-23 consists of 2 components: the p40 subunit, which is also found in IL-12, and the p19 subunit, which is unique to IL-23. IL-23 plays a key role in maintaining and amplifying T helper 17 cells and stimulating various innate immune cells, which are important in the pathogenesis of chronic inflammatory diseases, especially UC [6-9]. Therefore, several phase II and III trials were conducted to evaluate the efficacy and safety of IL-12/23 p40 and IL-23 p19 inhibitors compared to placebo in patients with moderate-to-severe UC. One trial demonstrated the efficacy and safety of ustekinumab, which targets the IL-12/23 p40 subunit, in both the induction and maintenance phases [10], while other trials demonstrated the efficacy of risankizumab, guselkumab, and mirikizumab, which target IL-23 p19, in both the induction and maintenance phases [11-15].

Despite demonstrating efficacy in both the induction and maintenance phases, direct comparative studies evaluating anti-IL-12/23 and IL-23 therapies in patients with UC are lacking, resulting in a significant gap in our understanding of their relative efficacy and safety. Accordingly, we sought to conduct a network meta-analysis to compare the efficacy of all anti-IL-12/23 and IL-23 drugs in terms of efficacy and safety at the end of the induction and maintenance phases in patients with moderate-to-severe UC.

Materials and methods

Results

Study characteristics and risk of bias assessment

Our search yielded 11,887 articles from the database. We selected 62 studies for full-text screening and deemed only 6 appropriate for data extraction and analysis. Fig. 1 depicts the details of our selection process. Six double-blind RCTs were included: 2 were Phase II and 4 were Phase III.

Figure 1 Study selection PRISMA flow diagram

RCT, randomized controlled trial

A total of 3808 patients from 6 studies were included in the induction phase [10-15], with 2422 (63.6%) in the intervention group and 1386 (36.4%) in the control group. For the maintenance phase, 5 studies were included [10-13,15], involving a total of 1697 patients. Of these, 957 (56.4%) were in the intervention group, and 740 (43.6%) were in the control group. Table 1 presents an overview of the baseline characteristics of the patients at the beginning of the induction and maintenance phases. Fig. 2 and 3 present the drugs included in each outcome.

Table 1 Baseline characteristics for the studies and patients in the induction and maintenance phase

Figure 2 (A-E) Network plots for different outcomes at the end of the induction phase

Figure 3 (A-H) Network plots for different outcomes at the end of the maintenance phase

All trials demonstrated a low risk of bias, as assessed by the Cochrane Risk of Bias tool. The results of the bias assessment are shown in Supplementary Table 2.

Overall patients

Induction phase

Risankizumab achieved the highest rate of clinical remission (OR 3.89, 95%CI 2.24-6.75) compared to placebo (Fig. 4). Based on SUCRA, risankizumab ranked first (80.7%) (Supplementary Fig. 1). For clinical response, guselkumab demonstrated the greatest benefit (OR 4.15, 95%CI 2.53-6.82) vs. placebo (Supplementary Fig. 2). SUCRA confirmed guselkumab as the highest ranked (86.4%) (Supplementary Fig. 1).

Figure 4 Comparative efficacy of anti-interleukin [IL]-12/23 and anti-IL-23 for clinical remission and endoscopic improvement at the end of the induction phase in patients with and without a history of inadequate response to advanced therapy, as well as in all patients with moderate-to-severe ulcerative colitis. Comparisons are read from left to right; odds ratios with 95% confidence intervals were used. Blue boxes indicate a statistically significant difference in clinical remission, while green boxes indicate a statistically significant difference in endoscopic improvement

Endoscopic outcomes were consistent. Risankizumab provided the largest increase in endoscopic improvement (OR 4.21, 95%CI 2.12-8.35) (Fig. 4), with a SUCRA ranking of 87.6% (Supplementary Fig. 1). For endoscopic remission, risankizumab had the highest rate (OR 3.39, 95%CI 1.77-6.50) compared to placebo (Supplementary Fig. 3), with SUCRA ranking risankizumab highest (73%) (Supplementary Fig. 1). Risankizumab also showed the highest statistically significant increase of HEMI (OR 3.87, 95%CI 2.49-6.07) compared to placebo, with SUCRA ranking risankizumab highest (87%).

Maintenance phase

Guselkumab achieved the highest rate of clinical remission (OR 4.28, 95%CI 1.58-11.59) (Fig. 5), ranking first by SUCRA (81.6%) (eFig. S4). For corticosteroid-free clinical remission, guselkumab had the highest increasing rate (OR 4.25, 95%CI 2.67-6.75) compared to placebo, and was superior to risankizumab directly (OR 2.17, 95%CI 1.14-4.14) (Supplementary Fig. 5). SUCRA ranked guselkumab first (95.8%) (Supplementary Fig. 4).

Guselkumab demonstrated the highest endoscopic improvement (OR 4.56, 95%CI 2.87-7.23) compared to placebo, and also outperformed risankizumab (OR 2.05, 95%CI 1.09-3.84) (Fig. 5). SUCRA ranked guselkumab highest (93.1%) (Supplementary Fig. 4). Mirikizumab achieved the greatest clinical response (OR 4.01, 95%CI 2.76-5.84), with superiority over risankizumab (OR 2.02, 95%CI 1.15-3.57) (Supplementary Fig. 6), and ranked first by SUCRA (98%) (Supplementary Fig. 4).

Figure 5 Comparative efficacy of anti-interleukin [IL]-12/23 and anti-IL-23 for clinical remission and endoscopic improvement at the end of the maintenance phase in patients with and without a history of inadequate response to advanced therapy, as well as in the overall patients with moderate-to-severe ulcerative colitis. Comparisons are read from left to right; odds ratios with 95% confidence intervals were used. Blue boxes indicate a statistically significant difference in clinical remission, while green boxes indicate a statistically significant difference in endoscopic improvement

For HEMI, guselkumab demonstrated the highest increase (OR 4.54, 95%CI 2.82-7.30) compared to placebo (Supplementary Fig. 5), and ranked first by SUCRA (98.3%) (Supplementary Fig. 4). Regarding endoscopic remission, guselkumab showed the highest increase (OR 2.82, 95%CI 1.72-4.64) compared to placebo (Supplementary Fig. 6), and ranked first by SUCRA (83.7%) Supplementary Fig. 4). Across all agents, no significant differences were observed in adverse events or serious infections (Supplementary Fig. 7).

Discussion

TNF inhibitors have been the first-line treatment for UC for a long time. However, about one third of patients fail to respond to initial therapy, highlighting the need for alternatives [4]. Recent trials have demonstrated the efficacy and safety of anti-IL-12/23 and IL-23 inhibitors in both the induction and maintenance phases of UC treatment [10-15]. In this network meta-analysis, we combined data to evaluate the efficacy and safety of anti-IL-12/23 and anti-IL-23 medications, from 6 RCTs during the induction phase and 5 RCTs during the maintenance phase. We also examined outcomes in a cohort of moderate-to-severe UC patients, including subgroup analysis for those with or without a history of inadequate response.

The central therapeutic goal in UC is to achieve and sustain clinical remission while minimizing corticosteroid exposure [19-21]. In our analysis, risankizumab demonstrated the highest rates of clinical remission during induction, both in the overall population and in the subgroup of patients without a history of inadequate response to advanced therapy. Conversely, among the subgroup of patients with a history of inadequate response to advanced therapy, ustekinumab produced the greatest induction-phase gains in clinical remission. At the end of maintenance, guselkumab was the most effective across key outcomes, showing the highest rates of clinical remission and corticosteroid-free remission in both the overall patient population, and in the subgroups with or without a history of inadequate response to advanced therapy. These findings are concordant with a prior network meta-analysis comparing IL-12/23 and IL-23 in Crohn’s disease, where guselkumab achieved the highest rates of clinical remission at the end of the induction and maintenance phases, as well as the highest corticosteroid-free remission rates at the end of the maintenance phase [22].

Mirikizumab also demonstrated strong maintenance-phase efficacy, emerging as the most effective therapy at the end of maintenance across all patient subgroups, irrespective of prior treatment history. During induction, guselkumab outperformed mirikizumab and ustekinumab in terms of clinical response among patients without a history of inadequate response to advanced therapy, while risankizumab demonstrated superiority over mirikizumab. In the subgroup of patients with a history of inadequate response to advanced therapy, ustekinumab yielded higher remission rates than mirikizumab during induction. During maintenance, guselkumab surpassed risankizumab for both clinical remission and corticosteroid-free remission in patients with a history of inadequate response to advanced therapy. Mirikizumab also outperformed risankizumab in clinical response at the end of maintenance in the overall population.

Discrepancies were observed between the rankings for clinical response and clinical remission within the same phase. These endpoints reflect distinct thresholds: response captures partial symptomatic improvement (including reductions in rectal bleeding), while remission requires more comprehensive disease control, typically including endoscopic healing. This divergence highlights the complexity of interpreting treatment effects, and underscores the importance of clearly defined endpoints. Mechanistic and pharmacologic distinctions probably contributed to these differences. Ustekinumab targets the shared p40 subunit of IL-12 and IL-23, while guselkumab, risankizumab and mirikizumab selectively inhibit the p19 subunit of IL-23. Selective p19 inhibition may preserve IL-12–mediated host defense while more precisely suppressing Th17-driven inflammation [20]. Structural and biophysical studies further demonstrate differences in p19 binding epitopes and affinities; risankizumab and guselkumab exhibit higher IL-23 affinity and in vitro potency compared with ustekinumab and tildrakizumab, while crystallizable fragment (Fc) modifications and epitope features may influence mucosal penetration and receptor blockade [23]. Pharmacokinetics, pharmacodynamics, and dosing schedules also vary. For example, risankizumab for UC induction is administered as 1200 mg IV at weeks 0, 4 and 8 [24], whereas guselkumab is administered as 200 mg IV at the same intervals [15]. Such variations in exposure may account for differential early mucosal healing and long-term durability. Moreover, heterogeneity in trial populations—including the proportion of patients with or without a history of inadequate response to advanced therapy, corticosteroid tapering protocols, and central vs. local endoscopic reads—probably amplifies apparent efficacy differences beyond drug mechanisms alone.

Endoscopic and histological endpoints are key UC targets, given their association with reduced steroid use, hospitalization and colectomy risk [20,25]. HEMI is particularly predictive of favorable outcomes. During induction, risankizumab achieved the highest rates of endoscopic remission, HEMI, and endoscopic improvement overall, while guselkumab was most effective across these outcomes during maintenance, both overall and in the subgroup of patients without prior inadequate response. In patients with such a history, guselkumab maintained superiority for endoscopic remission, ustekinumab led in induction of HEMI, and risankizumab outperformed mirikizumab and ustekinumab for induction of HEMI and endoscopic improvement.

UC therapies are limited by non-response, waning efficacy, and adverse events [26-29]. In our analysis, there were no statistically significant differences between IL-23 pathway agents (risankizumab, ustekinumab, guselkumab, mirikizumab) and placebo regarding overall adverse events or serious infections during maintenance. These findings suggest a favorable safety profile that is consistent across agents.

Our results align with broader evidence supporting IL-23 inhibition as a safe and effective treatment for moderate-to-severe UC. Meta-analyses have consistently shown that IL-23 and IL-12/23 blockade improves clinical, endoscopic and histological outcomes relative to placebo [30]. Our network meta-analysis extends this evidence by enabling indirect comparisons between individual IL-23–targeting drugs, thereby offering more granular insights into comparative efficacy and safety.

The 2024 American Gastroenterological Association (AGA) Evidence Synthesis provides an important comparator [31]. The AGA ranked risankizumab and guselkumab among the most effective induction agents in the subgroup of patients without a history of inadequate response to advanced therapy, consistently with our findings. For the subgroup of patients with a history of inadequate response to advanced therapy, ustekinumab, tofacitinib and upadacitinib ranked higher, with risankizumab and guselkumab showing intermediate efficacy. Taken together, these results suggest that risankizumab or guselkumab may be optimal for patients without a history of inadequate response to advanced therapy, whereas ustekinumab may be preferable for patients with a history of inadequate response to advanced therapy.

Therapeutic selection among IL-23 inhibitors may depend on clinical priorities. Risankizumab ranked highest for rapid induction of endoscopic improvement and remission, particularly in patients without a history of inadequate response to advanced therapy. For maintenance, guselkumab consistently outperformed comparators in remission, corticosteroid-free remission, HEMI and endoscopic outcomes. In the subgroup of patients with prior inadequate response, ustekinumab was superior for induction, while guselkumab provided the most durable maintenance, with comparable safety.

This study has several strengths. It is the first network meta-analysis to evaluate IL-23–selective and IL-12/23 inhibitors in moderate-to-severe UC, stratified according to the patients’ history of inadequate response to advanced therapy. It included both induction and maintenance phases, prioritizing clinically relevant endpoints (corticosteroid-free remission, HEMI), and all trials had a low risk of bias. While SEQUENCE compared risankizumab and ustekinumab in Crohn’s disease [32], no head-to-head UC trials exist; our indirect analysis addresses this gap.

However, important limitations must be acknowledged. Induction phase durations varied slightly, with 12 weeks for all trials except the ustekinumab trial, while maintenance ranged from 40 weeks (mirikizumab) to 52 weeks (risankizumab). This variability may affect cross-trial comparisons. Endoscopic remission definitions were not uniform: for example, the Phase III mirikizumab trial defined remission as a subscore ≤1 without friability, which is equivalent to “endoscopic improvement” in other studies, necessitating reclassification. Two of the included studies were phase II, which may have reduced the robustness of the pooled estimates. Safety analyses were restricted to adverse events and serious infections, with limited long-term follow-up. Lastly, the number of trials per agent was small, reducing statistical power for indirect comparisons. These factors mandate cautious interpretation until validated by additional trials.

In conclusion, selecting optimal therapy for moderate-to-severe UC remains challenging, particularly for subgroups of patients with a history of inadequate response to advanced therapy. Our network meta-analysis demonstrated that guselkumab provided the most consistent maintenance-phase benefits across clinical and endoscopic endpoints, whereas risankizumab and ustekinumab were particularly effective during induction in patients without and with a history of inadequate response to advanced therapy, respectively. Mirikizumab also showed strong efficacy at the end of maintenance. All agents demonstrated favorable safety profiles compared with placebo. Future head-to-head RCTs are essential to definitively establish the comparative efficacy and safety of these IL-23 pathway inhibitors in UC.