Background: Objective measures of Crohn's Disease (CD) are sought to monitor disease and therapeutic activity. Biopsy mRNA and serum protein analyses were completed to assess the molecular impact of ustekinumab (UST), an anti-IL-12/23 p40 monoclonal antibody, on CD.
Methods: The phase 3 UNITI-1 & 2 studies evaluated the safety and efficacy of IV UST induction (I) in patients (pts) with moderately-to-severely active CD who had failed ≥1 TNF antagonists (UNITI-1) or conventional therapies (UNITI-2). UST induction responders (R) entered IMUNITI to evaluate SC UST maintenance (M) therapy. Gene expression was profiled using biopsies from UNITI-1 (n=69) and UNITI-2 (n=170) pts at Iwk0, Iwk8 & Mwk44. UST effects on transcriptome were assessed via Gene Set Variation Analysis (GSVA). Ten protein analytes including SAA, IL17A & F, MMP1, 3 & 9, MPO, TNFa, IFNg & IL6 were measured in serum from UNITI-1 (n=766) & UNITI-2 (n=593) pts at Iwk0, Iwk6, Mwk8 & Mwk44. Serum and biopsies from 14–30 healthy subjects was analyzed as control (cntl) for each dataset. Modulations with |fold change| >1.5× and p<0.05 were considered significant.
Results: Biopsy Transcriptome: CD expression profiles in UNITI-2 pts were significantly normalized by both UST induction and maintenance therapies while the UST normalization was notable in R in UNITI-1 but not significant. A trend of greater effects with UST 90 mg SC q8w vs q12w was observed in both cohorts at Mwk44. Serum: CD pts in UNITI-1 & 2 had similar serum profiles with SAA, IFN-γ, IL-17A & MMP9 significantly elevated vs healthy cntl while the elevation of SAA was greater in UNITI-1. IFNγ was identified as a pharmacodynamic marker, with similar significant modulation by UST in R and non-responders (NR) (greater effects with UST 6 mg/kg vs. 130 mg dosing), but not in PBO-treated pts. SAA & IL-6 were significantly reduced by UST induction in R and less so or not at all in NR. These markers remained reduced with SC UST maintenance. Elevated IL-17A and MMPs in CD were normalized, but not significantly, following UST induction in R. The trend of normalization became larger during UST maintenance, with MMPs achieving significance in UST R in UNITI-2, and IL-17A in both cohorts. TNFa was uniquely elevated in UNITI-1 vs healthy cntl, but it was not significantly normalized by UST therapies. Placebo induction pts did not show notable changes in biopsy mRNA or serum proteins.
Conclusions: Transcriptomic and protein analyses in the Phase 3 UST studies demonstrate normalization of CD-associated markers in induction which were maintained or magnified during maintenance. These results provide insight into the mechanisms of UST efficacy and identify potential biomarkers to monitor CD activity and UST pharmacodynamics.