Azathioprine

證據等級: L5 預測適應症: 10

目錄

  1. Azathioprine
  2. Azathioprine: From Immunosuppression to Inflammatory Bowel Disease
    1. One-Sentence Summary
    2. Quick Overview
    3. Why is This Prediction Reasonable?
    4. Clinical Trial Evidence
    5. Literature Evidence
    6. Ulcerative Colitis — Secondary Prediction (Rank 9)
    7. Netherlands Market Information
    8. Other TxGNN Predictions
    9. Safety Considerations
    10. Conclusion and Next Steps
    11. Disclaimer

## 藥師評估報告

Azathioprine: From Immunosuppression to Inflammatory Bowel Disease

One-Sentence Summary

Azathioprine is a well-established purine antimetabolite immunosuppressant, historically used for organ transplant rejection prevention and autoimmune diseases. The TxGNN model predicts it may be effective for Inflammatory Bowel Disease (IBD), with 50 clinical trials and 20 publications currently supporting this direction — notably, this prediction aligns with azathioprine’s already-approved use for IBD in the Netherlands and internationally, serving as a strong validation of the model.

Quick Overview

Item Content
Original Indication Immunosuppression (autoimmune diseases, organ transplant rejection)
Predicted New Indication Inflammatory Bowel Disease
TxGNN Prediction Score 99.52%
Evidence Level L1 (multiple completed Phase 3 RCTs)
NL Market Status Marketed (Azathioprine is registered in the Netherlands; note: the evidence pack contains Taiwan regulatory data showing “Not marketed” in Taiwan)
Number of Authorizations Not available for NL in this dataset
Recommended Decision Proceed with Guardrails

Why is This Prediction Reasonable?

Azathioprine is a prodrug of 6-mercaptopurine, which inhibits purine synthesis, reduces lymphocyte proliferation, and suppresses pro-inflammatory cytokines (TNF-α, IL-6). By dampening the adaptive immune response, azathioprine effectively controls the dysregulated immune activation that drives chronic intestinal inflammation in IBD.

Inflammatory bowel disease — encompassing both Crohn’s disease and ulcerative colitis — is fundamentally an immune-mediated condition where the gut’s adaptive and innate immune systems mount an inappropriate response against intestinal flora and mucosal antigens. Azathioprine’s mechanism of suppressing lymphocyte-driven inflammation directly addresses this core pathology, making the TxGNN prediction mechanistically well-grounded.

Remarkably, this prediction is already clinically validated: azathioprine was approved for long-term therapy of Crohn’s disease in the Netherlands and is recognized worldwide as a first-line immunomodulatory maintenance therapy for both Crohn’s disease and ulcerative colitis. Multiple Cochrane systematic reviews and landmark trials (e.g., SONIC trial, NCT00094458) have confirmed its efficacy. The TxGNN model’s correct identification of this established drug-disease relationship serves as a powerful positive control for the model’s predictive accuracy.

Clinical Trial Evidence

Trial Number Phase Status Enrollment Key Findings
NCT00094458 Phase 3 Completed 508 SONIC trial: IFX + AZA combination vs. AZA or IFX monotherapy in CD naïve to immunomodulators and biologics; demonstrated superiority of combination therapy
NCT05040464 Phase 3 Recruiting 166 Head-to-head RCT comparing AZA vs. MTX as combination partner with adalimumab in Crohn’s disease
NCT03185611 Phase 3 Unknown 120 Rifaximin + thiopurine vs. thiopurine alone for preventing postoperative endoscopic recurrence in CD
NCT00976690 Phase 3 Completed 83 AZA vs. mesalazine for prevention of postoperative CD recurrence; assessed AZA superiority
NCT07424040 N/A Not Yet Recruiting 154 Infliximab monotherapy vs. IFX + AZA combination in pediatric Crohn’s disease
NCT02852694 Phase 4 Completed 192 Risk-stratified trial: MTX vs. AZA for maintaining remission in low-risk pediatric CD
NCT00554710 Phase 4 Completed 129 Top-down (early immunomodulators/biologics) vs. step-up strategy in newly diagnosed CD (Benelux study)
NCT00546546 Phase 4 Completed 120 Early immunosuppressant prescription vs. conventional strategy on 3-year CD course
NCT05584228 N/A Not Yet Recruiting 150 SMART trial: AZA + subcutaneous IFX vs. ileocecal resection in symptomatic stricturing small bowel CD
NCT03464136 Phase 3b Completed 386 Ustekinumab vs. adalimumab in biologic-naïve CD patients who failed conventional therapy including AZA

Literature Evidence

PMID Year Type Journal Key Findings
29293971 2018 Review J Crohn’s Colitis State-of-the-art overview of thiopurine (AZA/MP/TG) treatment in IBD: indications, efficacy, and safety by expert panel
19072367 2008 Mechanistic Review Expert Rev Gastroenterol Hepatol 45 years of clinical experience with thiopurines in IBD; strong data from RCTs and meta-analyses documenting efficacy
30889246 2019 Translational Research Inflamm Bowel Dis AZA induces autophagy via mTORC1 and PERK pathways — a novel molecular mechanism relevant to CD treatment
37586320 2023 Translational Research Cell Rep Med Commensal bacteria (B. wexlerae) promote AZA therapy failure in IBD by decreasing 6-MP bioavailability
22072847 2011 Clinical Review World J Gastroenterol Optimizing 6-MP and AZA therapy: 6-TGN levels correlate with efficacy, 6-MMP with toxicity
16048561 2005 Pharmacogenomics Review J Gastroenterol Hepatol AZA/6-MP pharmacogenetics and metabolite monitoring; TPMT polymorphisms and dosing implications
36462311 2023 Pharmacogenomics Biomed Pharmacother DNA methylation of TPMT affects AZA pharmacokinetics in VEO-IBD children
10499471 1999 Comprehensive Review Scand J Gastroenterol Suppl AZA clinical efficacy and safety update — documents approval for Crohn’s disease in the Netherlands
15177535 2004 Clinical Review Gastroenterol Clin North Am Critical review of 6-MP and AZA efficacy and toxicities in IBD
30954317 2019 Review Gastroenterol Hepatol Evidence on optimal duration and withdrawal of thiopurine therapy in IBD

Ulcerative Colitis — Secondary Prediction (Rank 9)

The TxGNN model also independently predicts azathioprine for ulcerative colitis (score: 99.33%, Evidence Level: L1), which is a subtype of IBD. This prediction is supported by equally robust evidence:

Trial Number Phase Status Enrollment Key Findings
NCT03101800 Phase 3 Unknown 84 Low-dose AZA + allopurinol vs. AZA monotherapy in UC — directly evaluates AZA dosing strategies
NCT02425852 Phase 4 Completed 65 Early AZA + IFX vs. corticosteroids + AZA for acute severe UC
NCT00537316 Phase 3 Terminated 242 IFX monotherapy vs. IFX + AZA vs. AZA monotherapy in moderate-to-severe active UC
NCT07235904 Phase 4 Recruiting 300 MIRACLE trial: Mirikizumab vs. AZA as standard of care in newly diagnosed moderate-to-severe UC
NCT07271069 N/A Not Yet Recruiting 150 Real-world: Ozanimod vs. AZA for UC in Japan

Key UC Literature:

PMID Year Type Journal Key Findings
40013523 2025 Cochrane Systematic Review Cochrane Database Syst Rev Updated Cochrane review: AZA and 6-MP for maintenance of remission in UC
39586616 2025 RCT Gut ACTIVE trial: Top-down IFX + AZA vs. AZA alone in acute severe UC responding to IV steroids
19392869 2009 Meta-analysis Aliment Pharmacol Ther Meta-analysis confirming AZA/6-MP efficacy in UC
27192092 2016 Cochrane Systematic Review Cochrane Database Syst Rev AZA and 6-MP for UC remission maintenance
9412914 1997 Clinical Study J Clin Gastroenterol AZA in steroid-resistant and steroid-dependent UC: clinical outcomes

Netherlands Market Information

The evidence pack contains Taiwan regulatory data (showing azathioprine as “Not marketed” in Taiwan with 0 licenses). However, azathioprine is well-established in the Netherlands:

Item Content
Market Status (NL) Marketed — azathioprine has been registered and widely prescribed in the Netherlands for decades
CBG-MEB Authorization Available as Imuran® and generic formulations
Known NL Approvals Crohn’s disease (documented since 1999, PMID 10499471), organ transplant rejection, autoimmune hepatitis, severe rheumatoid arthritis, SLE, dermatomyositis
Dosage Forms Oral tablets (25 mg, 50 mg); injectable formulations
SmPC Reference Consult the CBG-MEB Geneesmiddeleninformatiebank for current Dutch SmPC

Note: Detailed CBG-MEB license numbers (RVG numbers) were not included in this evidence pack. Please consult the CBG-MEB database for complete authorization details.

Other TxGNN Predictions

The TxGNN model generated 10 predictions for azathioprine. Beyond IBD and UC (discussed above), the remaining 8 predictions are all classified as Hold due to lack of mechanistic rationale or absence of clinical evidence:

Rank Disease TxGNN Score Evidence Level Recommendation Rationale
1 Colobomatous microphthalmia-rhizomelic dysplasia syndrome 99.99% L5 Hold Congenital developmental anomaly; no immune-mediated pathology; no clinical evidence
2 Brachydactyly-syndactyly syndrome 99.99% L5 Hold Genetic skeletal defect (GDF5/BMPR1B); not inflammatory; no clinical evidence
3 Osteoarthritis susceptibility 99.70% L5 Hold Genetic susceptibility phenotype; AZA cannot modify genetic predisposition
4 WHIM syndrome 99.68% L5 Hold Primary immunodeficiency (CXCR4 mutation); immunosuppression is contraindicated
6 Chronic granulomatous disease (AR type 5) 99.41% L5 Hold NADPH oxidase deficiency; further immunosuppression would worsen infection risk
7 Osteoarthritis 99.40% L4 Hold Degenerative/mechanical pathology; systemic immunosuppression risk outweighs benefit
8 Granulomatous disease with neutrophil chemotaxis defect 99.37% L5 Hold Innate immune deficiency; AZA would worsen immune function
10 Acromesomelic dysplasia, Hunter-Thompson type 99.27% L5 Hold GDF5 homozygous mutation; gene-driven developmental defect; no treatment logic

Important safety note: Predictions for WHIM syndrome (Rank 4), chronic granulomatous disease (Rank 6), and granulomatous disease with neutrophil chemotaxis defect (Rank 8) represent immunodeficiency conditions where azathioprine use would be mechanistically contraindicated, as further immunosuppression could cause life-threatening infections.

Safety Considerations

Please refer to the SmPC (Summary of Product Characteristics; Dutch: Samenvatting van de Productkenmerken) for comprehensive safety information. The SmPC is available via the CBG-MEB Geneesmiddeleninformatiebank.

Key safety considerations known from established clinical use include:

  • Myelosuppression: Dose-dependent bone marrow suppression (leukopenia, thrombocytopenia, anaemia); TPMT and NUDT15 genotyping recommended before initiation
  • Hepatotoxicity: Elevated 6-MMP metabolite levels associated with liver injury
  • Infection risk: Increased susceptibility to opportunistic infections due to immunosuppression
  • Malignancy risk: Long-term use associated with increased risk of lymphoproliferative disorders (particularly hepatosplenic T-cell lymphoma when combined with anti-TNF agents in young males)
  • Pharmacogenomics: TPMT and NUDT15 polymorphisms significantly affect drug metabolism; pre-treatment genotyping is recommended by EMA guidelines
  • Drug interactions: Allopurinol markedly increases 6-TGN levels (dose reduction to 25-33% required); 5-aminosalicylates may increase thiopurine toxicity

Note: Detailed DDI, warnings, and contraindications data were not available in this evidence pack. The items listed above are derived from the clinical trial and literature evidence reviewed.

Conclusion and Next Steps

Decision: Proceed with Guardrails

Rationale: Azathioprine for inflammatory bowel disease represents a validated prediction — the drug is already an approved and guideline-recommended therapy for both Crohn’s disease and ulcerative colitis in the Netherlands and globally. Multiple completed Phase 3 RCTs, Cochrane systematic reviews, and meta-analyses provide L1-level evidence confirming its efficacy in maintaining steroid-free remission. The TxGNN model’s correct identification of this established drug-disease relationship serves as a strong positive control for the model’s overall predictive validity.

To proceed, the following is needed:

  • Obtain NL-specific regulatory data: Retrieve current CBG-MEB authorization details (RVG numbers, approved indications, current SmPC) from the Geneesmiddeleninformatiebank
  • Complete safety profile: Download and parse the current Dutch SmPC for detailed warnings, contraindications, and drug interactions
  • Mechanism of action data: Query DrugBank API for structured MOA data (known: purine antimetabolite → 6-MP → 6-TGN → lymphocyte apoptosis via Rac1 inhibition)
  • Pharmacogenomic screening protocol: Ensure TPMT and NUDT15 genotyping is integrated into any prescribing pathway, per current EMA and DPWG (Dutch Pharmacogenetics Working Group) guidelines
  • Therapeutic drug monitoring: Establish 6-TGN and 6-MMP metabolite monitoring protocols for dose optimization
  • Evaluate model performance: Use this validated prediction as a benchmark to assess TxGNN model reliability for other, less-established drug-disease predictions

This report is for research purposes only and does not constitute medical advice. Drug repurposing candidates require clinical validation before application. All treatment decisions should be made by qualified healthcare professionals in accordance with current clinical guidelines and the applicable SmPC.

Data cutoff: 2026-04-03 | Evidence Pack version: v4 | Candidate ID: TW-DB00993-multi

Disclaimer

This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.



Copyright © 2026 NlTxGNN Project. For research purposes only. Not medical advice.

This site uses Just the Docs, a documentation theme for Jekyll.