Positioning Infliximab and Vedolizumab in the Treatment of Moderate-to-Severe Ulcerative Colitis

Narula N, Wong ECL, Marshall JK, et al. Comparative efficacy for infliximab vs vedolizumab in biologic naïve ulcerative colitis [published online ahead of print July 28, 2021]. Clin Gastroenterol Hepatol https://doi.org/10.1016/j.cgh.2021.07.038. In 2005, ACT-1 became the first clinical trial to show that the anti-tumor necrosis factor (TNF) agent, infliximab, was efficacious in treating moderate-to-severe ulcerative colitis (UC). Since that time, we have developed a growing armamentarium of medications with efficacy in treating moderate-to-severe UC, including vedolizumab, ustekinumab, tofacitinib, and ozanimod. Despite advancements in drug development, our knowledge on how best to position these medications in the treatment of UC is limited, because a critical gap in comparative effectiveness data exists in context of few head-to-head clinical trials. The VARSITY trial, which demonstrated superiority of vedolizumab over adalimumab in achieving clinical remission and endoscopic improvement in patients with UC, was the first study to directly compare the efficacy of two biologic agents, but more comparative effectiveness data are sorely needed. Narula et al conducted a post hoc analysis of 3 trials—ACT-1, VARSITY, and GEMINI-1—to compare the efficacy of infliximab and vedolizumab in biologic-naïve patients with moderate-to-severe UC who were exposed to infliximab or vedolizumab at standard induction and maintenance dosing. The primary outcome was clinical remission at 52 weeks. Secondary outcomes included 6-week clinical response and remission, 1-year endoscopic improvement and remission, corticosteroid-free clinical remission, and serious infections. To control for potential confounding, the authors included logistic regressions adjusted for baseline covariates and a sensitivity analysis using a propensity score-matched cohort. The study cohort included 228 patients treated with infliximab and 567 patients treated with vedolizumab. The results demonstrated significant variability between groups on baseline characteristics, including concomitant corticosteroid and immunomodulator use, albumin levels, and disease extent. Results demonstrated no difference in clinical remission at 1 year between infliximab and vedolizumab (adjusted odds ratio [aOR] 1.05; 95% confidence interval [CI], 0.75–1.47). However, the rates of corticosteroid-free clinical remission (aOR, 2.36; 95% CI, 1.27–4.49), and 1-year endoscopic remission (aOR, 1.55; 95% CI, 1.08–2.22) were significantly higher in the infliximab group. Rates of serious infection were also higher in the infliximab group (5.3% vs 0.7%; P < .001). Results from a sensitivity analysis of 223 infliximab-treated and vedolizumab-treated patients, propensity-score matched by baseline characteristics, demonstrated similar results. The authors concluded that infliximab is superior to vedolizumab at achieving 1-year steroid-free clinical and endoscopic remission, suggesting that infliximab may be more efficacious than vedolizumab in biologic-naïve patients with moderate-to-severe UC. With the emergence of novel therapeutics for induction and maintenance treatment of moderate-to-severe UC, additional data on the comparative effectiveness of these therapies are needed. To date, VARSITY is the only head-to-head trial to compare two first-line agents for the treatment of moderate to severe UC (N Engl J Med 2019;381:1215–1226). Although this trial demonstrated the superiority of adalimumab over vedolizumab for achieving clinical remission at 1 year and endoscopic improvement, the direct comparison of vedolizumab to other anti-TNF agents has not been explored. Singh et al sought to address this gap in comparative effectiveness through a recent network meta-analysis of first and second-line pharmacotherapies for patients with moderate-to-severe UC (Clin Gastroenterol Hepatol 2020;18:2179–2191). In this analysis, infliximab ranked highest for inducing clinical remission and endoscopic improvement in biologic-naïve patients, suggesting that it may be the most efficacious agent in this patient population. However, on indirect comparison, there was no difference between infliximab and vedolizumab in rates of clinical remission and endoscopic improvement. Network meta-analyses have become an increasingly important statistical tool in assessing comparative effectiveness of the multiple available therapies for UC. However, these analyses can be prone to heterogeneity, inconsistency, and bias (Ann Intern Med 2013;159:130–7). Therefore, some investigators have used post hoc analysis of clinical trial data to indirectly compare therapies. These types of analyses have previously demonstrated superiority of infliximab over adalimumab (Clin Gastroenterol Hepatol 2017;15:1218–1225) and golimumab (Clin Gastroenterol Hepatol 2020;18:424–431) in the treatment of moderate-to-severe UC. Narula et al are the first to use a post hoc analysis of clinical trial data to compare the efficacy of vedolizumab with infliximab. Narula et al demonstrate 3 clinically relevant points of interest. First, while the rates of 1-year clinical remission were similar between infliximab and vedolizumab, rates of steroid-free clinical remission were significantly higher in the infliximab group. Results from the recent SPIRIT consensus from the IOIBD demonstrated that early corticosteroid use in UC was associated with UC-related hospitalizations, surgery, and future corticosteroid use (Gastroenterology 2021;160:1452–1460). These findings, combined with the known adverse effects of long-term corticosteroid use suggest that steroid-free clinical remission is the more clinically meaningful end point than clinical remission alone. Second, patients treated with infliximab were 1.5 times more likely to achieve endoscopic remission than patients treated with vedolizumab, with a number needed to treat of 8.3 for infliximab over vedolizumab. With an increasing body of evidence suggesting that endoscopic remission predicts future clinical remission, colectomy, and corticosteroid-free clinical remission, endoscopic and histologic end points are becoming increasingly important therapeutic targets (Clin Gastroenterol Hepatol 2016;14:1245–1255). The importance of these end points are reflected in the recent STRIDE II guidelines, which include endoscopic healing as a long-term treatment target for patients with UC (Gastroenterology 2021;160:1570–1583). Finally, the Narula et al study showed a lower incidence of serious infection in the vedolizumab group compared with the infliximab group. Using the propensity score–matched cohort data, they report a number needed to harm of 25 for infliximab compared to vedolizumab. This finding emphasizes the importance of shared decision making and patient preferences when considering therapeutic options for UC. Many physicians may inappropriately make assumptions about patients’ unwillingness to accept risk, when, in fact, patients with UC have been shown to accept high levels of risk to avoid surgery (Inflamm Bowel Dis 2014;20:103–114). The main limitation of this study is that the use of post hoc analysis of clinical trials is subject to similar biases as observational studies. In this case, there are several differences in methodology between trials that could question the validity of the analysis. First, ACT-1 (N Engl J Med 2005; 253:2462–76) and GEMINI-1 (N Engl J Med 2013;369:699–710) both used blinded local investigators to review endoscopy data whereas VARSITY used centralized, blinded reviewers. The authors attempted to control for the differences in endoscopic grading by performing a subgroup analysis of ACT-1 and GEMINI, which both used local reviewers. The results were similar; however, the lack of a centralized review process in both the infliximab and vedolizumab groups limits the ability to infer clinically relevant differences in endoscopic remission rates. Despite these limitations, the Narula et al article has several potential implications for both clinical practice and research in the field of inflammatory bowel disease. First, the current clinical guidelines from both the American Gastroenterological Association (Gastroenterology 2020;158:1450–1461) and the American College of Gastroenterology (Am J Gastroenterol 2019;114:384–414) currently recommend the use of anti-TNF agent or vedolizumab as first-line therapy for biologic-naïve patients with moderate-to-severe UC. This study and other supporting data discussed above suggest that infliximab may be the most efficacious first-line treatment for biologic-naïve moderate-to-severe UC, which may be used to influence future guidelines and ultimately change clinical practice. This research also demonstrates the critical lack of head-to-head trials comparing therapies for UC. However, as more pharmacotherapies become available, direct comparisons between each of these medications will not be feasible. The use of post hoc analyses, as these authors demonstrate, is one methodological approach to address this feasibility concern. However, another approach would be to perform a limited number of pragmatic head-to-head clinical trials between therapies of different classes. This work would not only improve our ability to directly compare treatments, but would add to the robustness of future network meta-analyses. Finally, the issue of comparative effectiveness calls into question the appropriateness for placebo-controlled trials in UC and the ethical concerns surrounding withholding treatment when there are several effective therapies available (J Crohns Colitis 2016;10:S548–S552). In summary, Narula et al use clinical trial data from ACT-1, GEMINI-1, and VARSITY to demonstrate superiority of infliximab over vedolizumab at achieving 1-year corticosteroid-free clinical and endoscopic remission. Although these outcomes may be meaningful clinically, what is perhaps more illuminating is the very need to use post hoc analysis and observational data to draw conclusions on comparative effectiveness. The most meaningful contribution from Narula et al may in fact be their ability to highlight the need for pragmatic head-to-head trials and perhaps even the phasing out of placebo-controlled trials for moderate-to-severe UC.