Beyond class switching in multi-refractory inflammatory bowel disease: Nine case reports and review of literature

Abstract

BACKGROUND

A clinically relevant subset of patients with inflammatory bowel disease (IBD) remain partial responders despite multiple advanced therapies. In this setting, reflexive class switching risks forfeiting incremental mechanistic gains, whereas a deliberate “build-on-partial-response” strategy - through dose optimization, extended induction, and, in selected cases, advanced combination therapy (ACT) [an oral Janus kinase inhibitor (iJAK) plus a biologic] - may consolidate disease control. However, real-world data supporting ACT remain limited, particularly regarding phenotype-driven selection, objective reassessment, and safety considerations.

CASE SUMMARY

We report nine multi-refractory IBD patients [ulcerative colitis (UC), n = 5; Crohn’s disease, n = 4] managed with ACT after incomplete response to an initial advanced therapy. UC strategies included infliximab plus tofacitinib, ustekinumab plus tofacitinib, and vedolizumab combined with iJAK. Crohn’s disease cases illustrate three patterns: Add-on anti-tumor necrosis factor therapy after partial iJAK response (e.g., upadacitinib plus adalimumab in two patients with stenosing or perianal phenotypes), interleukin-23 pathway intensification combined with iJAK (guselkumab plus tofacitinib), and phenotype discordance between extraintestinal and luminal disease control. Overall, 8/9 patients (89%) achieved clinical response with improvement in biochemical markers and/or endoscopic findings; among these, 6/9 (67%) achieved documented deep remission (clinical and biochemical, with endoscopic confirmation in 5/6). One patient with severe refractory UC failed ustekinumab plus tofacitinib and infliximab rescue and proceeded to colectomy with ileal pouch construction (11%). Patients did not present any adverse event associated with therapy, only secondary to uncontrolled disease.

CONCLUSION

In multi-refractory IBD, add-on ACT may convert partial response into deep remission, averting surgery in selected patients.

Key Words: Refractory inflammatory bowel disease; Ulcerative colitis; Crohn’s disease; Advanced combination therapy; Janus kinase inhibitors; Vedolizumab; Tofacitinib; Upadacitinib; Guselkumab; Case report

Core Tip: Multi-refractory inflammatory bowel disease often stalls at a “near-working” partial response to an advanced therapy. This case series (n = 9) suggests that, in carefully selected patients, consolidating partial mechanistic control with structured optimization and add-on advanced combination therapy (an oral Janus kinase inhibitor plus a biologic) may convert incomplete response into deep remission and delay surgery. This approach requires phenotype-driven selection, objective reassessment anchored to endoscopic and biochemical endpoints, rigorous infection exclusion, close monitoring, and predefined stopping rules - recognizing that timely colectomy remains appropriate when objective control fails.

INTRODUCTION

A meaningful subset of patients with inflammatory bowel disease (IBD) remain only partial responders to advanced therapy after multiple prior exposures. In this setting, discontinuing a “near-working” agent can risk losing incremental gains, while escalation within the same treatment trajectory - through optimization and, in carefully selected cases, advanced combined therapy (ACT) [an oral Janus kinase inhibitor (iJAK) plus a biologic) - may consolidate response[1]. Real-world data on iJAK-biologic combinations remain limited, particularly regarding phenotype-driven selection and structured safety monitoring[2].

Managing IBD after patients have exhausted multiple advanced therapies remains one of the hardest problems in everyday practice[3]. Randomized trial populations are highly selected, leaving clinicians to navigate refractory phenotypes with limited real-world guidance on dose, duration, sequencing, and what to do with the common scenario of a “near-working” partial response. In this setting, pragmatic strategies that build on partial mechanistic control - rather than reflexively switching class - are increasingly being used, including extended induction and carefully selected combination or sequential advanced therapy[1,3]. Tzouvala et al[4] provide important 52-week real-world data in ulcerative colitis (UC), showing that extended induction with tofacitinib can deliver meaningful remission and durable treatment persistence, while also highlighting clinical features that may shape dosing decisions over time. These observations align with prior studies supporting individualized placement of iJAK in multi-exposed UC[5-7] and complement feasibility/safety signals reported with selective combination or sequential advanced therapy in highly refractory disease[8,9].

A parallel challenge exists in Crohn’s disease (CD), where therapeutic sequencing has become progressively more complex as the armamentarium expands[10]. The practice-oriented framework by Imbrizi et al[11] emphasizes that phenotype, inflammatory burden, pharmacokinetics, safety, and access should guide not only first-line choice but also what we do when the first agent fails; within that schema, upadacitinib is positioned as a rational option after anti-tumor necrosis factor (TNF) exposure, particularly in colonic and perianal disease and in patients with high biologic clearance or hypoalbuminemia. Emerging data now add real-world texture to this positioning: A large prospective cohort has shown rapid clinical and biochemical responses to upadacitinib in both UC and CD, including patients who have cycled through multiple advanced therapies, with an acceptable early safety profile[12]. Phase 3 programme post hoc analyses further suggest benefit in fistulizing phenotypes, demonstrating higher rates of drainage resolution and external opening closure for perianal fistulas under upadacitinib vs placebo in both induction and maintenance[13]. Case-based reports are also filling gaps left by trial subgroups, describing rapid resolution of refractory perianal disease - even in complex contexts such as diverting stomas, concomitant vulvar involvement, and penetrating rectovaginal fistulas treated after multiple biologic failures[14,15]. At the extreme end, experimental approaches such as triple biologic therapy underscore how far clinicians sometimes push targeted agents when conventional sequencing is exhausted[16], reinforcing the broader point - also echoed in earlier work on Janus kinase (JAK) inhibitors as rescue options in treatment-resistant UC - that partial response in refractory IBD is often the starting point for rational escalation rather than a dead end[5,16].

For clinicians working in mixed public-private or resource-limited environments, these are not theoretical debates. When access is delayed or sequencing fails, the consequences are frequently surgical, with a real risk of postoperative infectious complications in malnourished, inflamed IBD patients[17]. Regional epidemiology reinforces the scale of the problem: Population-based data from Southern Brazil describe a relatively young IBD cohort with a high burden of extensive colitis and clinically significant CD, foreshadowing substantial future surgical and biological workload[18]. Our group has previously used a case-plus-systematic-review approach to illuminate rare Crohn’s phenotypes precisely because single patients can expose blind spots left by trials and guidelines[5].

Against this backdrop, we present a case series of multi-refractory IBD in which partial response to an advanced agent was deliberately consolidated - through extended induction and/or add-on ACT - to convert near-failures into deep remission and preserve options short of colectomy. In UC, this includes an extensive (E3) biologic-experienced patient who appeared headed toward surgery but recovered with a personalized extended induction plan: Short-term intensification of tofacitinib, step-down maintenance at a lower dose, and concurrent biologic optimization under tight biomarker and endoscopic surveillance, mirroring the persistence signal reported by Tzouvala et al[4] and illustrating how bedside decisions on dose, duration, and sequence can be rationalized in day-to-day care[11]. In CD, we highlight the complementary principle: In complex colonic/perianal phenotypes - where upadacitinib may offer rapid symptom control - objective reassessment can reveal residual luminal burden that warrants add-on biologic therapy rather than abandoning a “near-working” trajectory, consistent with the phenotype-driven sequencing logic emphasized by Imbrizi et al[11] and supported by emerging cohort, trial, and case-based data[16].

CASE PRESENTATION

Chief complaints

Case 1: Recurrent rectal bleeding, urgency, and increased stool frequency despite prior therapies.

Case 2: Persistent severe disease activity after failing vedolizumab, infliximab, and azathioprine.

Case 3: Abdominal pain, bloody diarrhea, and progressive perianal complications.

Case 4: Ongoing bleeding and diarrhea despite escalated infliximab.

Case 5: Recurrent acute abdominal pain episodes and sub-occlusive symptoms.

Case 6: Multi-refractory course having failed azathioprine, mesalazine, golimumab plus tofacitinib, and infliximab.

Case 7: Significant relapse (6-7 bowel motions/day with constant bleeding) despite ongoing vedolizumab every 4 weeks.

Case 8: Long-standing colonic CD presented with a multi-refractory course, persistently high inflammatory burden, and repeated endoscopic confirmation of severe disease.

Case 9: Recurrent bleeding and mucus despite vedolizumab plus tofacitinib.

History of present illness

Case 1: A 57-year-old man presented in March 2019 with persistent endoscopic activity despite intermittent clinical remission. In February 2019, colonoscopy showed ongoing inflammation (transverse colon Mayo 2). Over October 2019, he relapsed with rectal bleeding and mucus, requiring high-dose oral mesalazine and enteric budesonide; azathioprine was then introduced, leading to a temporary clinical stability.

In June 2020, he flared again. Colonoscopy confirmed extensive colitis (E3) with Mayo 2 activity and significant histologic inflammation. By October 2020, infliximab was initiated (induction) and continued as the core biologic. Despite partial improvement, urgency and incomplete symptom control persisted, prompting monthly dose optimization. Given the partial response, tofacitinib was added as adjunctive therapy in May 2021, designed to consolidate the “near-working” anti-TNF response, rather than class switching.

After approximately 3 months on combination therapy, the patient achieved clinical improvement (2 stools/day, no blood/pus, restored continence). Although residual endoscopic activity persisted initially (sigmoid Mayo 2), the clinical trajectory supported consolidation rather than abandonment, and he continued infliximab 500 mg every 4 weeks + tofacitinib 5 mg twice daily during this high-severity phase. By February 2022, he achieved biochemical quiescence (fecal calprotectin < 15) and sustained symptom control, allowing simplification back toward infliximab-centered maintenance. By February 2026, his maintenance regimen is monotherapy subcutaneous infliximab 240 mg every 2 weeks, under deep remission.

Case 2: A 42-year-old woman with a history of extensive UC (Montreal *E3) ongoing severe disease activity despite multiple prior therapies, culminating in add-on ACT to consolidate a partial response. By October 2022, she presented with ongoing severe disease activity (E2M3) after failing mesalazine, vedolizumab and infliximab, with azathioprine intolerance; infectious causes were excluded [negative cytomegalovirus (CMV) and Clostridioides testing], supporting true refractory inflammatory activity. She commenced a dual-target “build on partial response” strategy with golimumab (Simponi) and tofacitinib: Golimumab was initiated on October 2022, and tofacitinib was started soon after, initially 5 mg twice daily, then escalated to 10 mg twice daily (together with escalation of golimumab to 100 mg monthly by January 2023) in response to persistent nocturnal symptoms, ongoing bleeding, and fecal calprotectin > 3000 μg/g. Partial clinical response was achieved, with improvement in bleeding and mucus decreased, but persistent high stool frequency (8 bowel motions/day) and significant inflammatory burden. In April 2023, colonoscopy confirmed severe pancolitis despite optimized golimumab + tofacitinib (Figure 1).

Figure 1 Colonoscopy findings in case 2 under tofacitinib monotherapy (severely symptomatic). Ulcerative colitis with stricture, friability and ulcers, Mayo endoscopic score 2. A: Descending colon; B: Sigmoid colon.

With persistent inflammation and partially effective iJAK “near-working” signal, the approach was not to abandon the trajectory but to add a new mechanism: Ustekinumab induction was first administered in June 2023, while tofacitinib was maintained to prevent rebound. To maximize early control, tofacitinib was temporarily intensified to 10 mg three times daily for 30 days, then reduced back to 10 mg twice daily as extended induction. The response was rapid and clinically meaningful. Within 6-8 weeks, stool frequency fell to around 4/day, bleeding and mucus resolved, nocturnal urgency ceased, appetite returned, and she gained 5 kg. Biomarkers began to move in parallel, with fecal calprotectin falling to 1679 (from > 3000). She continued ustekinumab maintenance initially every 8 weeks, then, because colonoscopy on December 2023 still showed residual activity (E3M2) despite major clinical improvement (Figure 2), ustekinumab was escalated to monthly dosing, while tofacitinib was maintained at 10 mg twice daily, then later de-escalated to 5 mg twice daily after transaminase changes.

Figure 2 Colonoscopy findings in case 2 on maintenance therapy with ustekinumab every 8 weeks and tofacitinib 5 mg twice daily (mildly symptomatic). Improvement, but still presenting ulcerative colitis with patchy areas of stricture, friability and ulcers, Mayo endoscopic score 2. A: Descending colon; B: Sigmoid colon.

Over 2024-2025, she transitioned from improvement to deep remission: By April 2025, she was asymptomatic for six months with fecal calprotectin 11.2, stable blood counts, and normal inflammatory markers, on ustekinumab 90 mg monthly plus tofacitinib 5 mg twice daily. A surveillance colonoscopy on October 2025 documented E3M0 (cicatricial) with normal biopsies (Figure 3). Importantly, a patient who had exhausted multiple mechanisms and was approaching colectomy consideration achieved and sustained colectomy-free deep remission through a pragmatic strategy: Recognize a partial iJAK signal, avoid premature class switching, and add on ustekinumab with short-term intensified tofacitinib as a structured rescue bridge.

Figure 3 Colonoscopy findings in case 2 on maintenance therapy with ustekinumab every 4 weeks and tofacitinib 5 mg twice daily - two years into deep remission. Scarred mucosa, Mayo endoscopic score 0. A: Transverse colon; B: Sigmoid colon.

Case 3: A 28-year-old man presented with colonic CD with complex perianal involvement (Montreal A2 L2B2p), characterized by rapid symptomatic and perianal improvement on JAK inhibitor therapy, but persistent luminal inflammatory burden requiring add-on biologic therapy (Figure 4).

Figure 4 Sequential healing course of perianal Crohn’s disease under upadacitinib therapy (case 3). A: Baseline postoperative appearance with seton in place at treatment initiation, showing active inflammation and exposed fistulous tract; B: Day 30: Intraoperative view demonstrating tract exposure and targeted debridement, with early reduction in local inflammatory burden; C: Day 60: Ongoing cicatrization with visible contraction of the external opening and resolution of surrounding erythema; D: Day 90: Complete closure of the external opening with mature cicatricial tissue, reflecting full postoperative fistula healing.

By April 2024, he had marked systemic inflammation [C-reactive protein (CRP): 152 mg/L] and endoscopy on May 2024 showed deep longitudinal colonic ulcers consistent with Crohn’s colitis. At his first full review in our service on June 2024, he was evaluated with aggressive colonic and perianal disease. Screening and risk mitigation were pursued in parallel, including infectious work-up (negative Clostridioides difficile; subsequent stool/parasite results guided treatment), and surgical collaboration was established for the perianal component.

Because the immediate “standard” choice (anti-TNF) was not straightforward logistically, therapy was started with upadacitinib in September 2024, following budesonide bridging. He commenced upadacitinib 45 mg daily, with a planned step-down to 30 mg daily after a 16 weeks induction. The half-response signal was strong and early. By October 2024, diarrhea had ceased (1-2 formed stools/day), abdominal pain had resolved, and postoperative perianal symptoms had improved after proctologic intervention; upadacitinib was reduced to 30 mg daily with budesonide taper.

Over December 2024 to February 2025, despite sustained clinical remission, objective inflammation persisted (e.g., calprotectin > 800 in December 2024, then 680 in February 2025; CRP fell but remained elevated), and ongoing perianal disease still required repeat surgical management. Objective reassessment then clarified the key problem: Colonoscopy on June 2025 showed persistent luminal disease with an inflammatory hepatic-flexure stenosis and Simple Endoscopic Score for Crohn’s disease (SES-CD 8), indicating clinical response with incomplete endoscopic control - a partially effective iJAK that was “near working”, but not enough alone.

Rather than switching away from upadacitinib (and risking loss of the clear clinical/perianal gains), the strategy was to add on adalimumab to consolidate mechanistic control. The plan (July 2025) was to start adalimumab every 14 days and step down upadacitinib to 15 mg daily once adalimumab was in hand; he confirmed receipt of adalimumab in July 2025 and proceeded with the combined regimen. The result was conversion from partial to deep control. By 10 November 2025, he reported full symptomatic remission and weight gain, with resolution of the right-sided abdominal pain after starting adalimumab, while maintained on adalimumab every 14 days + upadacitinib 15 mg daily. On November 2025, colonoscopy documented minimal activity (SES-CD 1), and at follow-up on December 2025 he was assessed as having complete response to upadacitinib + adalimumab, remaining colectomy-free and clinically stable.

Case 4: A 28-year-old woman with colonic CD presented with persistent bloody diarrhea and systemic inflammation (CRP 45 mg/L, erythrocyte sedimentation rate 63 mm/hour) despite escalated infliximab. Colonoscopy during admission confirmed severe active colitis, and infectious causes were excluded (CMV and herpes simplex virus negative). Ustekinumab induction (390 mg intravenous) was initiated on February 2024, with a corticosteroid bridge at discharge. Partial clinical improvement was achieved; however, to consolidate disease control, tofacitinib 10 mg twice daily was added on April 2024, while maintaining ustekinumab (Stelara) in parallel, rather than switching mechanism again. Over the next months she improved clinically and biochemically, reaching a low-inflammation state by July 2024 (CRP < 2 mg/L) and fecal calprotectin of 27 μg/g, consistent with deep response on ustekinumab + tofacitinib.

During follow-up, an unplanned pregnancy required stopping of tofacitinib, while continuing ustekinumab 90 mg subcutaneous every 8 weeks with close laboratory surveillance during gestation. At the second semester, disease activity recurred, with symptom relapse and elevated calprotectin, requiring dose intensification of ustekinumab to 90 mg every 4 weeks.

After a successful delivery of a healthy baby, she later demonstrated biochemical relapse despite ongoing advanced therapy access constraints, and was transitioned to guselkumab 200 mg monthly; however, follow-up documented persistent active disease on Tremfya, prompting a decision to reintroduce the association of tofacitinib 10 mg every 12 hours (given prior proven response). At that point, breastfeeding counseling and the lack of robust long-term safety data were explicitly discussed in the record, reinforcing the real-world complexity of add-on iJAK strategies in women of childbearing age.

Case 5: A 42-year-old woman was referred in March 2025 after two acute abdominal pain episodes in October and December 2024, the latter requiring a 7-day admission with marked distension and endoscopic findings suggesting CD. In April 2025, upper endoscopy and colonoscopy demonstrated post-bypass anatomy and ulcerated, scarred colonic stenoses with chronic ileitis and colitis. During a flare on May 2025 with sub-occlusive symptoms and vomiting, the plan was to stop interleukin (IL)-17 therapy and bridge with advanced therapy. She commenced upadacitinib 45 mg once daily (induction) plus budesonide, with the intent to complete 4 months of 45 mg and then transition to 30 mg daily maintenance.

Initial clinical improvement was observed by August 2025, with minimal abdominal pain and resolution of arthritis symptoms; however, biochemistry suggested residual inflammatory burden (e.g., fecal calprotectin 678 μg/g). In September 2025, she relapsed with weight loss, nausea, and diarrhea despite ongoing therapy. By late October 2025, symptoms again improved and her joint pain was better on upadacitinib - an important determinant in not discontinuing iJAK.

Objective reassessment in November 2025 confirmed only a partial endoscopic response: Colonoscopy showed ileocolonic Crohn’s with a stenosing phenotype (Montreal A2 L3B2) and moderate activity (SES-CD 12); histology demonstrated chronic erosive colitis with CMV/herpes immunohistochemistry negative and fungal/acid-fast bacilli stains negative. Given this partial-response pattern - symptomatic improvement on upadacitinib with persisting endoscopic activity - the strategy shifted from class switching to add-on therapy: In late 2025, while on upadacitinib 30 mg daily, the team elected to add adalimumab induction and down-titrate upadacitinib to 15 mg daily as combination advanced therapy, with planned endoscopic reassessment after approximately 6 months. Notably, upadacitinib was intentionally maintained rather than stopped because it was controlling her inflammatory arthropathy (and prior adalimumab had not), while adalimumab was introduced to consolidate luminal control in the setting of persistent stenosing ileocolonic activity.

Case 6: A 21-year-old male with extensive UC, started upadacitinib 45 mg daily induction for 16 weeks on February 2025. Early response was partial: By April/2025, bleeding had reduced, but stool frequency remained markedly elevated (reported approximately 20/day), and he required ongoing bridging steroids/budesonide. By May 2025, he described some symptomatic improvement (around 10/day, less bleeding/urgency), with biochemical improvement.

Given this incomplete luminal control on JAK1 inhibition alone, vedolizumab was added as an “add-on” strategy rather than abandoning a partially effective agent on Aug 2025, continued alongside upadacitinib 30 mg daily. Over the following months, he transitioned to a clear clinical/biochemical response on upadacitinib + vedolizumab: By October 2025 fecal calprotectin was 60 μg/g with CRP < 2, and by November 2025 he reported 4 stools/day without blood, had stopped prednisone, with clinical and laboratory response to the combination (vedolizumab 300 mg every 8 weeks and updacitinib 30 mg daily). He escaped colectomy under this regimen.

Case 7: A 54-year-old woman with long-standing extensive UC. After a significant relapse (6-7 bowel motions/day with constant bleeding despite ongoing vedolizumab every 4 weeks), colonoscopy on August 2023 confirmed active extensive disease (E3M3 in the rectum). Tofacitinib was started the day after colonoscopy, with an early symptomatic improvement over 10 days (no diarrhea and no bleeding), suggesting an initial “signal”, but disease control proved non-durable.

By 2024, given persistent disease activity, she was escalated to ustekinumab monthly plus tofacitinib twice daily for at least six months, yet she continued with daily bleeding and high inflammatory burden (fecal calprotectin 3510 μg/g on December 2024). She had to be admitted to the hospital and a colonoscopy on December 2024 showed severe activity (E3M3 in the rectosigmoid), with CMV and herpes immunohistochemistry negative. This represented objective failure of the ustekinumab + tofacitinib combination.

A final rescue attempt with infliximab induction was made in early 2025 (“three doses”), without meaningful reversal (ongoing bleeding; calprotectin 1140 μg/g on February 2025). She proceeded to total colectomy with ileal pouch creation on March 2025, followed by ileostomy reversal on August 2025.

In the late postoperative period, she was subsequently followed for pouchitis and cuffitis (colonoscopy on December 2025 showing five aphthous ulcers in the ileum and cuffitis; histology: Erosive ileitis and chronic erosive colitis in the cuff with cryptitis and crypt microabscesses). Initial management was antibiotics and VSL#3 probiotics; if refractory, vedolizumab was considered as the next step as prior response to vedolizumab had lasted seven years.

Case 8: A 36-year-old woman with long-standing colonic CD followed an aggressive, multi-refractory course with persistently high inflammatory burden and repeated endoscopic confirmation of severe disease (colonoscopy March 2024 SES-CD 25; July 2024 SES-CD 32). Risankizumab was initiated in February 2025, achieving only a partial response, with persistent symptoms (approximately 10 bowel motions/day) and objective inflammation (calprotectin 1550 μg/g in April 2025; later 3500 μg/g in September 2025).

Given ongoing symptoms (approximately 10/day) and persistent biomarker elevation (> 800 μg/g), the plan shifted in November 2025 to guselkumab at intensified dosing (200 mg, maintenance every 4 weeks) combined with tofacitinib 10 mg twice a day, explicitly aiming to convert a partial biologic response into deeper disease control. After approximately six months of guselkumab + tofacitinib, she has achieved a good clinical response and is currently awaiting confirmatory colonoscopy to document endoscopic response/remission.

Case 9: A 30-year-old woman with extensive UC (pancolitis; E3 with severe flares historically E3M3) illustrates a “near-working” trajectory where upadacitinib was added/switch-in to consolidate control while keeping vedolizumab in play, ultimately achieving stability on upadacitinib + vedolizumab. Colonoscopy in April 2023 confirmed extensive severe disease (E3M3) and she was started on vedolizumab.

Because symptoms and objective activity persisted on vedolizumab-based therapy, tofacitinib was introduced as an adjunct in June 2023 (initially 5 mg twice a day, then escalated to 10 mg twice a day), alongside ongoing vedolizumab (later intensified to every 4 weeks) and topical/oral 5-aminosalicylic acid optimization. The key “partial response” signal was clear but incomplete: She could reach near-normal bowel habits at times, yet repeatedly relapsed when tofacitinib was interrupted - most notably in February 2024, when stopping tofacitinib was followed by recurrent bleeding/diarrhea within approximately 10-12 days, requiring prompt re-initiation and re-escalation of tofacitinib, with steroid bridging.

Objective reassessment showed improvement but persistent residual disease: Colonoscopy July 2024 documented E3M2 with intense histologic activity (CMV immunohistochemistry negative), and colonoscopy May 2025 again showed ongoing left-sided activity (Mayo 2) with quiescent/“cicatricial” segments elsewhere. Despite being “reasonably controlled” on vedolizumab every 4 weeks + tofacitinib 10 mg twice daily, she worsened again in mid-2025 with recurrent bleeding and mucus, and the decision was made to switch from tofacitinib to upadacitinib rather than abandon the vedolizumab backbone.

Upadacitinib 45 mg daily was stated on September 2025 (planned extended induction, then step-down), and within the first month she reported a striking clinical improvement: One formed stool/day, no blood or mucus, and improved abdominal pain, while continuing vedolizumab every 4 weeks and high-dose mesalazine. By October 2025, she was under clinical remission, and the working maintenance strategy became upadacitinib + vedolizumab with ongoing monitoring and objective reassessment planned.

History of past illness

Case 1: Long-standing UC diagnosed approximately in 2013, previously managed with mesalazine and azathioprine.

Case 2: The patient was diagnosed in 2013 with UC, initially limited to the rectum (E1), after a 30-day episode of rectal bleeding, responding to oral mesalazine over 3 months. She stopped therapy and relapsed in 2015, remaining in near-continuous flare for almost a year; mesalazine (oral and suppository) produced incomplete control. A third flare in 2018 again failed optimized mesalazine, and by January 2020 the disease had extended to the splenic flexure (E2M2). Biologic therapy was started with vedolizumab 300 mg in 2020, with an initial response followed by secondary loss of response. She then switched in 2021 to infliximab plus azathioprine; however, azathioprine was discontinued due to cytopenia and anemia requiring transfusion, and infliximab was ultimately stopped in August 2021. Objective pharmacokinetic failure was later documented: Infliximab level > 15 in May 2021, then < 0.4 with anti-drug antibodies positive by October 2021, consistent with immunogenic loss of response despite prior optimization attempts.

Case 3: He first developed abdominal pain and diarrhea around 2020, later evolving (by early 2024) into diarrhea with blood/mucus and progressive perianal complications following an anal fissure. Initial management elsewhere included mesalazine and repeated antibiotic courses, without durable control.

Case 4: Colonic CD, initially misclassified as UC, followed a refractory and steroid-exposed course. Prior therapies included mesalazine and azathioprine, followed by infliximab, which was initiated at 5 mg/kg every 8 weeks and later escalated to 5 mg/kg every 4 weeks due to persistent symptoms and high inflammatory burden (fecal calprotectin 3171 μg/g).

Case 5: Prior Roux-en-Y gastric bypass (2017) and cholecystectomy (2004). She had a history of psoriatic arthritis treated for approximately 2 years with ixekizumab (Taltz, IL-17) with prior inadequate response to adalimumab.

Case 6: Extensive UC (Montreal E3M3) evolved into a multi-refractory course, having failed azathioprine, oral mesalazine, and became steroid dependant. Prior advanced therapies included golimumab combined with tofacitinib and subsequent infliximab, without sustained response.

Case 7: Diagnosed with pancolitis in 2009, the patient developed a prolonged steroid-exposed course with repeated relapses despite “full therapy”. She had received prolonged high-dose mesalazine and vedolizumab (including monthly schedules during earlier phases), with intermittent systemic corticosteroids over several years.

Case 8: Long-standing colonic CD diagnosed in 2018 with an aggressive, multi-refractory course. Prior therapeutic history included failure to infliximab (with attempted dose increase, still without benefit), adalimumab, azathioprine, and ustekinumab, with ongoing diarrhea and bleeding. In late 2024, a short trial of upadacitinib 45 mg daily produced no meaningful clinical response after 4 weeks, leading to discontinuation.

Case 9: She was diagnosed with UC in 2015 and later required escalation to advanced therapy. She had previously been treated with infliximab in 2020-2021, which she stopped in 2022 after an intercurrent pneumonia; she reported being asymptomatic on infliximab, but relapsed again in early 2023 with diarrhea and bleeding despite azathioprine + high-dose mesalazine + corticosteroids, prompting re-staging.

Personal and family history

Cases 1, 2, 3, 5, 6, 7, 8, and 9: No relevant family history of IBD or colorectal malignancy reported.

Case 4: No relevant family history of IBD or colorectal malignancy reported. Pregnancy occurred during follow-up, requiring modification of ACT.

Physical examination

Case 1: Abdomen soft with mild left lower quadrant tenderness on deep palpation; no peritoneal signs.

Case 2: Abdomen distended with diffuse tenderness; no peritoneal signs.

Case 3: Perianal examination revealed an active fistulous tract with seton in place; abdominal tenderness in the right lower quadrant.

Case 4: Abdomen tender with active rectal bleeding noted at examination.

Case 5: Abdominal distension with right-sided tenderness; no peritoneal signs.

Case 6: Abdomen soft; no peritoneal signs at time of ACT initiation.

Case 7: Abdomen tender with active rectal bleeding and urgency.

Case 8: Abdomen tender; active diarrhea with blood and mucus.

Case 9: Abdomen soft; mild left lower quadrant tenderness; active rectal bleeding.

Laboratory examinations

Case 1: Inflammatory markers elevated at time of escalation; fecal calprotectin greater than 200 μg/g at flare; normalized to less than 15 μg/g by February 2022.

Case 2: Fecal calprotectin greater than 3000 μg/g at initiation of ACT; fell to 1679 μg/g after 6-8 weeks; normalized to 11.2 μg/g by April 2025. Transaminase elevation noted during tofacitinib dose of 10 mg twice daily, prompting de-escalation to 5 mg twice daily.

Case 3: CRP 152 mg/L at presentation; fecal calprotectin greater than 800 μg/g in December 2024, 680 μg/g in February 2025. Clostridioides difficile testing negative. Infectious workup including stool parasitology performed; Blastocystis hominis identified and treated prior to biologic initiation.

Case 4: CRP 45 mg/L, erythrocyte sedimentation rate 63 mm/hour, fecal calprotectin 3171 μg/g at admission; fell to less than 2 mg/L CRP and 27 μg/g calprotectin by July 2024 on ustekinumab plus tofacitinib.

Case 5: Fecal calprotectin 678 μg/g in August 2025 despite clinical response; CMV/herpes simplex virus immunohistochemistry negative; fungal and acid-fast bacillus stains negative on November 2025 biopsy.

Case 6: Fecal calprotectin 60 μg/g and CRP < 2 mg/L by October 2025 on vedolizumab plus upadacitinib combination.

Case 7: Fecal calprotectin 3510 μg/g (December 2024); 1140 μg/g after infliximab rescue (February 2025). CMV PCR and herpes simplex virus immunohistochemistry negative.

Case 8: Fecal calprotectin 1550 μg/g (April 2025) and 3500 μg/g (September 2025) on risankizumab monotherapy; greater than 800 μg/g at ACT initiation.

Case 9: Inflammatory markers and fecal calprotectin elevated during flares; biochemical normalization anticipated on upadacitinib plus vedolizumab with objective reassessment planned.

Imaging examinations

Case 1: Colonoscopy in June 2020 confirmed extensive colitis (E3) with Mayo endoscopic score (MES) 2. Follow-up colonoscopy in August 2021 showed residual moderate activity (sigmoid MES 2).

Case 2: Colonoscopy in April 2023 confirmed severe pancolitis (E3 MES 3) despite optimized golimumab plus tofacitinib. In December 2023, it showed residual activity (E3 MES 2) despite clinical improvement, prompting ustekinumab intensification to monthly dosing. Surveillance colonoscopy October 2025 documented E3 MES 0 (cicatricial) with normal biopsies.

Case 3: Colonoscopy May 2024: Deep longitudinal colonic ulcers consistent with Crohn’s colitis. Colonoscopy June 2025: Ongoing luminal disease with inflammatory hepatic-flexure stenosis, Simple Endoscopic Score for CD (SES-CD) 8. Colonoscopy November 2025: Minimal activity (SES-CD 1), confirming conversion from partial to deep endoscopic control.

Case 4: Colonoscopy during February 2024 admission: Severe active colitis (CMV/herpes simplex virus negative). Subsequent colonoscopies documented progressive biochemical and clinical improvement.

Case 5: Esophagogastroduodenoscopy/colonoscopy April 2025: Post-bypass anatomy with ulcerated, scarred colonic stenoses with chronic ileitis/colitis. Colonoscopy November 2025: Ileocolonic Crohn’s with stenosing phenotype (Montreal A2 L3B2), SES-CD 12, moderate activity.

Case 6: Endoscopic assessment confirmed E3 MES 3 at baseline. Biochemical normalization (fecal calprotectin 60 μg/g, CRP < 2 mg/L) by October 2025 without confirmatory colonoscopy yet performed.

Case 7: Colonoscopy August 2023: E3 MES 3. Colonoscopy December 2024: E3 MES 3 in rectosigmoid despite combination therapy. Post-colectomy colonoscopy December 2025: Five aphthous ulcers in ileum and cuffitis.

Case 8: Colonoscopy March 2024: SES-CD 25. Colonoscopy July 2024: SES-CD 32. Confirmatory colonoscopy to document endoscopic response/remission on guselkumab plus tofacitinib is pending.

Case 9: Colonoscopy July 2024: E3 MES 2 with intense histologic activity. Colonoscopy May 2025: Ongoing left-sided activity (MES 2) with quiescent/cicatricial segments elsewhere. Objective reassessment planned after upadacitinib induction.

FINAL DIAGNOSIS

All nine patients had established diagnoses of IBD, including UC in five cases and CD in four, confirmed by endoscopy, histology, and clinical criteria, in accordance with current standard guidelines. Within this case series, patients shared a common clinical scenario of multi-refractory disease with a documented partial response to at least one prior advanced therapy. This pattern served as the working clinical framework guiding management decisions and supporting escalation to advanced combination therapy (ACT).

TREATMENT

Treatment decisions followed a structured “build-on-partial-response” approach. True inflammatory activity was first confirmed through objective reassessment, including endoscopy, fecal calprotectin, and C-reactive protein. The partially effective agent was then optimized, either through dose adjustment or extended induction when appropriate.

Infectious triggers, such as Clostridioides difficile and CMV, as well as regionally relevant pathogens, were systematically excluded before and during escalation. Add-on ACT was introduced in patients with confirmed partial response and disease phenotype supporting combination therapy.

Also, stopping rules including reassessment was planned at predefined intervals, typically within 3 months to 6 months, with ongoing multidisciplinary discussion and surgical collaboration maintained throughout. Specific ACT combinations and dosing are detailed in the individual case presentation above.

OUTCOME AND FOLLOW-UP

The summary outcomes by case are presented in Table 1. Overall, 8 of 9 patients (89%) achieved clinical response, defined as symptomatic improvement and objective reduction in inflammatory activity or planned confirmatory endoscopic reassessment. Among these, 6 of 9 patients (67%) achieved documented deep remission, defined as clinical and biochemical remission with endoscopic confirmation; endoscopic documentation was available in 5 of these 6 cases. One patient (case 7; 11%) with severe refractory UC did not respond to combination therapy (ustekinumab plus tofacitinib) or infliximab rescue and ultimately required total colectomy with ileal pouch construction. This outcome represents the appropriate role of surgical management as the endpoint when adequate disease control cannot be achieved with medical therapy.

Table 1 Case series summary.

Case	Age	Sex	Disease	Classification	Failed/prior treatments	Final combined therapy	Response	Complications/key notes
1	57	M	UC	E3; Mayo 2 (residual Mayo 2, August, 2021)	Mesalazine; budesonide; azathioprine; infliximab escalated to monthly (partial) → add-on iJAK	IFX (IV 500 mg every 4 weeks during severe phase) + TOFA 5 mg twice a day (later simplified to SC IFX 240 mg every 2 weeks)	Yes (deep remission)	Partial response to IFX → add-on TOFA; no surgery
2	42	F	UC	E3 (prior E1 → E3); E2M3 (October, 2022) → E3M2 (December, 2023) → E3M0 (October, 2025)	Mesalazine (multiple incomplete); vedolizumab (secondary loss of response); IFX + AZA (AZA stopped cytopenia/anemia; IFX immunogenic failure ADA+); golimumab + TOFA (partial despite TOFA 10 twice a day + golimumab 100 monthly)	USK + TOFA (TOFA 10 mg three times daily for 30 days, then 10 mg twice a day → 5 mg twice a day; USK escalated to every 4 weeks)	Yes (deep remission)	Transaminase changes → TOFA de-escalation; colectomy avoided
3	28	M	CD	Montreal A2 L2B2p (colonic + perianal; inflammatory stricture)	Mesalazine; multiple antibiotics; upadacitinib 45 → 30 mg (partial: Perianal + symptoms improved but SES-CD 8, stenosis)	UPA + ADA (UPA 15 mg daily + ADA every 14 days)	Yes (deep response; SES-CD 1)	Perianal disease requiring procedures; hepatic flexure inflammatory stenosis
4	28	F	CD	Colonic Crohn’s	Mesalazine; azathioprine; IFX 5 mg/kg every 8 weeks → escalated every 4 weeks (no response; FC 3171; CRP/ESR high); ustekinumab IV induction; pregnancy required TOFA stop; later guselkumab	USK + TOFA (10 mg twice a day) (later stopped for pregnancy; later TOFA reintroduced with guselkumab)	Yes (initial deep response); later fluctuating	Pregnancy; relapse in 2nd trimester → USK every 4 weeks; postpartum relapse and switching
5	42	F	CD	Montreal A2 L3B2, stenosing; SES-CD 12	Prior IL-17 (ixekizumab) for PsA; prior ADA for arthritis ineffective; UPA 45 mg → 30 mg (partial clinical; FC 678; persistent SES-CD 12)	UPA + ADA (ADA induction + UPA 15 mg)	Partial/ongoing (awaiting reassessment)	Stenosing disease; sub-occlusive episodes; arthritis control drove UPA continuation
6	21	M	UC	E3M3	Azathioprine; mesalazine; steroid-dependent; golimumab + TOFA (failed); infliximab (failed); UPA induction (partial)	UPA + VDZ (UPA 30 mg + VDZ 300 mg every 8 weeks)	Yes (clinical + biochemical response)	Severe stool frequency early; steroid dependence; colectomy avoided
7	54	F	UC	E3M3 (rectosigmoid severe December, 2024)	Mesalazine high-dose; long-term VDZ; steroids; TOFA initial signal but non-durable; USK + TOFA failed; IFX rescue induction (3 doses) failed	USK + TOFA (failed) → IFX rescue (failed)	No (colectomy)	Colectomy + IPAA (March, 2025), reversal August, 2025; late pouchitis/cuffitis December, 2025
8	36	F	CD	Colonic Crohn’s; severe SES-CD 25 → 32	IFX (dose increase, no benefit); ADA; AZA; USK; UPA trial (no response); risankizumab partial; add-on TOFA; access constraints	Guselkumab + TOFA	Yes (clinical); colonoscopy pending	Multi-refractory; awaiting endoscopic confirmation
9	30	F	UC	E3 (E3M3 April, 2023; E3M2 July, 2024; Mayo 2 May, 2025)	IFX 2020-21 (stopped 2022 after pneumonia); AZA + mesalazine + steroids; VDZ partial; TOFA 5 mg → 10 mg twice a day + VDZ every 4 weeks partial; relapse when TOFA stopped	UPA + VDZ (UPA 30 mg maintenance; VDZ 300 mg every 8 weeks)	Yes (clinical); objective reassessment planned	Pneumonia led to IFX stop; repeated relapse on TOFA interruption; now stable on UPA + VDZ

UC: Ulcerative colitis; CD: Crohn’s disease; IFX: Infliximab; iJAK: Janus kinase inhibitor; TOFA: Tofacitinib; SC: Subcutaneous; AZA: Azathioprine; ADA: Adalimumab; USK: Ustekinumab; SES-CD: Simple Endoscopic Score for Crohn’s disease; IL: Interleukin; UPA: Upadacitinib; PsA: Psoriatic arthritis; FC: Fecal calprotectin; VDZ: Vedolizumab; IPAA: Ileal pouch-anal anastomosis.

DISCUSSION

This case series reflects real-world clinical practice from a mixed public-private IBD population in Southern Brazil. Patients were included if they had a confirmed diagnosis of UC or CD based on endoscopic, histologic, and clinical criteria, and had shown a partial response to at least one prior advanced therapy. Partial response was defined as objective improvement in at least one domain - symptoms, biomarkers, or endoscopic findings - without achieving full clinical or biochemical remission. All included patients subsequently received ACT (an oral JAK inhibitor plus a biologic) as a deliberate escalation strategy, rather than switching therapeutic class. Before initiation of ACT, active inflammatory disease was confirmed through objective assessment, including endoscopy and/or fecal calprotectin and C-reactive protein, and infectious causes such as Clostridioides difficile, CMV, and regionally relevant pathogens were excluded. Outcomes were defined at the time of clinical decision-making as follows: Clinical response, corresponding to symptomatic improvement with objective reduction in inflammatory markers; deep remission, defined as clinical and biochemical remission with endoscopic confirmation; and treatment failure, defined as the need for colectomy or absence of objective improvement at planned reassessment, typically within 3 months to 6 months. Given the descriptive nature of this case series, no formal statistical analysis was performed, and outcomes are presented using exact proportions.

Across all nine cases, the dominant signal is that partial response was repeatedly leveraged as a platform rather than treated as failure, with escalation decisions anchored to objective reassessment. In case 1 (UC), infliximab produced an early bleeding signal but left persistent urgency and endoscopic activity despite optimization, prompting add-on tofacitinib and subsequent conversion to biochemical quiescence and long-term deep remission on simplified maintenance. Case 2 (UC) illustrates the “multi-exposed, everything-failed” phenotype where a half-response to an iJAK plus anti-TNF (golimumab + tofacitinib) still left severe pancolitis; instead of discarding the partial iJAK signal, ustekinumab was layered in while tofacitinib was temporarily intensified as an extended induction bridge, ultimately achieving sustained deep remission with endoscopic normalization. The same logic drove the two gut-selective combinations in severe UC: Case 6 converted an early, incomplete upadacitinib response into clear clinical/biochemical control after vedolizumab add-on (with steroid withdrawal and colectomy avoidance), while case 9 demonstrates a long “near-working” vedolizumab backbone that required an iJAK partner - first tofacitinib with frequent relapse on interruption, then a switch to upadacitinib while maintaining vedolizumab - leading to rapid symptom resolution and ongoing consolidation.

In CD (cases 3-5 and 8), the trigger for add-on was often phenotype discordance: Rapid symptomatic or extraintestinal improvement contrasted with residual luminal inflammatory burden on endoscopy. Case 3 (A2 L2B2p) is the clearest example: Upadacitinib delivered early symptom and perianal gains, yet persistent objective luminal disease (inflammatory stenosis; SES-CD 8) prompted adalimumab add-on rather than abandoning the iJAK trajectory; the result was conversion to minimal activity (SES-CD 1) and durable clinical remission. Case 5 similarly combined adalimumab with continued upadacitinib in a stenosing ileocolonic phenotype (A2 L3B2) where arthritis control and prior inadequate joint response to adalimumab argued for preserving iJAK while adding anti-TNF primarily to consolidate mucosal control; this case is intentionally positioned as “partial response with planned reassessment”, reflecting a pre-defined endpoint rather than open-ended combination. Case 8 represents the extreme multi-refractory end of the spectrum (failed anti-TNF, thiopurine, ustekinumab, non-response to a short upadacitinib trial, and only partial benefit on risankizumab), where guselkumab plus tofacitinib was used explicitly to deepen a partial IL-23-axis response; she now has a good clinical response and is awaiting confirmatory colonoscopy, emphasizing that symptom control alone is insufficient in refractory disease.

The single clear failure (case 7) is clinically important because it defines the boundary of this strategy. Despite an early symptomatic “signal” to tofacitinib and subsequent escalation to ustekinumab plus tofacitinib, case 7 had persistent severe objective inflammation and then failed infliximab rescue, ultimately requiring colectomy with pouch formation and later pouchitis/cuffitis - highlighting that combination therapy cannot replace timely surgical transition when objective markers and endoscopy remain unfavorable. Case 4 further refines the limits: Although ustekinumab plus tofacitinib produced deep biochemical response, pregnancy forced iJAK cessation and later required dose intensification and subsequent biologic switching, illustrating how life-stage constraints can fragment otherwise effective pathways and necessitate pragmatic re-introduction of previously effective combinations. Taken together, cases 1-9 support a practical framework after “advanced therapy exhaustion”: Identify a measurable partial response, optimize and reassess objectively, add on deliberately when phenotype or kinetics justify it, de-escalate when control is achieved, and maintain explicit stopping rules - because the endpoint is deep remission (or timely surgery), not combination for its own sake.

Two practical safeguards deserve emphasis before ACT is interpreted as a “default” solution in multi-refractory IBD. First, competing infectious drivers must be actively excluded before and during escalation, with targeted evaluation for Clostridioides difficile and CMV, tuberculosis where relevant, and a low threshold to re-investigate when the course deviates from the expected trajectory; in real-world practice this also includes regionally prevalent pathogens that can mimic or amplify colitis activity (e.g., intestinal paracoccidioidomycosis in endemic areas) and protozoa such as Blastocystis hominis, as well as systematic assessment for perianal sepsis in fistulizing CD[17-25]. Second, early surgical partnership is essential: Even when medical therapy is “working”, refractory UC can deteriorate abruptly and require urgent colectomy, and perianal CD can progress to abscesses and complex tracts that demand timely operative input; postoperative infectious risk is tightly linked to nutritional status, albumin, and inflammatory burden[17,18]. A related, pragmatic lesson across our series is disciplined avoidance of prolonged systemic corticosteroids. Steroids mask symptoms without addressing fistulizing biology, delay definitive therapy, worsen infection risk, and impair healing in phenotypes already primed for sepsis and postoperative complications - making them a detour into harm rather than a reliable bridge, particularly in perianal CD[21,26-29].

Mounting real-world and clinical data strengthen the rationale for a temporally flexible, data-guided use of small molecules and add-on strategies in advanced IBD. In UC, beyond standard 8-week induction, multiple series demonstrate that extending tofacitinib induction to 16 weeks can convert partial responders into sustained remitters, with colectomy-free survival exceeding 75% at six months in some cohorts, mirroring the “extended induction” logic embedded in the UC cases presented[4]. Evidence is also emerging for dual-target or sequential approaches: Small cohorts describe combining tofacitinib with agents such as vedolizumab or ustekinumab with acceptable safety signals and meaningful remission rates in highly refractory disease[22,24]. A multicenter Chinese case series reported clinical response in > 85% of refractory IBD patients after three months of dual-target therapy, with complete steroid withdrawal among those who started on corticosteroids[23]. Systematic reviews similarly suggest that carefully selected combinations - anti-TNF plus tofacitinib, or ustekinumab plus vedolizumab - can achieve endoscopic and biochemical remission in a substantial proportion of previously non-responsive patients[25]. In CD, the broader JAK inhibitor story reinforces the same principle: Although early phase II tofacitinib programs did not meet primary endpoints, dose-related biomarker signals and later real-world cohorts - including off-label tofacitinib as monotherapy or in combination - suggest meaningful benefit in selected refractory phenotypes, including stricturing colonic disease, supporting the plausibility of iJAK add-on strategies when the biology is “near controlled”[26,30,31]. Upadacitinib has since converted that class-level signal into more robust evidence, with prospective real-world data showing rapid clinical/biochemical responses in UC and CD[12], phase 3 data demonstrating improvements in luminal outcomes, quality of life and work productivity through 52 weeks[32], and post hoc fistulizing analyses showing higher rates of perianal drainage resolution and external opening closure vs placebo in both induction and maintenance[13]. Comparative/contextual data in UC further suggest that potency and kinetics may matter in high-burden phenotypes, with propensity-matched real-world analyses indicating lower 12-month risks of steroid-requiring hospitalization and colectomy with upadacitinib vs tofacitinib, and multicenter acute severe UC data suggesting a faster early response signal with upadacitinib in the first week[33-36]. Taken together, these lines of evidence support a shift away from rigid step-up paradigms toward adaptive decision-making - also reflected in bridge strategies (e.g., calcineurin inhibitors plus ustekinumab) and dual-biologic rescue for extraintestinal manifestations - provided monitoring is precise and escalation is disciplined rather than indiscriminate[37,38].

In that sense, the present series should be read less as advocacy for “more drugs”, and more as a pragmatic framework for what to do with partial response in advanced IBD: Confirm true inflammatory activity, optimize what is already working, and add on deliberately when phenotype and kinetics justify it - while maintaining explicit stopping rules and early surgical collaboration. This is not a call to abandon established algorithms; anti-TNF therapy remains first-line for many fistulizing phenotypes where feasible, and long-term experience still matters[11,13]. Rather, our data echo the more nuanced message emerging from sequencing frameworks and real-world constraints: In selected patients - young, high-burden, colonic-perianal phenotypes[20], or those limited by access, reimbursement or service organization - oral, rapidly acting iJAK strategies may be intentionally positioned earlier or used as add-on consolidation when the alternative is prolonged steroids, repeated admissions, or premature surgery[11,39]. This perspective aligns with national Brazilian registry data linking younger age, colonic disease and perianal involvement to Crohn’s severity, and with physician surveys emphasizing persistent regional barriers to diagnostics and therapeutic delivery[40,41]. It also mirrors updated consensus recommendations emphasizing early identification of aggressive phenotypes and phenotype-tailored sequencing in colonic-perianal CD[42,43]. The overarching conclusion is therefore practical: With rigorous infection exclusion, careful cardiovascular/thrombotic risk assessment, and multidisciplinary follow-up, extended induction and combination/sequential strategies can function as pragmatic tools to convert near-failures into durable remission and preserve the option space short of colectomy - while recognizing, as our failure case underscores, that timely surgery remains the correct endpoint when objective disease control cannot be achieved[42,43].

Beyond observational series, two prospective datasets now give extra support to the specific “gut-selective biologic + iJAK” concept highlighted in our UC cases. First, the VETO study (a prospective cohort nested within a randomised trial) specifically enrolled patients with moderate-to-severe UC who were refractory to anti-TNF therapy and had failed a second-line advanced agent; among non-responders to vedolizumab or tofacitinib monotherapy, escalation to vedolizumab + tofacitinib achieved high rates of response and corticosteroid-free remission by week 24, with endoscopic remission in a meaningful subset and no severe safety signals reported, supporting ACT as a rational “salvage consolidation” strategy when monotherapy stalls[35]. Second, a multicentre randomised open-label trial presented at European Crohn’s and Colitis Organization tested a “hit-hard-and-early” induction strategy with vedolizumab + upadacitinib (45 mg daily for 8 weeks) vs vedolizumab alone in moderate-to-severe UC, demonstrating superior week-8 endoscopic remission (MES = 0) and clinical remission with comparable short-term adverse event rates, reinforcing that combining a rapidly acting iJAK with a gut-selective biologic can overcome the monotherapy efficacy ceiling in selected patients[36].

DUET-CD and DUET-UC, presented at Digestive Disease Week 2026, further support dual advanced therapy in highly refractory IBD, showing that combined TNF-α/IL-23 blockade with JNJ-78934804, a golimumab + guselkumab-based co-antibody, produced dose-dependent benefits in patients with CD and UC who had failed ≥ 2 systemic therapy mechanisms[44,45]. In DUET-CD, high-dose JNJ-4804 achieved higher week-48 clinical remission than golimumab or guselkumab alone in this subgroup (49.2% vs 23.1% and 27.3%) and improved endoscopic response (33.3% vs 12.3% and 20.0%). In DUET-UC, the same high-dose strategy improved clinical remission (26.7% vs 2.1% with golimumab and 16.0% with guselkumab) and endoscopic improvement (35.6% vs 2.1% and 20.0%), reinforcing the biologic plausibility of pathway-complementary escalation after repeated mechanism failure[44,45]. Still, dual advanced therapy should remain reserved for selected refractory phenotypes: Mild-to-moderate UC often remains controlled with conventional strategies, including aminosalicylates[46], while emerging real-world experience with anti-TNF plus JAK inhibitor combinations supports individualized use in difficult-to-manage luminal, perianal, or extraintestinal disease[47]. Before escalation, clinicians must also exclude mimickers or superimposed infections, including recurrent Clostridioides difficile, CMV, fungal/parasitic disease, and bacterial overgrowth syndromes[48-50].

This report has important limitations inherent to its design. As a small, single-region case series without a comparator group, it cannot quantify relative efficacy or safety, establish causality, or define an optimal sequencing algorithm; treatment decisions were individualized and not guided by a predefined protocol - the absence of a standardised selection and escalation algorithm is an acknowledged limitation that future prospective studies should address - and combinations were used off-label in highly selected, multi-refractory phenotypes, introducing strong selection bias. The cases are heterogeneous (UC and CD; luminal, perianal, and extraintestinal drivers; differing prior exposures, induction durations, and de-escalation strategies), and objective endpoints were not uniform across all patients, with at least one case still awaiting confirmatory endoscopy. Follow-up duration also varied, and rare adverse events (serious infection, thromboembolism, malignancy) cannot be meaningfully assessed in a sample of this size. Finally, access constraints and life-stage factors (e.g., pregnancy) influenced therapy choice and continuity, which may limit generalizability to settings with different reimbursement structures and monitoring capacity.

CONCLUSION

In conclusion, this case series reinforces that personalized, temporally flexible use of advanced therapies can meaningfully redirect the course of multi-refractory IBD when standard algorithms stall. In UC, extended induction with tofacitinib - guided by biomarkers, supported by vigilant infection screening, and paired with rational biologic optimization or add-on dual targeting when needed - can convert partial response into durable remission and colectomy avoidance, providing a bedside parallel to the population-level persistence/remission signals. In CD, the evolving JAK literature and phase 3 fistulizing data suggest that upadacitinib can achieve rapid luminal improvement and meaningful perianal responses in selected high-burden phenotypes, and that - when partial response leaves residual endoscopic disease - add-on consolidation can be more rational than abandoning a “near-working” trajectory. Future prospective studies should refine optimal timing, dosing, risk stratification and sequencing of small molecules and combinations, but the message is already clear: In carefully selected patients, with meticulous infection exclusion, disciplined steroid avoidance, and early surgical collaboration, combination and extended-induction strategies deserve consideration before the colectomy table or irreversible surgical escalation.