Background: Despite increasing therapeutic options in the treatment of ulcerative colitis (UC), achieving disease remission remains a major clinical challenge. Nonresponse to therapy is common and clinicians have little guidance in selecting the optimal therapy for an individual patient. This study examined whether patient-derived materials could predict individual clinical responsiveness to the Janus kinase (JAK) inhibitor, tofacitinib, prior to treatment initiation.

Method: In 48 patients with UC initiating tofacitinib, we longitudinally collected clinical covariates, stool, and colonic biopsies to analyze the microbiota, transcriptome, and exome variations associated with clinical responsiveness at week 24. We established patient-derived organoids (n = 23) to determine how their viability upon stimulation with proinflammatory cytokines in the presence of tofacitinib related to drug responsiveness in patients. We performed additional biochemical analyses of organoids and primary tissues to identify the mechanism underlying differential tofacitinib sensitivity and potential rescue by upadacitinb.

Results: The composition of the gut microbiota, transcriptome, biomarkers, and exome variants were indistinguishable among UC patients prior to tofacitinib treatment. A subset of patient-derived organoids displayed reduced sensitivity to tofacitinib as determined by loss of viability upon cytokine stimulation (Figure 1). UC organoids in which tofacitinib increased viability in both co-stimulation conditions (rTNFα+rIFNβ, rTNFα+rIFNγ) were categorized as tofacitinib-sensitive (UC-S). UC patient organoids insensitive to tofacitinib (UC-IS) were divided into 2 subgroups: (UC-IS1) organoids in which tofacitinib failed to increase viability in both conditions, and (UCIS-2) organoids in which tofacitinib increased viability with rTNFα+rIFNβ, but not rTNFα+rIFNγ. In UC-IS organoids, tofacitinib fails to inhibit pSTAT1 phosphorylation in the presence of cytokines, resulting in cell death (Figure 2). The sensitivity of organoids to tofacitinib predicted individual patient response. Reduced response to tofacitinib was associated with decreased levels of MATE1, a cationic transporter. Higher MATE1 levels in primary tissues were maintained in organoids and predicted clinical response to tofacitinib. Upadacitinb rescued the viability of organoids insensitive to tofacitinib due to reduced MATE1. Inhibitors of OCT1, another cationic transporter, negated the ability of upadacitinb to rescue viability.

Conclusions: Patient-derived intestinal organoids from patients with UC predict and identify mechanisms of individual tofacitinib and upadacitinb responsiveness. Expression of the transporter MATE1 in the gut epithelium may be a clinical biomarker for determining tofacitinib responsiveness whereas OCT 1 expression may predict response to upadacitinib.

Defne Ercelen

Appearances