Comparative Effectiveness in Positioning Therapies

With the development of new therapies with new mechanisms of action, it is important to understand how drugs compare in terms of safety, efficacy, and patient preference.¹ Various approaches can be used to compare efficacy and safety outcomes in IBD subpopulations. Randomized controlled trials are the mainstay of comparative research, but they may be limited by several factors, such as a small study population that limits the power to detect a difference between therapies; a follow-up time that is too short; the expense associated with managing large numbers of patients, often at different treatment centers; and the difficulty of recruiting patients with IBD. Observational studies can offer access to a larger patient population, which may be particularly valuable for evaluating rare safety events, but such studies do not contain randomized patient populations and can be limited by confounding variables or selection bias. Alternatively, a network meta-analysis can compare 3 or more treatments in a single analysis. Evidence is gathered from several clinical trials and may have the advantage of including large patient populations. A network meta-analysis relies on the concept of transitivity, such that the only difference between comparisons lies in the treatments. In other words, studies being compared should have a similar patient population, median age, disease duration, concomitant medication use, and other factors.

Only a few randomized controlled trials that directly compare 2 or more advanced therapies have been conducted in patients with IBD. The phase 3 VARSITY trial was the first study to compare 2 biologic therapies in patients with moderately to severely active UC.² The study randomized 769 patients for treatment with adalimumab vs vedolizumab. VARSITY showed a significant improvement in clinical remission rate at 52 weeks with vedolizumab compared with adalimumab in the overall study population (P=.006). However, no difference in therapeutic outcome was observed between groups that were taking steroid medication or immunomodulators. Moreover, no dose escalation was permitted, and drug levels were not assessed; thus, drug levels may not have been optimized.

Levels of evidence vary across comparative effectiveness studies in IBD, necessitating an understanding of which study best informs clinical decision-making in each clinical scenario. Although each mechanism of action carries therapy-specific safety concerns, the safest agent is the one that best controls the IBD. Positioning and sequencing decisions should be patient-centric and informed by shared decision-making.

– Stephen B. Hanauer, MD

Recent network meta-analyses have attempted to derive data to help prioritize the positioning of therapies in patients with UC.^3-5 A Bayesian network meta-analysis compared the efficacy and safety of targeted therapies in patients with moderately to severely active UC (Figure 5).⁵ Efficacy outcomes were evaluated in the intention-to-treat populations from 23 clinical trials. Although induction and maintenance data were initially evaluated separately, they were later combined to provide a single overview of targeted therapy efficacy in UC, and the maintenance data were adjusted with the induction response data. In the biologic-naive population, the analysis showed the highest efficacy with upadacitinib (45 mg for induction and 30 mg for maintenance), based on the rates of clinical response (66.7%), clinical remission (40.6%), and endoscopic improvement (55.0%). Upadacitinib at the same doses for induction and maintenance also showed the best efficacy in the biologic-exposed population, based on the rates of clinical response (59.8%), clinical remission (51.0%), and endoscopic improvement (54.1%). Network meta-analyses have also been performed in CD populations.^6-8

Figure 5.

Comparative effectiveness via a Bayesian network meta-analysis for patients with UC.

ADA, adalimumab; FIL, filgotinib; GOL, golimumab; INF, infliximab; ITT, intention-to-treat; NA, not available; OZA, ozanimod; PBO, placebo; Q2W, every 2 weeks; Q4W, every 4 weeks; Q8W, every 8 weeks; Q12W, every 12 weeks; TOF, tofacitinib; UPA, upadacitinib; UST, ustekinumab; VED, vedolizumab.

Adapted from Long et al. Comparative effectiveness in positioning therapies. Presented at the Advances in Inflammatory Bowel Disease Conference; Orlando, Florida; December 14-16, 2023.¹

Optimal therapeutic decision-making includes several considerations beyond drug efficacy. Patient characteristics to consider include age, comorbidities, and personal preferences. Advanced therapy choices in IBD must take individual patient characteristics into consideration, including pregnancy, pharmacokinetics, combination with existing therapies, and perianal disease. Disease characteristics include the extent of disease, complications, early vs late, and prior outcomes from therapy. Safety concerns vary with drug class and sometimes among drugs that have the same target or a similar mechanism of action.^9,10 Common safety concerns associated with drugs for IBD include infection, cancer, and cardiovascular events.

An observational study evaluated the efficacy and safety outcomes in patients with CD initiating a new biologic therapy at 5 health systems in California.¹¹ Using propensity score matching, the study evaluated the risk of infection, hospitalization, or surgery in patients treated with ustekinumab, vedolizumab, or anti-TNF agents. Ustekinumab was associated with the lowest risk of infection, and no differences between agents emerged for risk of hospitalization or surgery. Vedolizumab and anti-TNF agents were associated with a similar risk of infection. Although safety outcomes should be considered, the best therapy is the one that controls the disease.