Novel Microbial-Based Immunotherapy Approach for Crohn's Disease

Abstract

Background: Current Crohn's disease (CD) therapies focus on suppressing immune function and come with consequent risk, such as infection and cancer. Notwithstanding, most CD patients still experience disease progression. There is a need for new CD treatment strategies that offer better health outcomes for patients. Aims: To assess safety, efficacy, and tolerability of a novel microbial-derived immunotherapy, QBECO, that aims to restore rather than suppress immune function in CD. Methods: A randomized, double-blind, placebo-controlled trial was conducted in 68 patients with moderate-to-severe CD. Primary endpoints: safety and Week 8 clinical improvement. Secondary endpoints: Week 8 clinical response and remission. Week 8 responders continued blinded treatment through Week 16; non-responders received open-label QBECO from Weeks 9-16. Exploratory analyses included immune biomarker and genotype assessments. Results: QBECO was well-tolerated. Mean reduction in Crohn's Disease Activity Index (CDAI) score was -68 for QBECO vs. -31 for placebo at Week 8. Improvement with QBECO continued through Week 16 (-130 CDAI reduction). Week 8 QBECO clinical response, improvement and remission rates were 41.2%, 32.4%, 29.4% vs. 26.5%, 23.5%, 23.5% for placebo. TNFα inhibitor-naïve subjects achieved higher response rates at Week 8 with QBECO (64%) vs. placebo (26%). Specific immune biomarkers were identified that linked to QBECO response. Conclusion: This proof-of-concept study supports further investigation for the use of QBECO as a novel immunotherapy approach for CD. Biomarker analyses suggests it may be feasible to personalize CD treatment with QBECO. Larger trials are now needed to confirm clinical improvement and the unique biological findings. Clinical Trial Number: NCT01809275 (https://ichgcp.net/clinical-trials-registry/NCT01809275).

Keywords: Crohn's disease; biologic; biomarkers; immunotherapy; innate immunity; microbial-based therapy; randomized placebo-controlled trial.

Figures

Figure 1

Patient flow by treatment assignment. Subjects with moderate-to-severe Crohn's disease were randomized 1:1 to QBECO or Placebo for 8 weeks. Those responding to allocated blinded treatment continued blinded treatment for another 8 weeks; all 8 Week non-responders commenced open-label QBECO for 8 weeks.

Figure 2

Mean change in Crohn's Disease Activity Index (CDAI) score from baseline to Week 8 by treatment group. The mean reduction in CDAI from baseline to Week 8 in patients with moderate-to-severe Crohn's disease blinded to QBECO (red bar) or Placebo (blue bar) treatment. ITT, Intention to Treat analysis; PP, Per Protocol analysis.

Figure 3

Mean change in Crohn's Disease Activity Index (CDAI) score in study groups Week 8 and Week 16. Mean change in CDAI score for subjects randomized to QBECO (solid red line) and placebo (solid blue line) from baseline to Week 8 and Weeks 16 of study treatment. Those responding to allocated blinded treatment at week 8 continued blinded treatment for another 8 weeks [light blue solid line for responders originally blinded to placebo treatment (n = 9); light red solid line for QBECO responders (n = 13)]. All placebo (n = 22) and QBECO (n = 14) non-responders at week 8 received open-label QBECO (dashed red lines) for weeks 9–16. Dark solid red line represents the average of all subjects on QBECO for weeks 9–16.

Figure 4

Serum cytokines that differed in QBECO responders and non-responders. A 42-plex cytokine/chemokine analysis was performed on serum. Four cytokines- IFNγ, IL-12p70, IL-17A, and TGFα- differentiated QBECO responders from non-responders over the study period after adjusting for multiple comparisons.

Figure 5

Gene risk score separates QBECO responders from non-responders. One hundred and thirteen inflammatory bowel disease (IBD)-related SNPs were included in computing the gene risk score for response to QBECO to assess the potential contribution of subjects' genetics to treatment outcome. The derived gene risk score could successfully distinguish QBECO responders from non-responders, p = 0.0000243.