IBD-associated SpA · Joints, Spine & Gut
When Gut and Joints Respond Differently
Evidence reviewed as of October 5, 2026
On this page
- Start with separate assessments
- Four combinations of activity
- Name the failure before changing treatment
- Reactive TDM: answer a specific intestinal question
- Proactive monitoring is a separate question
- Combination therapy: explain what the second drug is doing
- After the first TNF inhibitor: sequence by reason and phenotype
- New joint symptoms after TNFi → vedolizumab
- Diarrhoea or possible new IBD during IL-17 inhibition
- Paradoxical psoriasis and lupus-like syndromes during TNFi
- When one targeted agent cannot control both domains
- Finish with a shared, testable plan
- Sources and evidence notes
A treatment can control intestinal inflammation while leaving spondyloarthritis active, or control the joints while bowel disease progresses. In IBD-associated spondyloarthritis—also called enteropathic arthritis—the next decision therefore requires two assessments and one coordinated plan. “Treatment failure” is incomplete until the clinician names the affected domain, establishes whether inflammation persists, and identifies what the current drug was expected to treat.
This chapter concerns adult management. Its decision framework combines IBD guidance, axSpA guidance and interdisciplinary consensus; it is not a validated IBD-SpA treatment algorithm. Direct trials comparing complete gut–joint strategies remain limited. Patient priorities, previous complications and the consequences of losing an effective treatment belong in every decision.1
Start with separate assessments
For the bowel, distinguish symptomatic improvement from inflammatory control. Record stool frequency, bleeding, nocturnal symptoms, weight trajectory and abdominal or perianal symptoms alongside recent objective findings. In Crohn’s disease, the location and behaviour of disease determine whether endoscopy, intestinal ultrasound, enterography or another investigation is needed. A normal CRP alone cannot establish remission. Discordant symptoms and biomarkers may require endoscopic confirmation before changing treatment; the AGA biomarker guideline specifically discourages empiric escalation solely because biomarkers are elevated during symptomatic remission.2
In UC, stool markers and symptoms should also be interpreted together. Persistent symptoms with reassuring markers, or elevated markers despite apparent clinical remission, create a question to investigate rather than a complete treatment decision. Recent infection, changes in medication and the timing of the sample affect interpretation. Avoid a single composite statement such as “IBD is stable” when the last endoscopic assessment is old or current evidence is contradictory.3
For the musculoskeletal system, identify the actual phenotype. Separate axial symptoms, swollen peripheral joints, enthesitis and dactylitis from tenderness without demonstrable inflammation. Describe function and the examination, not simply a global pain score. Reconsider mechanical disease, structural damage, osteonecrosis, fracture and nociplastic pain when appropriate. ASAS–EULAR recommends reassessing diagnosis and comorbidities when treatment fails. MRI or ultrasound can answer a focused unresolved question; neither a historical positive scan nor nonspecific imaging abnormalities prove that today’s pain requires more immunosuppression.4
CRP does not identify the responsible organ. A useful consultation note states the evidence for intestinal inflammation and the evidence for musculoskeletal inflammation separately. Fatigue, sleep disruption, anaemia and gastrointestinal symptoms can influence patient-reported disease activity. A higher score merits attention, but the proposed treatment should address the mechanism generating it.
Four combinations of activity
The following matrix is an organising aid. “Active” refers to clinically established inflammatory disease, not symptoms alone; “controlled” should specify what evidence supports that judgement. The approach is consistent with the domain-based coordination proposed by the IG-IBD/SIR consensus.1
| Bowel | Joints or spine | Decision to resolve |
|---|---|---|
| Active | Active | Can an optimised shared treatment control both, or has one mechanism genuinely failed? |
| Controlled | Active | Is inflammatory musculoskeletal disease undertreated, and how can intestinal control be preserved? |
| Active | Controlled | Can bowel treatment be improved without withdrawing the treatment protecting joints or spine? |
| Controlled | Controlled | Can the effective regimen be sustained safely, and what explains any remaining symptoms? |
Both domains active
First establish severity and exclude complications requiring hospital, surgical or infection management. For an ambulatory patient, review treatment exposure before declaring a mechanistic failure. A shared treatment is attractive only when the particular molecule has evidence for the patient’s intestinal disease and musculoskeletal phenotype. Monoclonal TNF inhibition has a central role in IBD-associated axial disease; etanercept cannot be substituted on the assumption that all TNF inhibitors treat IBD. ECCO does not recommend vedolizumab or ustekinumab as treatment for IBD-associated axial inflammation. Peripheral synovitis requires a different discussion from axial disease.6
Bowel controlled, joints or spine active
Do not invalidate persistent inflammatory symptoms because intestinal treatment worked. Conversely, establish inflammation before abandoning a successful bowel regimen. Limited peripheral synovitis may permit a local intervention or selected conventional DMARD while preserving intestinal treatment; this is not a substitute for effective therapy of active axial disease. ECCO supports TNF antagonists for non-axial SpA and notes supportive, more limited evidence for sulfasalazine, methotrexate and ustekinumab. The options depend on phenotype and severity.6
Bowel active, joints or spine controlled
Ask what is protecting the musculoskeletal domain before replacing it. A bowel-effective alternative may have little or no axial efficacy. For active IBD with axSpA, BSR favours a suitable monoclonal TNF inhibitor or JAK inhibitor and advises against starting IL-17 inhibition. A history of inactive IBD is a different situation, requiring gastroenterology review rather than an automatic class-wide prohibition. Drug choice still depends on the individual molecule, prior treatment and safety profile.7
Both domains controlled, symptoms remain
Persistent diarrhoea may reflect bile acid malabsorption, a disorder of gut–brain interaction, bacterial overgrowth or another non-inflammatory mechanism. The AGA functional-symptom update recommends a stepwise evaluation before assigning functional symptoms. Treating a demonstrated alternative cause is active care; escalating immunosuppression without an inflammatory target is unlikely to resolve that mechanism. Assess musculoskeletal pain with the same discipline while continuing appropriate disease surveillance.8
Name the failure before changing treatment
Primary non-response means an adequate initial treatment course never produced the intended response. Secondary loss of response means a meaningful response occurred and subsequently deteriorated. Distinguish both from intolerance, an access-related interruption, incomplete adherence and a new unrelated diagnosis. Record this history separately for the bowel and the musculoskeletal system: infliximab may be a primary bowel failure but an effective axial treatment.
“Adequate treatment” includes the relevant induction regimen, maintenance schedule and time to assess response for that indication. Reconstruct actual administration dates, missed injections, infusion delays, dose changes, steroid exposure and prior objective response. Pharmacy access problems should be corrected without describing the patient as pharmacologically refractory. Ongoing severe disease or complications may preclude waiting through a routine reassessment interval.
The PANTS prospective Crohn’s cohort supports the importance of early drug exposure and immunogenicity in later anti-TNF outcomes. Low exposure may reflect clearance and antibody formation rather than failure of TNF as a therapeutic target. These are observational intestinal data, not a validated serum target for axial inflammation. They explain why “the drug did not work” should lead to a mechanism-focused review rather than automatic serial switching.9
Reactive TDM: answer a specific intestinal question
In adults with objectively active IBD receiving an anti-TNF agent, AGA suggests reactive therapeutic drug monitoring: drug concentration, with anti-drug antibodies when informative, used to guide a treatment change. The question is whether inflammatory failure is associated with insufficient exposure or persists despite adequate exposure. Testing can be relevant even when active bowel disease produces few symptoms.10, 11
| Pattern | Interpretation to consider | Potential response |
|---|---|---|
| Low drug; absent or low antibodies | Insufficient exposure, including non-immune clearance or missed treatment | Correct administration problems; consider dose or interval optimisation. |
| Low or undetectable drug; clinically significant antibodies | Immune-mediated pharmacokinetic failure | Consider another agent; selected low-titre situations may permit optimisation or an immunomodulator. |
| Adequate exposure; persistent objective inflammation | Possible pharmacodynamic or mechanistic failure | Reassess diagnosis and complications; consider a different mechanism that addresses the required domains. |
| Symptoms without established inflammation | No demonstrated inflammatory target for escalation | Clarify the cause of symptoms before treating a concentration result. |
These are clinical interpretations, not automatic laboratory commands. Antibody thresholds and concentration targets vary with assay, drug, timing, phenotype and intended outcome. Obtain the sample at an interpretable time, usually near the next dose for trough-based strategies, and document the last dose. A drug-sensitive antibody assay may behave differently when circulating drug is present. Repeat an unexpected or borderline result when clinically appropriate.11, 12
Do not transfer an IBD trough target into an axSpA treatment rule. EULAR permits consideration of reactive TDM in selected rheumatic-disease situations but does not endorse routine proactive testing.12 The 2026 ACR axSpA summary conditionally recommends against management of active axSpA based on serum drug levels or anti-drug antibodies. This differs from AGA’s recommendation for active IBD and should remain visible when the same patient has both conditions.13
Proactive monitoring is a separate question
Routine scheduled testing during remission should not be equated with reactive testing during confirmed loss of response. The 2017 AGA guideline made no recommendation for routine proactive anti-TNF TDM in quiescent IBD; that was a knowledge gap, not proof of no value.10 Subsequent NOR-DRUM B randomised evidence found better sustained disease control with proactive monitoring during maintenance infliximab across several immune-mediated diseases. Its findings cannot establish the same benefit for every biologic, induction setting or patient with combined IBD-SpA.14
Discuss proactive testing within the relevant specialist protocol, particularly when results could change a defined decision. State who will interpret the assay, what action is contemplated and which outcome is being protected. An isolated “low” concentration in a clinically stable patient is not, by itself, proof that effective treatment must be abandoned.
Combination therapy: explain what the second drug is doing
Adding an immunomodulator may serve intestinal efficacy or reduce anti-TNF immunogenicity. That rationale is distinct from expecting methotrexate or a thiopurine to treat axial inflammation. The PANTS cohort associated concomitant immunomodulator use with lower risk of antibodies linked to undetectable drug concentrations; it does not prove that all combinations are appropriate or equally beneficial.9
Recommendations differ by disease and molecule. AGA’s 2025 Crohn’s guideline suggests infliximab plus a thiopurine rather than infliximab alone in thiopurine-naïve patients starting infliximab. It makes no recommendation favouring adalimumab plus a thiopurine or methotrexate over adalimumab alone. AGA’s UC living guideline supports combining infliximab, adalimumab or golimumab with an immunomodulator over corresponding monotherapy. Neither statement justifies adding an immunomodulator to every non-TNF biologic.15, 16
Document the expected benefit, infection and malignancy considerations, marrow and liver monitoring, reproductive implications and intended duration. Reassess the combination after sustained remission. In SPARE, selected adults with Crohn’s disease on infliximab combination therapy had more relapse after withdrawing infliximab than after withdrawing the immunosuppressant. The trial does not establish a withdrawal rule for active axial disease or all fistulising phenotypes. A drug used for a second organ must remain part of the de-escalation discussion.17
After the first TNF inhibitor: sequence by reason and phenotype
A second TNF inhibitor can be reasonable after a previously effective agent loses response through immunogenicity or becomes intolerable. Persistent inflammatory disease despite adequate exposure raises a different question about mechanism. BSR notes lower response to a second TNF inhibitor after primary non-response than after secondary failure, but insufficient evidence for a universal axSpA sequence.7 The 2026 ACR summary conditionally favours switching mechanism over another agent in the same class after initial targeted-therapy failure; this is conditional advice and does not remove intestinal constraints.13
Reconsider which bowel and joint outcomes must be achieved before selecting the next agent. AGA’s intestinal efficacy groupings are not rankings for combined IBD-SpA. A therapy with strong Crohn’s or UC evidence may leave axial disease untreated. Separate label eligibility from clinical evidence, and confirm the indication-specific regimen. Shared decision-making should include injection or infusion preferences, previous adverse events, pregnancy plans and the practical ability to sustain treatment.15, 16
A biosimilar switch is not a mechanism switch. NOR-SWITCH found switching stable patients from originator infliximab to CT-P13 non-inferior overall, across several diseases; individual disease subgroups were not powered for equivalent certainty. Record product, schedule and symptoms around a switch. New symptoms deserve assessment, but their timing alone neither demonstrates biosimilar failure nor permits dismissal as expectation-related.18
New joint symptoms after TNFi → vedolizumab
Construct a timeline including the final TNF-inhibitor dose, the first vedolizumab dose, steroid reduction and the onset of symptoms. Possibilities include loss of systemic musculoskeletal suppression, previously unrecognised disease, steroid-withdrawal symptoms, an unrelated disorder and a treatment-associated adverse event. These possibilities require different responses.
Post hoc GEMINI analyses found that steroid reduction was associated with new or worsening arthritis/arthralgia in Crohn’s disease regardless of assigned treatment. The outcomes combined symptoms and diagnoses and were not a definitive imaging-based SpA assessment.19 In a 2024 prospective study with serial rheumatological assessment and MRI, no new axial or peripheral SpA developed among the 17 patients completing 24-week evaluation after vedolizumab initiation. Its small size and duration cannot exclude uncommon or delayed events.20
Accordingly, “arthralgia after vedolizumab” and “vedolizumab-induced inflammatory arthritis” are not interchangeable diagnoses. Confirm the musculoskeletal phenotype and bowel response before deciding whether to retain intestinal treatment, add a phenotype-appropriate intervention or replace therapy. Record the reasoning so a subsequent clinician can distinguish suspected causality from a temporal association.
Diarrhoea or possible new IBD during IL-17 inhibition
Evaluate persistent diarrhoea, bleeding, nocturnal symptoms, weight loss or abdominal pain promptly, including infection and other plausible causes. A temporal association alone does not diagnose drug-induced IBD. Nevertheless, intestinal safety is a substantive concern: the randomised secukinumab Crohn’s trial failed to show benefit and identified adverse outcomes, including worsening Crohn’s disease.21
Coordinate gastroenterology assessment and the prescribing specialist’s decision about interruption or discontinuation when IBD is suspected or confirmed; severe presentations require urgent care. Do not continue an otherwise successful joint regimen unquestioningly while bowel inflammation evolves. Conversely, do not label brief nonspecific diarrhoea as permanent IBD without appropriate evaluation. The replacement plan should identify how both intestinal and musculoskeletal control will be maintained.
Paradoxical psoriasis and lupus-like syndromes during TNFi
New psoriasis is not automatically failure of bowel treatment. Confirm the skin diagnosis, severity and functional impact with dermatology. A limited eruption may be treated while preserving an effective TNF inhibitor; extensive, distressing or refractory disease may justify changing therapy. A 2024 nationwide IBD cohort illustrates the trade-off between skin improvement and subsequent intestinal relapse after treatment changes. Because it was observational, treatment selection and disease severity limit causal comparisons.22
New arthralgia with rash, serositis or systemic features also raises a lupus-like reaction. Positive ANA or anti-dsDNA alone is insufficient to diagnose anti-TNF-induced lupus. Evaluate the clinical syndrome, chronology, alternative causes and possible organ involvement. An IBD cohort used compatible clinical and serological findings plus a temporal relationship and improvement following withdrawal. Its incidence estimates are setting-dependent. If a convincing syndrome is present, management includes specialist assessment and reconsideration of the implicated agent; autoantibody positivity without a syndrome should not trigger automatic withdrawal.23
When one targeted agent cannot control both domains
The 2026 ACR axSpA summary conditionally supports dual targeted therapy when axSpA or associated IBD remains active despite targeted monotherapy, with very low certainty. It remains a summary-stage recommendation.13 It must be reconciled with the specific drugs’ prescribing information: the current US upadacitinib label does not recommend combination with biological therapies for UC or Crohn’s disease, other JAK inhibitors or specified potent immunosuppressants. Guideline discussion does not confer regulatory approval for a combination.24
Such decisions belong in an experienced multidisciplinary setting after objective reassessment, adequate monotherapy trials and consideration of safer alternatives. Specify each drug’s target, the evidence supporting the combination, infectious and other risks, monitoring responsibilities and stopping criteria. Avoid informal accumulation of separate prescriptions for each organ without a shared risk review.
Finish with a shared, testable plan
The following documentation checklist translates the chapter into a follow-up decision; it is a practical aid rather than a validated instrument.
- Current state: bowel and musculoskeletal activity, evidence and unresolved uncertainty.
- Reason for change: inadequate exposure, confirmed inflammatory failure, intolerance, complication or another symptom mechanism.
- Intervention: the exact treatment change, its intended organ effects and protection of an already controlled domain.
- Reassessment: named clinical, laboratory or imaging outcomes, an appropriate review interval and an earlier-contact plan.
- Ownership: the clinician responsible for each test, prescription, safety issue and communication with the patient.
Repeated isolated steroid courses can obscure the pattern without establishing durable control. Record all rescue exposure and agree on the longer-term strategy. At follow-up, ask whether the planned outcomes occurred in each domain—not merely whether the prescription was issued. If they did not, revisit the stated mechanism before repeating the same intervention.
Sources and evidence notes
Primary studies, society guidance and prescribing information are distinguished below. Treatment examples explain clinical reasoning and do not replace indication-specific prescribing information. The chapter should be reviewed after publication of the full 2026 ACR axSpA manuscripts or a material safety update.
- Macaluso et al. IG-IBD/SIR recommendations for IBD-associated SpA. 2024. Interdisciplinary pseudo-Delphi consensus.
- AGA. Biomarkers in Crohn’s disease. 2023. Intestinal assessment; not a joint-activity measure.
- AGA. Biomarkers in UC. 2023.
- Ramiro et al. ASAS–EULAR axSpA recommendations: 2022 update. 2023. AxSpA evidence; applicability assessed by phenotype.
- Ravn et al. SANJO guideline. 2023.
- Gordon et al. ECCO extraintestinal manifestations guideline. 2024. Statements 18–19; axial and non-axial disease distinguished.
- BSR. AxSpA biologic and targeted synthetic DMARD guideline. 2025.
- Colombel et al. AGA clinical practice update on functional GI symptoms in IBD. 2019. Expert guidance.
- PANTS. Mechanisms and management of anti-TNF loss of response: three-year cohort. 2024. Prospective Crohn’s cohort.
- AGA. Therapeutic drug monitoring in IBD. 2017.
- AGA technical review on TDM in IBD. 2017. Intestinal pharmacokinetic interpretation.
- Krieckaert et al. EULAR TDM points to consider. 2023. Rheumatic-disease context.
- ACR/SAA/SPARTAN axSpA guideline summary. Posted 24 June 2026. Official summary, not the full journal manuscripts; recommendations retain their stated certainty.
- Syversen et al. NOR-DRUM B. JAMA. 2021. Randomised maintenance-infliximab TDM trial across immune-mediated diseases.
- AGA living Crohn’s pharmacological guideline. 2025. March 2026 evidence review retained recommendations.
- AGA living UC pharmacological guideline. 2024. March 2026 evidence review retained recommendations.
- Louis et al. SPARE randomised trial. 2023. Selected adults with Crohn’s disease in steroid-free remission.
- Jørgensen et al. NOR-SWITCH. 2017. Randomised originator-to-biosimilar infliximab switch trial.
- Feagan et al. Arthritis/arthralgia in GEMINI. 2019. Post hoc trial analyses; online 2018.
- Rohekar et al. Vedolizumab initiation and new-onset SpA. 2024. Small prospective clinical/MRI study.
- Hueber et al. Secukinumab in Crohn’s disease. 2012. Randomised placebo-controlled trial.
- Sanz Segura et al. Management of anti-TNF-induced psoriasis in IBD. 2024. Nationwide observational cohort.
- Picardo et al. Anti-TNF-induced lupus in IBD. 2020. Retrospective cohort.
- RINVOQ US prescribing information. Revised June 2026. Combination limitations; US regulatory context.
Educational information. Individual care depends on the clinical history, examination, investigations and applicable prescribing information.
Chapters in this collection
- Understanding IBD-associated SpA
- The Gut–Joint Connection
- Recognizing the Clinical Phenotypes
- Finding Disease in Both Directions
- Diagnosis, Tests & Important Mimics
- Imaging Without Overcalling Disease
- Initial Treatment & Symptom Control
- Choosing Therapy Across Organ Domains
- When Gut and Joints Respond Differently
- Monitoring Both Diseases
- Beyond Gut and Joints
- Special Populations & Life Stages
- Surgery, Colectomy & Pouch-related Questions
- Difficult Disease & Emerging Evidence