Upadacitinib Outcomes After Tofacitinib Failure in Patients With Moderate to Severe Ulcerative Colitis Authors: Krishan Pratap, Abdulrahman Bassam A. AlBassam, Neeraj Narula Published: 2026 ABSTRACT ABSTRACTEvidence for upadacitinib in ulcerative colitis (UC) after tofacitinib failure is limited, with unclear long-term outcomes. A single-centre retrospective case series at a Canadian tertiary hospital included adult UC patients who lost response to tofacitinib and were subsequently treated with upadacitinib (January 2016 to March 2024). Nine patients, who failed a median of four therapies, received upadacitinib. Eight completed induction, and four archived steroid-free clinical remission during maintenance, supported by endoscopic and histologic improvement. Durable remission beyond one year was observed in these four patients, suggesting that upadacitinib may be effective in a subset of patients after tofacitinib failure. INTRODUCTION Ulcerative colitis (UC) is a subtype of inflammatory bowel disease (IBD) which occurs because of immune dysregulation and microbial dysbiosis.1 Currently, there are 6 broad classes of advanced therapies (biologic and small molecule therapies) available to manage UC. Currently, there are 2 approved Janus kinase (JAK) inhibitors to treat UC, tofacitinib and upadacitinib. Both are oral small molecules with tofacitinib inhibiting JAK1 and 3 receptors and upadacitinib inhibiting JAK1 to modulate the JAK-signal transducer and activator of transcription (STAT) signaling pathway.2,3 These agents are both highly efficacious and have rapid onset within symptom improvement typically seen in the first 3–7 days.4,5 Owing to these characteristics, these agents have emerged as a favored therapy for patients with moderate to severe UC. This was highlighted in the TRIUMPH study which found that tofacitinib is an effective rescue therapy in steroid refractory acute severe UC.6 A recent systematic review and meta-analysis by Shehab et al confirmed that upadacitinib was found to have the highest likelihood of clinical response as well as endoscopic and histologic in patients with UC.7 Furthermore, a systematic review and Bayesian network meta-analysis by Jairath et al demonstrated that upadacitinib was able to achieve the highest rates of response and remission in patients previously exposed to advanced therapies.8 While tofacitinib and upadacitinib are within the same drug class, this demonstrates that there are notable differences between the clinical efficacy of upadacitinib and tofacitinib. In patients with UC with previous tofacitinib failure, there is limited evidence regarding the responsiveness of upadacitinib in these patients.9 A recent systematic review and meta-analysis by Hasan et al, demonstrated that there is clinical response to upadacitinib in patients with UC that had previous tofacitinib failure; however, most of the studies included in their review assessed outcomes during the induction phase and for a brief period after the induction phase with limited data regarding durable longer term outcomes. There remain a paucity of data regarding the utility of upadacitinib in the maintenance phase for these patients. Therefore, we aimed to describe a case series of patients with UC with failure to have response or lost response to tofacitinib and their long-term outcomes after commencement of upadacitinib. CASE REPORT Study design This study was designed as a single-center retrospective case series at the McMaster University Medical Center. Patients were identified through a search of patient lists obtained from patient support programs and confirmed through search of electronic medical records at the Digestive Diseases Clinic. Inclusion criteria were patients aged 17 or older diagnosed with UC, patients diagnosed with UC, and are prescribed upadacitinib with any history of primary or secondary loss of response to tofacitinib, with upadacitinib prescriptions made between January 2016 and the March 1, 2024. Patients were excluded if there were missing or incomplete medical records. Outcomes Our primary outcomes were to determine if patients with UC with previous tofacitinib failure can achieve steroid-free clinical remission with upadacitinib. This is defined based on a patient reported outcome (PRO)-2 score ≤2 based on stool frequency (SF) and rectal bleeding (RB) or physician's determination of clinical remission. Secondary outcomes include determining if clinical response and remission of upadacitinib to upadacitinib are durable in the subsequent 12 months after commencement of upadacitinib in patients with UC following previous use and tofacitinib failure, defined as maintenance of steroid-free clinical remission with PRO2 ≤2 without the need for interim systemic steroids, hospitalizations for UC, or surgery. Other outcomes of interest include endoscopic improvement [Mayo endoscopic score (MES) 0 or 1], clinical response (≥50% decrease from baseline PRO2 score), histologic response (inactive/resolved, mild, moderate, and severe), and adverse events. Furthermore, need for corticosteroids, surgery, or hospitalization was collected. In those who discontinued upadacitinib, outcomes included duration of time using upadacitinib and reasons for discontinuation. Additional exploratory outcomes were to evaluate predictors of disease response to upadacitinib in patients with UC following loss of response to tofacitinib including demographics, previous advanced therapies used (number and type), and duration. Data extraction Two investigators (K.P. and A.A.) performed data extraction. Extracted data included patient demographics, IBD history, biochemical data, endoscopic data, and clinical response data. We collected data regarding the upadacitinib treatment regimen, clinical response, endoscopic response, histological response, and biochemical response. RB, SF, and PRO2 scores were retrospectively derived from clinic notes at baseline and at their last clinic follow-up appointment. Data analysis Data were summarized using descriptive and comparative analyses. Median and range were used to describe continuous variables. Categorical variables were reported as proportions. Missing data were handled with simple case-wise deletion for each analysis. Only patients with available data for a given outcome and time point were included in that specific analysis. Research ethics This study was reviewed and approved by the Hamilton Health Sciences research ethics board. Patient characteristics A total of 9 patients with UC with previous tofacitinib failure and subsequently treated with upadacitinib were included in this study. Eight of these patients were able to complete upadacitinib induction. Patient age when commencing upadacitinib ranged from 23 to 52 years with a median age of 35 years. Three patients (33%) were women. For disease extent, 3 patients (33%) had left-sided colitis with 6 patients (67%) having pancolitis. The median age at diagnosis was 29 years (18–46 years) with a median disease duration of 7 years (2–23 years). No patients had previous IBD-related surgeries. The 9 patients included had a previously failed median of 4 therapies (2–9) before upadacitinib commencement and failure of a median of 3 advanced therapies (1–5). One patient had previously trialed 1 advanced therapy, while 1 patient had trialed 2 advanced therapies, and the remaining 7 patients had been exposed to ≥2 advanced therapies. All patients had previously been treated with tofacitinib for a median duration of 12 months (4–57 months) with all patients having secondary loss of response to therapy. Upadacitinib dosing All patients were started on upadacitinib to treat their UC. Patients started on an induction dose of 45 mg once daily for 8 weeks followed by maintenance of 30 mg once daily. One patient required reinduction with upadacitinib 45 mg once daily after an initial response but loss of response while on maintenance dosing and was able recapture response after reinduction followed by a maintenance dose of 30 mg once daily. Three patients failed to proceed to maintenance therapy for different reasons: 1 had refractory disease requiring colectomy, 1 had refractory disease requiring transition to a clinical trial, and the last 1 had issues with insurance reimbursement and side effects. The median follow-up duration of patients who progressed to maintenance dosing was 15.5 months (4–34 months). Clinical response Out of the 8 patients that completed upadacitinib induction, 2 patients (25%) had a primary nonresponse, 2 patients (25%) had a secondary loss of response, and 4 patients (50%) were able to achieve steroid-free clinical remission. From the 2 of 8 patients that had primary nonresponse, 1 patient proceeded to colectomy while the other patient entered a clinical trial. The remaining 6 of 8 patients reported some clinical improvement after 8 weeks of upadacitinib induction. The most frequently reported symptomatic improvements were in SF, RB, urgency, nocturnal bowel movements, and abdominal pain. Patients' clinical response is outlined in Tables 1 and 2. Table 1.Upadacitinib treatment regimens and clinical, endoscopic, and histologic outcomes following previous tofacitinib failurePatientUpadacitinib treatment courseTime from upadacitinib commencement to last colonoscopy (mo)Clinical response—SF, RB, PRO2 scoreEndoscopic response—MESHistologic responseOutcome at the end of the study period18 wk 45 mg induction ->30 mg maintenance (11 mo) -> 45 mg (2 mo) -> 30 mg (10 mo)20SF: 0, RB: 0, PRO2: 0MES 0Colonic mucosa with focal chronic active inflammation. Histological improvementSteroid-free clinical remission. Following single repeat induction of upadacitinib28 wk 45 mg induction -> 30 mg (34 mo)5SF: 0, RB: 0, PRO2: 0MES 1Mild chronic colitis and detached fragments of granulation tissue in keeping with active inflammation. Histological improvementSteroid-free clinical remission38 wk 45 mg induction -> 30 mg (4 mo). CeasedN/ASF: 3, RB: 1, PRO2: 4——Secondary loss of response. Subsequently commenced on ozanimod with clinical response and then lost to follow-up445 mg induction (10 mo) -> ceased10SF: 3, RB: 1, PRO2: 4MES 3Ulceration with granulation tissue and reactive epithelial changes are identified. Histological worseningPrimary nonresponse. Proceeded to fecal transplant clinical trial545 mg induction -> ceased (8 wk), colectomyN/A———Primary nonresponse requiring colectomy68 wk 45 mg induction -> 30 mg (12 mo)8SF: 0, RB: 0, PRO2: 0MES 1Focal active proctitis. Histological improvementSteroid-free clinical remission75 wk 45 mg induction -> ceased (cost and fatigue)N/A———Ceased due to cost and fatigue. Then commenced on estranimod with primary nonresponse and now planned to commence mirikizumab88 wk 45 mg induction -> 30 mg (11 mo)11SF: 3, RB: 0, PRO2: 3MES 3 (although extent of inflammation improved)Benign colonic mucosa with severe acute colitis with focal mucosal ulceration and granulation tissue. Histological worseningSecondary loss of response. Escalated to combination upadacitinib and vedolizumab98 wk 45 mg induction -> 30 mg (17 mo)N/ASF: 0, RB: 0, PRO2: 0——Steroid-free clinical remissionMES, MAYO endoscopic score; N/A, not applicable; PRO2, patient reported outcome-2; RB, rectal bleeding; SF, stool frequency. Upadacitinib treatment regimens and clinical, endoscopic, and histologic outcomes following previous tofacitinib failure MES, MAYO endoscopic score; N/A, not applicable; PRO2, patient reported outcome-2; RB, rectal bleeding; SF, stool frequency. Table 2.Baseline demographic and disease characteristics, including clinical, endoscopic, histologic findings, and treatment historyPatientDemographics and disease distributionPreviously failed therapiesBaseline clinical activityEndoscopic baselineBaseline histologic activity127 M with E3 UCVedolizumab, tofacitinib, methotrexate, 5-ASASF: 2, RB:2, PRO2: 4MAYO 2 inflammation to the descending colonModerate chronic active colitis248 F with E2 UCUstekinumab, infliximab, vedolizumab, methotrexate, tofacitinibSF: 2, RB:0, PRO2: 2MAYO 3 inflammation 20–25 cmMarked chronic active proctitis with ulceration339 M with E2 UCInfliximab, vedolizumab, tofacitinib, ustekinumab, adalimumabSF: 2, RB: 1, PRO2:4——441 F with E3 UCTacrolimus, ustekinumab, tofacitinib, infliximab, vedolizumab, ozanimod, azathioprine, methotrexate, 5-ASASF: 3, RB: 2, PRO2: 4MAYO 2 inflammationDiffuse active chronic colitis with ulceration527 M with E3 UCUstekinumab, infliximab, tofacitinib, 5-ASASF: 3, RB:2, PRO2: 5——635 M with E3 UCTofacitinib, 5-ASASF: 2, RB: 3, PRO2: 5Mayo 2 proctosigmoiditisModerate chronic, focally active proctitis734 M with E2 UCInfliximab, tofacitinibSF: 3, RB:2, PRO2: 5——823 M with E3 UCVedolizumab, ustekinumab, adalimumab, infliximab, tofacitinib, azathioprineSF: 3, RB: 0, PRO2: 3Mayo 3 pancolitisModerate chronic colitis and hyperplastic change952 F with E3 UCInfliximab, adalimumab, tofacitinib, azathioprineSF: 3, RB: 1, PRO2: 4——Histological change was defined to be improved or worsening if there was improvement or worsening respectively in the severity, chronicity, or distribution of inflammation.5-ASA, 5-aminosalicylic acid; PRO2, patient reported outcome-2; RB, rectal bleeding; SF, stool frequency; UC, ulcerative colitis. Baseline demographic and disease characteristics, including clinical, endoscopic, histologic findings, and treatment history Histological change was defined to be improved or worsening if there was improvement or worsening respectively in the severity, chronicity, or distribution of inflammation. 5-ASA, 5-aminosalicylic acid; PRO2, patient reported outcome-2; RB, rectal bleeding; SF, stool frequency; UC, ulcerative colitis. Endoscopic response After initiation of upadacitinib, 5 patients underwent follow-up colonoscopy (Table 2). One of the 5 patients with follow-up colonoscopy had endoscopic evidence of worsening inflammation from an MES of 2–3. Another patient with follow-up colonoscopy had a persistent MES of 3 inflammation, although they had subjectively some interval improvement compared with their baseline colonoscopy. Notably, the remaining 3 of 5 (60%) patients with follow-up colonoscopy performed had endoscopic improvement (Figure 1). One of these patients achieved complete endoscopic remission, whereas the remaining 2 had a reduced severity of inflammation and a reduced distribution. The median time from upadacitinib commencement to final endoscopic assessment was 10 months (5–20 months). Endoscopic images of the rectal mucosa in patient 2. (A) Moderate-to-severe rectal inflammation during treatment with tofacitinib. (B) Mucosal healing with a residual inflammatory polyp after 2 years of upadacitinib. Histologic response All 5 patients who had follow-up colonoscopy also had biopsies collected. Histology samples showed a reduction in the pathological descriptions in the severity of the inflammation for 3 of 5 (60%) patients that underwent follow-up colonoscopy. In addition, 2 of 5 patients that underwent follow-up colonoscopy demonstrated worsening histological descriptions of the severity of their inflammation which correlated with their endoscopic response. Variable biopsy numbers and variable distribution of biopsies were taken throughout the colon. Adverse events From the total of 9 patients that were commenced on upadacitinib following tofacitinib failure, 1 reported fatigue in the setting of upadacitinib use which in conjunction with issues related to insurance coverage led to cessation of upadacitinib before completion of the induction period. Biochemical outcomes This study assessed biochemical outcomes, including C-reactive protein (CRP), fecal calprotectin, albumin, hemoglobin, and eosinophil count, at baseline, end of induction, and at 12 months after initiation of upadacitinib therapy (see Tables 3 and 4, Supplementary Appendix, http://links.lww.com/ACGCR/A45). All 9 patients had a baseline CRP, albumin, hemoglobin, and eosinophil count with 5 of 9 patients having a baseline fecal calprotectin. At the end of induction, among the 8 patients who completed 8 weeks of upadacitinib induction, 4 had CRP results, 5 had albumin, hemoglobin, and eosinophil results, and 3 had fecal calprotectin results. At 12 months, of 8 patients, 5 had available CRP level, albumin, hemoglobin, and eosinophils, and 1 had a fecal calprotectin result. Among the 4 patients who had steroid-free clinical remission by the end of the study, 1 had an elevated CRP at baseline, with 4 patients having normal CRP levels at 12 months. In contrast, among the 4 patients who completed induction but had primary or secondary nonresponse, 2 had increase in CRP level at 8 weeks, while 1 had a reduction in CRP level at 8 weeks which continued to improve by 12 months. For hemoglobin, 2 of 4 patients in steroid-free clinical remission had a low baseline hemoglobin, and 4 had a normal hemoglobin at 12 months. Among the 4 patients with primary or secondary nonresponse, 3 had a normal baseline hemoglobin and 1 had a low hemoglobin level. At 8 weeks, 1 of 4 still had a low hemoglobin level, though there was some interval improvement. Fecal calprotectin results were available in 5 of 9 patients at baseline, in 3 of 8 patients after induction, and 1 of 8 patients at 12 months. All baseline fecal calprotectin results which were available were elevated. Among the patients who achieved steroid-free clinical remission, 1 of 4 had had a fecal calprotectin result measured at 8 weeks which demonstrated a downtrend, and 1 had fecal calprotectin measured at 12 months which was normal. Among the 4 patients with primary or secondary nonresponse, 2 had follow-up with a fecal calprotectin at 8 weeks, which showed an uptrend in their fecal calprotectin level. All albumin and eosinophils results were normal where available. DISCUSSION UC is an inflammatory condition which develops as a consequence of immune dysregulation, microbial dysbiosis, and dietary exposures.1 The immune dysregulation is thought to occur due to yet to be determined polygenetic variations for proinflammatory genes.10 Although the specific signaling cascades involved are yet to be elucidated, there is emerging evidence of the roles of interleukin-6, Treg cells and T-helper cells as well as the nuclear factor kappa light chain enhancer of activated B cells (NF-κB), mitogen-activated protein kinase (MAPK), Notch, phosphoinositide 3-kinase/protein kinase B (PI3K/Akt), and JAK-STAT signaling cascade being involved in the development of the mucosal inflammation seen in UC.11–16 In the JAK/STAT pathway, noncovalent bonds between JAKs and cytokine receptors form with STATs coupling with tyrosine phosphorylation signaling pathways leading to downstream gene transcription.11 Variations in the JAKs and STATs activated and the subsequent changes in gene transcriptions lead to changes in cell apoptosis, proliferation, differentiation, and cell migrations.11 Ultimately, dysregulation in this pathway can lead to a proinflammatory environment and contribute to the inflammation seen in UC.11 Tofacitinib was the first JAK inhibitor available to treat UC and works by inhibiting the JAK1 and JAK2 receptors preventing the subsequent inflammatory cascade.17 Upadacitinib is a subsequent drug in this same class; however, it is selective in inhibiting JAK1.18 Previous meta-analyses which investigated the efficacy of upadacitinib after previous tofacitinib failure reviewed the induction period with limited data available on long-term outcomes.9 Furthermore, the studies available at the time of their meta-analysis potentially could have been at risk of selection bias due to their positive findings. More recently, a case series by Chatto et al presented 4 cases of patients with UC responding to upadacitinib during the induction phase.19 This study demonstrated that during the maintenance phase, all 4 cases developed flares which required reinduction, steroids, or combination with another advanced therapy.19 As such further long-term data on real-world outcomes in patients with UC treated with upadacitinib post tofacitinib failure are needed to determine its utility in this clinical scenario. Our results are in keeping with previously published data which reported a pooled clinical response rate of 57%.9 Our case series reports 50% of the patients were able to demonstrate long-term response to upadacitinib during the maintenance phase. This demonstrates that durable maintenance response may be lower than the induction response. The 50% durable response may also be resultant from disease refractoriness as ≥2 advanced therapies failed in 75% of patients in the case series. Nonetheless, 3 patients (75%) that did achieve steroid-free clinical remission had ≥2 advanced therapies that previously failed, which demonstrates that upadacitinib can still be effective in this cohort of patients with UC. This highlights the significance of the JAK/STAT signaling pathway, and particularly the role of the JAK1 receptor. In addition to its efficacy, upadacitinib after tofacitinib failure in patients with UC appears to be safe. One patient ceased upadacitinib during induction due to a combination of fatigue and issues related to insurance coverage. None of the patients had any serious adverse events, and no other patient reported any nonserious adverse events during their follow-up with upadacitinib otherwise being well tolerated. Our study includes follow-up of an average of 23.5 months in our patients that were maintained upadacitinib. This demonstrates the long-term durable response to upadacitinib in this patient cohort. While 50% of our patients that completed induction therapy were able to remain in remission during the maintenance phase, 1 patient required reinduction dosing as well to recapture remission, which was successful without subsequent reinduction courses. This further demonstrates challenges in the management of these patients that was not described in most previously published articles. Reassuringly, no patients required steroid therapies after commencing of upadacitinib. Biochemical outcomes correlate directly with clinical remission.20 Two of the most well-established clinical makers in UC are CRP and fecal calprotectin, which reflects underlying mucosal inflammation.20 In UC, a key marker in determining acute flare is an elevated CRP level which is also a prognostic indicator for remission.20 In the 4 patients who are in clinical remission, the CRP level had declined and achieved near-normalization at 1-year interval, which reflects clinical efficacy of upadacitinib. While fecal calprotectin is also a major component in diagnosing and monitoring UC, only a substantial decrease was reported in only 1 patient at 1-year post upadacitinib. This decrease in biochemical markers in the steroids-free clinical remission patients supports the efficacy of upadacitinib in maintaining remission in refractory cases suggesting that selective JAK1 inhibition with upadacitinib can suppress residual JAK-STAT signaling that escapes pan-JAK inhibition. These observations align with larger upadacitinib trials and real-world studies showing that early biochemical improvement (particularly CRP and fecal calprotectin within the first 8–12 weeks) is highly predictive of longer term mucosal healing and clinical remission in UC.5,18 Our study is limited by its design as a retrospective study at a single Canadian center with a relatively small sample size of 9 patients demonstrating an emerging clinical practice. This study was based on a retrospective clinical chart review where clinical documentation was used to inform the PRO2 scoring. Endoscopic and biochemical data were not available for all the patients included which demonstrates the differences in clinical practice among their treating clinicians. In addition, there were differences in the timing of repeat endoscopic assessments too. Our case series was also limited by a lack of an appropriate comparator cohort, such as an alternative advanced therapy, which limits our findings to a casual inference on relative benefit. Furthermore, although adherence of upadacitinib was reported by patients based on chart review, objective assessments with pill counting were not conducted with nonadherence potentially contributing to a lack of response in some of the patients. There may have also been an element of selection bias toward patients that were able to access upadacitinib. Finally, in most patients ≥2 advanced therapies had failed, and the failure of upadacitinib in 50% of our patients may potentially represent outcomes in patients with an inherently resistance disease phenotype. Future research should be directed toward prospective multicenter studies which would have standardized interval assessments with a long-term investigation period, which should hopefully allow for the recruitment of patients which have failed multiple lines and limited previous advanced therapies. In addition, as the underlying pathogenesis of UC is clarified, a clinical tool may be developed to help predict patient responsiveness to upadacitinib following tofacitinib under a personalized medical care model. Upadacitinib can induce and maintain steroid-free clinical remission and results in mucosal healing in some patients with UC who have previous tofacitinib failure. Of the patients in this cohort, 50% of patients were able to achieve durable remission for an average of 23.5 months. Upadacitinib use was steroid sparing with no serious side effects and 1 nonserious side effect. Larger, prospective studies with a larger sample size as well as standardized clinical and objective assessments are needed to confirm the clinical efficacy of upadacitinib following tofacitinib failure in patients with UC. DISCLOSURES Author contributions: K. Pratap and ABA AlBassam acquired, compiled, and interpreted the data and wrote the manuscript. N. Narula conceived and designed the concept, wrote the manuscript, and is the article guarantor. Financial disclosure: N. Narula holds a McMaster University Department of Medicine Internal Career Award. He has received honoraria from Janssen, Abbvie, Takeda, Pfizer, Merck, Sandoz, Novartis, and Ferring. No other authors have any relevant conflicts of interest. Informed consent was obtained for this case report. Supplementary Material