Concomitant Use of Immunosuppressive Therapy with Tumor Necrosis Factor (TNF) Antagonists in Inflammatory Bowel Disease

Crohn’s disease

Ulcerative colitis

SONIC [7]

COMMIT [9]

UC-SUCCESS [8]

IFX

IFX + AZA

IFX

IFX + MTX

IFX

IFX + AZA

Clinical remission (%)

Mucosal healing (%)

–

Antidrug antibody (%)

14.6

0.9

20.4

4.0

19.0

3.0

IFX concentration

1.6 μg/mL

3.5 μg/mL

3.8 μg/mL

6.4 μg/mL

–

IFX infliximab, AZA azathioprine, MTX methotrexate

At face value, these data would suggest that when using concomitant immunosuppressive therapy, providers should only use azathioprine, but several considerations need to be made when interpreting these results. First, COMMIT had no minimum disease activity requirement for entry which resulted in the recruitment of patients with a much milder disease course. Second, COMMIT used a high-dose steroid induction regimen in both treatment arms. Given the known treatment benefits of concomitant steroid use and impact of triple induction (steroids + azathioprine + infliximab) on treatment success [10], the use of high-dose steroids and recruitment of a population with a milder disease course and increased propensity to respond to therapy may have obscured the clinical benefit of concomitant methotrexate therapy [11, 12]. The measurable impact of methotrexate on antidrug antibodies and infliximab drug concentrations would suggest that a therapeutic benefit does exist when using this immunosuppressive agent.

No comparative effectiveness studies are currently available comparing TNF-antagonist monotherapy versus TNF-antagonist combination therapy for other TNF-antagonists such as adalimumab, certolizumab, and golimumab. In post hoc analyses, when stratifying RCTs by baseline immunosuppressive use, the concomitant use of an immunosuppressive appears to impact the pharmacokinetics of these TNF-antagonists, but this did not directly translate to improved treatment outcomes within these trials [6, 13–18] (Table 7.2).

Table 7.2

Stratified analysis of randomized controlled trials for TNF-antagonists according to baseline immunosuppressive use

	Agent	Antidrug antibody		Clinical remission
	Agent	TNF-antagonist monotherapy	Combination therapy	TNF-antagonist monotherapy	Combination therapy
PRECISE 2	CTZ	12%	2%	64%	61%
CLASSIC II	ADA	3.8%	0	45%	48%
PURSUIT	GOL	3.8%	1.1%	50%	44%

IFX infliximab, CTZ certolizumab, ADA adalimumab, GOL golimumab

Pooled analyses of RCTs and observational data for adalimumab have suggested that the use of concomitant immunosuppressive therapy results in improved rates of remission at 12 weeks compared to adalimumab monotherapy (OR 0.78, 95% 0.64–0.95). Although this would suggest that a clinical benefit may exist, this improved efficacy at 12 weeks did not translate to improved rates of remission at 52 weeks within this meta-analysis (OR 1.08, 95% CI 0.87–1.33) [19]. A second meta-analysis that pooled patient level data from three TNF-antagonist RCTs in CD (infliximab, adalimumab, certolizumab) similarly observed that no clinical benefit was present when adding an immunosuppressive agent to adalimumab [20]. Within this second meta-analysis, an interesting observation was that the use of concomitant immunosuppressive therapy was associated with a trend toward improved rates of remission at 6 months for infliximab (OR 1.73, 95% CI 0.97–3.07), but not adalimumab (OR 0.88, 95% CI 0.58–1.35) or certolizumab (OR 0.93, 95% CI 0.65–1.34) [20]. When interpreting these data, we must remember that a significant proportion of the adalimumab and certolizumab patients enrolled had failed infliximab therapy, and thus the use of concomitant immunosuppressive agents represents the continuation of an immunosuppressive agent when switching TNF-antagonists as opposed to starting an immunosuppressive agent de novo in these patients. Furthermore, this meta-analysis excluded patients naïve to immunosuppressive therapy and thus represents a step-up approach to combination therapy as opposed to the more efficacious top-down approach. These variations in observations help to highlight the fact that the timing of adding an immunosuppressive agent to TNF-antagonist is as important as the potential impact it has on TNF-antagonist pharmacokinetics . This concept is further supported by two RCTs showing that early combined immunosuppression is superior to traditional step-up algorithms.

The “top-down” trial is a randomized trial where 133 patients were randomized to either early combined immunosuppression with infliximab (ECI; n = 67) or conventional management (CM; n = 66) where patients received steroids followed in sequence by azathioprine and infliximab [21]. At 26 weeks, a higher proportion of patients in the ECI group were in steroid-free clinical remission without surgical resection as compared to the CM group (60% vs. 35.9%, p = 0.006), and this difference continued through week 52 (61.5% vs. 42.2%, p = 0.0278). At week 104, the rates of mucosal healing (absence of ulcers) were significantly higher in the ECI group as compared to the CM group (73% vs. 30.4%, p = 0.0028). Notably, the rates of serious adverse events were similar between both groups (30.8% vs. 25.3%, p = 1.0). This study was the first to demonstrate that the early use of combined immunosuppressive therapy impacted treatment outcomes. Although they were able to demonstrate statistically significant differences in outcomes that correlate with long-term disease-related complications (i.e., mucosal healing), the small size of the study precludes its ability to directly quantify the impact on outcomes of interest such as hospitalization, surgery, and overall complications.

The REACT trial is a cluster randomization trial in which community practices in Canada (n = 34) and Belgium (n = 5) were randomly assigned in a 1:1 ratio to either ECI with a TNF-antagonist (ECI; n = 21 centers, n = 1084 patients) or CM where immunosuppression and TNF-antagonist use were determined by the primary provider (CM; n = 18 centers, n = 898 patients) [22]. The primary outcome (remission as defined by a Harvey-Bradshaw score (HBS) ≤4 in the absence of steroids) was achieved in a similar proportion of the ECI and CM groups at 12 (66% vs. 62%, p = 0.65) and 24 months (73% vs. 65%, p = 0.35). Within this study, however, a significantly higher proportion of patients in the ECI group received combination immunosuppressive/TNF-antagonist combination therapy at 12 months (15.1% vs. 6.5%, p < 0.001) and 24 months (19.7% vs. 9.6%, p < 0.001), and the ECI group had highly significant and clinically important reductions in the rates of complications (HR 0.74, 95% CI 0.62–0.89) and surgeries (HR 0.68, 95% CI 0.49–0.95) and the combined outcome of hospitalizations, complications, and surgeries (HR 0.74, 95% CI 0.62–0.87). Within this trial, an important point to be noted is that they followed a treat-to-target algorithm where adjustments in therapy were made if patients had not achieved clinical remission at 3–6-month intervals. This approach may have factored into the overall impact of ECI and suggests that the timing of concomitant immunosuppressive therapy and the manner in which we monitor and adjust dosing are equally important.

In aggregate, direct comparative effectiveness studies demonstrate that the concomitant use of an immunosuppressive agent improves treatment outcomes and reduces disease-related complications in IBD. The optimal approach to using concomitant immunosuppressive therapy with TNF-antagonists is early in the disease course with frequent monitoring and adjustments in dosing or therapies when clinical remission has not been achieved. The ideal choice of which immunosuppressive agent to use appears to be azathioprine (6-mercaptupurine can likely be used as well) based on efficacy, but providers will need to take into consideration differences in trial characteristics and variations in outcomes across trials. Consideration for immunosuppressive safety will therefore likely drive the decision as to which immunosuppressive agent is chosen on an individualized basis.

Safety

When taking into consideration the optimal use of a therapeutic agent or the combination of therapies, we must take into consideration the impact safety will have on patient outcomes and adherence. Specifically, we must understand the safety profile when immunosuppressive agents are added or continued alongside biologic agents and the populations at greater risk for adverse events when using concomitant immunosuppressive therapy. Two of the most notable safety concerns with concomitant immunosuppressive therapy are serious infections and malignancy.

Serious Infections

Treatment-related serious infections can be broadly categorized as those resulting in the interruption or discontinuation of therapy, hospitalization, or death. Although randomized controlled trials have not demonstrated an increased risk for serious infections with the addition of immunosuppressive agents to biologics, population-based studies have observed an increased incremental risk, with the majority of this risk being attributed to the concomitant use of steroids [23–27]. Any prednisone use can increase the risk of serious infections, but doses higher than 20 mg of prednisone for 2 or more weeks are associated with the most significant risk for serious infections, and this risk persists for up to 90 days after exposure [28, 29]. This risk can be further augmented in certain patients who are already at an increased baseline risk for treatment-related se rious infections. Two subgroups of particular importance are the elderly (≥65 years) and those on chronic narcotics [6, 28, 30, 31]. The exact mechanism through which narcotics increase the risk for serious infections and mortality is unclear, and this may simply serve as a proxy for more complicated disease, disease-related complications, or disease severity, the latter also being independently associated with an increased risk for infectious complications [28, 32–35].

Two important opportunistic infections that should be specifically considered when starting a concomitant immunosuppressive agent are hepatitis B and Clostridium difficile (C. diff). The use of immunosuppressive medications increases the risk for hepatitis B reactivation, with the greatest risk being seen in patients who are hepatitis B DNA and/or surface antigen positive being treated with long-term combination therapy with TNF-antagonists [36–38]. The occurrence of C. diff in IBD is associated with higher morbidity and mortality as compared to the general population [39, 40], and the use of immunosuppressive therapy, but not biologics (TNF-antagonists), has been associated with an increased risk of developing C. diff [36].

Malignancy

One of the most important considerations to be made when using concomitant immunosuppressive therapy in IBD is the potential increased risk for developing malignancy [41]. IBD patients are at an increased risk for malignancy at baseline [42–45], and the use of TNF-antagonists does not appear to increase this risk overall [46]. The concomitant use of immunosuppressive therapy, however, is clearly linked to an increased risk for malignancy and, in particular, an increased risk for lymphoma [47–49]. This increased risk for lymphoma is seen with concomitant thiopurine use, and the two populations at greatest risk for lymphoma development are the elderly (≥65 years) and young (≤35 years) males who are at a particular increased risk for the development of a fatal lymphoma subtype, hepatosplenic T-cell lymphoma (HSTCL). In both subgroups, the risk of lymphoma is duration dependent, with the greatest risk being seen after 2 years of use [49–54]. Another important malignancy linked to thiopurine use is skin cancer. Several studies have now demonstrated that thiopurines increase the risk for nonmelanoma skin cancer (NMSC) development, and this increased risk is nearly doubled when used in combination with a TNF-antagonist [48, 55, 56]. The risk of melanoma, however, appears to be increased by the use of TNF-antagonists but not immunosuppressive agents, and this risk is potentially higher among patients receiving long-term TNF-antagonist therapy [56].

Opportunities to Optimize the Use of Concomitant Immunosuppressive Therapy

When combining safety and efficacy data for the use of concomitant immunosuppressive therapy, several opportunities arise to optimize the personalization of these treatment decisions. (Fig. 7.1) Based on prior RAND appropriateness panels, systematic reviews, and our review of the literature, the use of concomitant immunosuppressive therapy as a general rule of thumb appears to be most appropriate for IBD patients with extensive disease or those at risk for disease-related complications (i.e., steroid dependent or refractory) [6, 18, 57]. Among these individuals, the decision to personalize the use of concomitant immunosuppressive therapy can be made through an assessment of safety and long-term risks. In young males at risk for HSTCL, short-term use of concomitant thiopurines, TNF-antagonist monotherapy, or the use of methotrexate as the concomitant immunosuppressive agent may be most appropriate given the fatal nature of this lymphoma. In the elderly or individuals at risk for lymphoma with extended use of thiopurines, TNF-antagonist monotherapy, discontinuation of the thiopurine after 1–2 years of use, or switching to methotrexate may be reasonable options. The risk of malignancy in patients using thiopurines rises exponentially after 2 years of use [50, 58], and the risk of malignancy in patients discontinuing thiopurines appears to return to the baseline risk seen in patients without prior exposure [58]. Thus, withdrawing the thiopurine after 2 years of use can be considered, particularly in patients at low risk for relapse upon immunosuppressive withdrawal [18] (Table 7.3). This approach is however associated with reductions in TNF-antagonist drug concentrations and the development of antidrug antibodies, so patients should be followed up closely to optimize TNF-antagonist dosing as needed [59–62].

Fig. 7.1

Considerations when determining whether to use a concomitant immunosuppressive agent with TNF-antagonists. IBD inflammatory bowel disease, AE adverse event, AZA azathioprine, 6MP 6-mercaptopurine, anti-TNF tumor necrosis factor antagonist, MTX methotrexate, IS immunosuppressive

Table 7.3

Factors associated with disease relapse after stopping an immunosuppressive agent

Extensive disease or elevated inflammatory markers (CRP, platelet count, white blood count)

Evidence of mucosal activity on endoscopy

Short duration in remission prior to stopping

Short duration of steroid-free remission

In individuals at an increased risk for serious infections or other thiopurine-related adverse events, opportunities to optimize the use of concomitant immunosuppressive therapy can be taken through modifying risk factors and thiopurine metabolite assessments [63–67] (Fig. 7.2). The traditional therapeutic efficacy window for thiopurine monotherapy is a thioguanine (TGN) level of between 235 pmol/8 × 10⁸ and 450 pmol/8 × 10⁸ RBCs. When using thiopurines as concomitant immunosuppressive agents, however, the therapeutic efficacy window for reducing antidrug antibodies may be lower (125 pmol/8 × 10⁸ RBCs) [68–70]. A crude measure of TGN levels is peripheral RBC mean corpuscular volume (MCV) [71]. In a post hoc analysis of SONIC, patients who had achieved a mean increase in MCV of 7 were more likely to be in steroid-free remission, achieve mucosal healing, and obtain an infliximab concentration of >3, as compared to those who hadn’t achieved a delta change in MCV of 7 or more. As patients will need regular blood test monitoring while on thiopurines, following peripheral MCV measurements could serve as a reliable interim surrogate for achieving optimal thiopurine concentrations.