Abstract
There has been an ongoing expansion in available biologic and small molecule therapies to treat moderate to severe ulcerative colitis (UC). The sequence in which these therapies are used has been a topic of high interest among clinicians. In the absence of precision medicine for UC, it would be useful to define the sequence of therapies that may predict the best outcomes. Limited data exists regarding the most effective treatment algorithms for achieving optimal patient outcomes. As such, this study endeavors to identify treatment algorithms when considering the UC therapies of vedolizumab, infliximab, ustekinumab, and tofacitinib.
A Markov model was used to identify the UC treatment algorithm that would potentially yield the greatest quality adjusted life years (QALYs) over 1 year. The base case was a 35-year old male with moderate to severe UC who had not previously received biologic or immunomodulator therapy. Four medical therapies were considered: (1) infliximab + azathioprine, (2) tofacitinib, (3) ustekinumab, and (4) vedolizumab. Colectomy was assumed to be the least desirable treatment option. These four therapies were combined in eight separate potential treatment algorithms (Figure 1). Tofacitinib was only included after anti-TNF therapy per its USA prescribing indication. With each therapy, simulated individuals could enter remission or response, develop a significant adverse event or infection, or lose response to the therapy. Transition probabilities for response, remission, and adverse event rates were derived from relevant clinical trials. Lymphoproliferative risks were derived from age-specific SEER (Surveillance, Epidemiology, and End Results) data and treatment-specific risks from CESAME. Utility estimates were derived from previously published studies. The time horizon was 1 year. Primary analyses consisted of first order Monte Carlo simulation of 100 trials of cohorts consisting of 100,000 individuals.
Algorithms incorporating vedolizumab as first-line therapy resulted in the greatest quality of life and proportion of patients in remission at 1 year in comparison to strategies where combination therapy with infliximab + azathioprine or ustekinumab were included as first line therapy (Table 1). Vedolizumab followed by infliximab + azathioprine and then ustekinumab resulted in the greatest quality of life advantage. The benefit of this treatment sequence over other sequences ranged from 0.008 to 0.011 QALYs at 1 year.
In this model, using vedolizumab prior to infliximab + azathioprine, ustekinumab and tofacitinib is predicted to yield greater quality of life and a greater percentage of patients with clinical remission at the end of 1 year. Further analyses assessing model inputs, probabilistic analyses, and strategy cost-effectiveness are required.
Figure 1. Schematic of eight simulated treatment algorithms. Abbreviations: AZA, azathioprine; IFX, infliximab; SAE, significant adverse event; VDZ, vedolizumab; UST, ustekinumab; TOF, tofacitinib; SAE, significant adverse event; NHL, non-Hodgkin’s lymphoma; LOR, loss of response
Table 1. Quality adjusted life years, incremental effectiveness, and percentage of cohort in steroid-free remission at the end of one year by treatment algorithms. Abbreviations: IFX, infliximab; QALYs, quality adjusted life years; SAE, significant adverse event; VDZ, vedolizumab; UST, ustekinumab; TOF, tofacitinib.
