Baseline Serum and Stool Microbiome Biomarkers Predict Clinical Efficacy and Tissue Molecular Response After Ritlecitinib Induction Therapy in Ulcerative Colitis

Abstract

Background and aims: Ritlecitinib, an oral JAK3/TEC family kinase inhibitor, was well-tolerated and efficacious in the phase 2b VIBRATO study in participants with moderate-to-severe ulcerative colitis [UC]. The aim of this study was to identify baseline serum and microbiome markers that predict subsequent clinical efficacy and to develop noninvasive serum signatures as potential real-time noninvasive surrogates of clinical efficacy after ritlecitinib.

Methods: Tissue and peripheral blood proteomics, transcriptomics, and faecal metagenomics were performed on samples before and after 8 weeks of oral ritlecitinib induction therapy [20 mg, 70 mg, 200 mg, or placebo once daily, N = 39, 41, 33, and 18, respectively]. Linear mixed models were used to identify baseline and longitudinal protein markers associated with efficacy. The combined predictivity of these proteins was evaluated using a logistic model with permuted efficacy data. Differential expression of faecal metagenomics was used to differentiate responders and nonresponders.

Results: Peripheral blood serum proteomics identified four baseline serum markers [LTA, CCL21, HLA-E, MEGF10] predictive of modified clinical remission [MR], endoscopic improvement [EI], histological remission [HR], and integrative score of tissue molecular improvement. In responders, 37 serum proteins significantly changed at Week 8 compared with baseline [false discovery rate of <0.05]; of these, changes in four [IL4R, TNFRSF4, SPINK4, and LAIR-1] predicted concurrent EI and HR responses. Faecal metagenomics analysis revealed baseline and treatment response signatures that correlated with EI, MR, and tissue molecular improvement.

Conclusions: Blood and microbiome biomarkers stratify endoscopic, histological, and tissue molecular responses to ritlecitinib, which may help guide future precision medicine approaches to UC treatment. ClinicalTrials.gov NCT02958865.

Keywords: Biomarkers; JAK inhibitor; ulcerative colitis.

Figure 1

Differential analysis. [A] Differentially expressed genes in inflamed and noninflamed tissues at baseline [left] and Week 8 following ritlecitinib therapy [right] in responders and nonresponders [MR]. [B] First column from left: baseline gene expression changes between inflamed and noninflamed tissues; other columns: gene expression changes from baseline following Week 8 treatment. [C] Gene set improvement scores using treatment-modulated genes in all participants; responders vs nonresponders are defined by endoscopic improvement and modified clinical remission, respectively. [D] Pathways modulated following ritlecitinib therapy at Week 8 in responders and nonresponders based on inflamed tissue proteomics and transcriptomics. CFB, change from baseline; ECM, extracellular matrix; GSI, gene set improvement; IL, interleukin; JAK, Janus kinase; LS, lesions; NL, nonlesions; Th, T helper.

Figure 2

Genes with treatment changes in responders and nonresponders in endoscopic improvement and modified clinical remission: [A] genes changing in responders but not in nonresponders at Week 8 of ritlecitinib therapy; [B] gene set improvement scores at all dose levels in responders and nonresponders.

Figure 3

[A] Serum protein baseline markers to determine responders and nonresponders in participants with EI or MR in the combined 70- and 200-mg ritlecitinib treatment groups. [B] Area under the ROC using a logistic model predicting EI and MR at Week 8 with HLA-E, LTA, CCL21, and MEGF10 proteins based on participants treated with 200 mg ritlecitinib. AUC, area under the curve; EI, endoscopic improvement; MR, modified clinical remission; ROC, receiver operating curve.

Figure 4

Serum protein signatures of EI and HR after ritlecitinib treatment at Week 8: [A] area under the ROC using a logistic model predicting EI and HR at Week 8 with IL4R, TNFRSF4, SPINK4, and LAIR-1 changes based on participants treated with 200 mg ritlecitinib; [B] baseline gene expression changes of IL4R, TNFRSF4, SPINK4, and LAIR1, between inflamed and noninflamed colon biopsy. AUC, area under the curve; EI, endoscopic improvement; HR, histological remission; ROC, receiver operating characteristic.

Figure 5

Faecal metagenomic analysis: differential expression analysis between responders and nonresponders for [A] endoscopic improvement and [B] modified clinical remission; [C] baseline taxa biclustering using the k-means algorithm, with differential responses for each cluster in participants with active treatment; [D] GSI of meta-analyses derived from the UC transcriptome, using clusters 1 and 2; [E] inosine 5-phosphate degradation pathway in responders and nonresponders at Week 8 based on modified remission. GSI, gene set improvement; NR, nonresponder; R, responder; UC, ulcerative colitis.