Combination therapy is not associated with improved rates of clinical or endoscopic remission in patients with inflammatory bowel disease treated with ustekinumab or vedolizumab: a retrospective study

Introduction

Treatment options for patients with inflammatory bowel disease (IBD), which includes ulcerative colitis (UC) and Crohn’s disease (CD), have been growing over the last decade. Steroids, immunomodulators such as thiopurines and methotrexate, and tumor necrosis factor (TNF)-α inhibitors such as infliximab (IFX) and adalimumab (ADA) have been the mainstay of treatment for the past decade [1-3]. There are now newer biologic monoclonal agents that target their own distinct immunological pathway, notably the anti-integrin antibody vedolizumab (VDZ) and the interleukin 12/23 antibody ustekinumab (UST).

Immunomodulators have commonly been used in combination with an anti-TNF-α to provide a synergistic effect from treatment, as well as to prevent immunogenicity of the biologic agent. The rate of immunogenicity and loss of response is quite variable amongst studies, with an estimated risk of around 13% per patient-year for IFX [4-6]. Subsequently, the use of combination therapy with immunomodulators and IFX to reduce rates of immunogenicity has been demonstrated in the SONIC and UC-SUCCESS trials [7,8]. The benefit of combination treatment for ADA is less clear, with mixed results in the literature [9,10]. It is now considered standard practice to use combination therapy with anti-TNF-α agents, in particular with IFX [11]. It is less clear whether combination therapy confers any benefit with the newer agents, including VDZ and UST, given that they work on a different immunological pathway. VDZ has a significantly lower rate of immunogenicity than IFX, at around 4% [12], while UST is around 4.6% [13].

Evidence for VDZ combination therapy is mixed, with one study of 164 patients showing treatment failure at 1 year in 9.3% of UC patients on combination therapy, compared to 26.3% for VDZ monotherapy [14]. Another multicenter cohort study of 136 patients found better clinical response and remission in CD patients on combination therapy, with an odds ratio (OR) of 2.71 (95% confidence interval [CI] 1.11-6.57), but not for UC (OR 0.22, 95%CI 0.05-0.88) [15]. This contrasted with a post hoc analysis in the GEMINI trial, along with a real-world cohort study showing no difference in outcome for combination therapy for both UC and CD compared to VDZ monotherapy [16,17].

There are also mixed results with UST. A study of 122 anti-TNF-α refectory CD patients demonstrated that immunomodulator use in combination with UST was a predictive factor for clinical efficacy at 3 months (OR 5.43, 95%CI 1.14-25.77; P=0.03) [18], whereas 3 other studies found no difference in clinical outcomes [13,17,19].

Based on the current literature, it is unclear whether combination therapy for VDZ or UST is associated with better clinical and endoscopic outcomes for UC and CD, in terms of either a synergistic effect from treatment or the prevention of immunogenicity by the biologic agent. Thiopurines and methotrexate also have several adverse effects, with particular concerns being an increased risk of infection and malignancy, including lymphoma [20]. Avoiding their use would improve the safety outcomes for patients with IBD. The aim of this retrospective cohort study was to determine whether there was a difference in clinical or endoscopic outcomes for UST or VDZ as monotherapy, compared to their combination with thiopurines or methotrexate.

Patients and methods

Results

A total of 159 patients were included in the study. Of these, 85 (53%) of patients received treatment with VDZ and 74 (47%) with UST. In the VDZ group, 61 (72%) patients were diagnosed with UC and 24 (28%) with CD. In the UST group, all patients (100%) were diagnosed with CD. The characteristics of the VDZ and UST groups are detailed in Tables 1 and 2, respectively.

Table 1 Characteristics of patients with inflammatory bowel disease on vedolizumab monotherapy or in combination with another immunomodulator

Table 2 Characteristics of patients with inflammatory bowel disease on ustekinumab monotherapy or in combination with another Immunomodulator

In the VDZ group, 56 (66%) were on VDZ monotherapy, with the remainder (33%) on combination therapy. The median age was higher in the VDZ monotherapy group, at 33±11.0 years, compared to 26±13.0 years in the combination group (P=0.03). Disease duration prior to starting VDZ was similar in both groups (9.4±9.1 years and 9.9±9.0 years for VDZ monotherapy and combination therapy, respectively; P=0.79). Steroid use at the time of induction did not differ significantly between the groups: 15 (27%) for VDZ monotherapy and 5 (17%) for combination therapy (P=0.30). For those on VDZ monotherapy, 44 (79%) had previously used a thiopurine and 6 (11%) had used methotrexate. For those on combination therapy 26 (90%) were being treated with a thiopurine, compared to 3 (10%) on methotrexate. There was no difference in previous exposure to anti-TNF-α agents: 31 (55%) and 19 (66%) for VDZ monotherapy and combination therapy, respectively (P=0.55). Only 24 (43%) of the VDZ monotherapy and 10 (35%) of the combination therapy patients were biologic naïve prior to starting VDZ (P=0.86). Baseline mean HBI and Mayo scores were similar between the groups for VDZ monotherapy (HBI 7.6±5.0, partial Mayo 3.9±1.8; P=0.29) and combination therapy (HBI 5.8±2.0, partial Mayo 3.9±2.1; P=0.79).

In the UST group, 53 (72%) were on UST monotherapy and 21 (28%) on combination therapy. The median age was similar, 32 years and 34 years for UST monotherapy and combination group, respectively (P=0.69), as was disease duration prior to starting UST: 12.3±11.0 years and 13.5±9.5 years (P=0.67). Steroid use at the time of induction did not differ significantly between the groups: 9 (17%) for UST monotherapy and 5 (24%) for combination therapy (P=0.50). For those on UST monotherapy, 38 (72%) had previously used a thiopurine and 6 (11%) had used methotrexate. Among those on combination therapy, 16 (76%) were being treated with a thiopurine, compared to 5 (24%) on methotrexate. There was no difference in the previous rate of anti-TNF-α agents: 36 (68%) and 16 (76%) for VDZ monotherapy and combination therapy, respectively (P=0.48). A smaller proportion of patients were biologic naïve compared to the VDZ group, with 16 (30%) on UST monotherapy and 5 (24%) on combination therapy (P=0.39). Baseline mean HBI scores differed significantly between the groups: 4.9±3.2 for UST monotherapy and 6.8±3.8 for combination treatment (P=0.04).

There was no difference in the primary endpoint of clinical response or remission for VDZ monotherapy compared to combination therapy (Table 3) at week 12 (80% vs. 70%; P=0.38), week 26 (82% vs. 80%; P=0.83) and week 52 (82% vs. 91%; P=0.32). Twenty-five of the 56 (45%) patients on VDZ monotherapy and 18 of the 29 (62%) on combination therapy had a 1-year colonoscopy, with no difference in endoscopic remission (66% vs. 78%; P=0.36). Similarly, there was no difference in the rate of treatment failure or relapse (23% vs. 19%; P=0.80).

Table 3 Outcomes for patients with inflammatory bowel disease on vedolizumab monotherapy or in combination with another Immunomodulator

In the UST group (Table 4), there was also no difference in clinical response or remission for UST monotherapy compared to combination therapy at week 12 (78% vs. 93%; P=0.21), week 26 (86% vs. 89%; P=0.74) and week 52 (97% vs. 88%; P=0.23). Twenty of the 53 (38%) patients on VDZ monotherapy and 18 of the 29 (62%) on combination therapy had a 1-year colonoscopy, with no difference in endoscopic remission (70% vs. 63%; P=0.70). There was also no difference in the rate of treatment failure or relapse (21% vs. 24%; P=0.17).

Table 4 Outcomes for IBD patients on ustekinumab monotherapy or in combination with another immunomodulator

Discussion

VDZ and UST are safe and effective treatment options for the treatment of both CD and UC. This study demonstrated similar rates of clinical response and remission, as well as endoscopic remission, in those treated with monotherapy compared with combination therapy. There were no differences in the rate of treatment failure or relapse between those on monotherapy and those on combination therapy.

Importantly, when the VDZ and UST combination and monotherapy cohorts were compared the groups were relatively well matched, with no major differences. It is noteworthy that, while a proportion of the patients were on steroids during induction for monotherapy and combination therapy with VDZ and UST, there was no statistical difference between these groups. Similarly, previous biologic exposure did not differ significantly among all groups. Given that these groups were similar with regard to both steroid and previous biologic use, we would not expect these variables to affect our findings. There were also varied reasons for treatment failure or relapse, spread among the subgroups of dose escalation, reinduction, change in agent, IBD-related surgery, or IBD-related hospitalization.

This study supports the existing evidence suggesting that there is no benefit from combination therapy for VDZ or UST. Notably, this has been shown in a post hoc analysis for a small subgroup of the GEMINI trial, and a larger cohort study of patients in the United States, Canada and France, which both showed no difference in the outcomes of combination therapy for both UC and CD compared to VDZ monotherapy [16,17]. The latter study found that, for VDZ, there was no difference in clinical response or remission with combination therapy compared to monotherapy at week 54 (78.3% vs. 72.9%; P=0.33). For UST, the difference at week 54 was similar (62.1% vs. 67.0%; P=0.52).

The mechanism for immunomodulators in the prevention of immunogenicity is still slightly unclear. It is thought that the addition of immunomodulators to biologic therapy (monoclonal therapy in particular) causes the immune system to be suppressed, which then results in a decrease in antibody formation. Therefore, the addition of thiopurines or methotrexate to biologic therapy could prevent the activation of immune cells and reduce the number of anti-drug antibodies [21]. As the risk of immunogenicity is much lower for VDZ (around 4%) and UST (around 4.6%), it may be that any benefit that could be realized from the addition of an immunomodulator is non-significant [12,13].

Furthermore, there are significant drawbacks to the use of immunomodulators [20]. These include quite significant nausea and leukopenia, as well as an increased risk of serious infection or cancer, and lymphoma in particular. The issue of serious infection is compounded when these drugs are used in addition to biologic therapy.

There are some limitations to this study. It was a retrospective study and there may have been an underlying bias in the selection of patients with more severe disease for combination therapy. For VDZ there was no statistical difference between baseline HBI or partial Mayo scores between the groups; however, in the UST group those on combination therapy had a higher baseline HBI. The study was also limited by the data available from our cohorts, with limited numbers in some subgroups. In addition, as this was a retrospective trial, some variables were incomplete for each time period assessed and were not included in the analysis. Similarly, additional variables, including drug levels and anti-drug antibodies, C-reactive protein and fecal calprotectin, were not included because of missing data. Future research could include a structured prospective study, which would allow for more regular and systematic collection of biomarkers and drug levels, as well as removing any inherent bias in the patient selection.

In conclusion, patients with IBD commencing either VDZ or UST have similar rates of clinical and endoscopic efficacy whether the drug is used as a monotherapy or as part of a combination therapy. There is conflicting evidence for monotherapy in the literature, and this paper adds to the growing evidence that combination VDZ or UST therapy does not lead to better outcomes. Combination therapy with an immunomodulator (thiopurine or methotrexate) can be avoided, which reduces the risks associated with these agents.