AGA 2024 - Moderate-to-Severe UC Pharmacotherapy Guideline

Citation: Singh S, Loftus EV Jr, Limketkai BN, Haydek JP, Agrawal M, Scott FI, Ananthakrishnan AN, on behalf of the AGA Clinical Guidelines Committee. “AGA Living Clinical Practice Guideline on Pharmacological Management of Moderate-to-Severe Ulcerative Colitis.” Gastroenterology 2024;167(7):1307–1343. Published online December 2024; updates the AGA’s 2020 UC guideline. Source file: raw/Singh 2024 - AGA Living Guideline Pharmacological Management Moderate-Severe UC.pdf.

Type: Society “living” clinical practice guideline (GRADE-based, with an explicit network meta-analysis underpinning positioning recommendations — see Limitations for how this differs methodologically from ACG 2025 - Ulcerative Colitis Guideline).

Scope

Narrower than the ACG 2025 guideline: covers only pharmacological management of moderate-to- severely active UC in adult outpatients (defined here as stool frequency ≥2 and rectal bleeding ≥2 on the 2-item PRO scale, or Mayo endoscopic subscore ≥2) — 14 recommendations across use/ positioning of advanced therapies, immunomodulator monotherapy, combination biologic + immunomodulator therapy, de-escalation of therapy, and step therapy vs. early advanced-therapy use. Explicitly excludes mild-to-moderate UC, hospitalized ASUC, diagnosis/monitoring, and surgery — all of which the ACG 2025 guideline does cover. A “living guideline”: recommendations are reviewed and updated every 6 months as new phase 3/4 trial data emerge, unlike the fixed-point-in-time ACG document.

Key points

• Recommends (strong, moderate-high quality) using any of: infliximab, golimumab, vedolizumab, tofacitinib, upadacitinib, ustekinumab, ozanimod, etrasimod, risankizumab, guselkumab — over no treatment. Suggests (conditional, moderate quality): adalimumab, filgotinib, or mirikizumab over no treatment.
• Efficacy-bucket positioning — the guideline’s central methodological move: rather than ranking by drug class, the panel grouped individual drugs into “higher/intermediate/lower efficacy” buckets using a combination of absolute risk difference over placebo (phase 3 RCTs) and network meta-analysis (NMA) p-scores from indirect comparisons:
  • Advanced-therapy-naive patients: Higher efficacy = infliximab, vedolizumab, ozanimod, etrasimod, upadacitinib, risankizumab, guselkumab. Intermediate = golimumab, ustekinumab, tofacitinib, filgotinib, mirikizumab. Lower = adalimumab.
  • Patients previously exposed to ≥1 advanced therapy (particularly TNF antagonists): Higher = tofacitinib, upadacitinib, ustekinumab. Intermediate = filgotinib, mirikizumab, risankizumab, guselkumab. Lower = adalimumab, vedolizumab, ozanimod, etrasimod. Note vedolizumab moves from the higher bucket (naive) to the lower bucket (biologic-exposed) — efficacy bucketing is population-specific, not a fixed drug ranking.
  • Recommendation in both populations: use a higher- or intermediate-efficacy medication rather than a lower-efficacy one (conditional, low certainty of evidence in both cases).
• Combination therapy: infliximab + immunomodulator (thiopurine) recommended over infliximab or immunomodulator alone (conditional, moderate quality) — direct trial evidence from UC-SUCCESS (corticosteroid-free remission 39.7% combination vs. 22.1% infliximab-alone vs. 23.7% azathioprine-alone at week 16; mucosal healing 62.8% vs. 54.6% vs. 36.8%). Adalimumab or golimumab
  • immunomodulator also recommended over either monotherapy (conditional, low quality — extrapolated from TNF-class mechanism similarity and indirect CD-trial evidence [SONIC], not a UC-specific RCT for these two agents specifically). No recommendation (knowledge gap) for combining immunomodulators with non-TNF biologics (vedolizumab/ustekinumab/anti-IL-23s) — observational data and one RCT subgroup show no clear benefit, unlike the TNF class.
• Safety trade-off of combination therapy: no increase in serious infections vs. monotherapy in UC-SUCCESS itself, but observational French national data show combination therapy associated with higher lymphoma risk than TNF-antagonist monotherapy (HR 2.53) or thiopurine monotherapy (HR 2.35) — this lymphoma-risk number is more specific than anything in the ACG 2025 guideline, which only broadly mentions a “6-fold increased risk…compared with unexposed patients” for combination vs. no therapy.
• De-escalation: suggests against withdrawing TNF antagonists in patients who’ve reached corticosteroid-free remission ≥6 months on TNF + immunomodulator combination therapy (conditional, very low quality — withdrawal associated with a 2-fold relapse-risk increase, 31.5% vs 11.2% at 12 months). No recommendation on withdrawing the immunomodulator instead while continuing the biologic (knowledge gap — trial data mostly drawn from Crohn’s populations, relapse risk not significantly different in the available data: 16.8% withdrawal vs 14.9% continuing). This whole topic (de-escalation) is not addressed at all in the ACG 2025 guideline.
• 5-ASA discontinuation on escalation: suggests stopping 5-ASA once a patient has escalated to immunomodulators or advanced therapy after 5-ASA failure (conditional, low quality) — based on a pooled individual-patient-data analysis across 10 RCTs (6044 patients) showing no treatment-effect modification from concomitant 5-ASA (adjusted RR 1.04, 95% CI 0.78–1.39). Exception: patients with residual proctitis/incomplete response may still benefit from adding rectal 5-ASA.
• Step therapy: suggests early use of advanced therapies (with or without immunomodulator), rather than gradual step-up after 5-ASA failure, once a patient meets moderate-to-severe criteria (conditional, very low quality). Explicit comment: patients who value 5-ASA’s safety profile over immunosuppressive-therapy efficacy may reasonably choose gradual step therapy instead — framed as a genuine shared-decision point, not a hard rule.
• Immunomodulator monotherapy: suggests against thiopurine monotherapy for induction (very low quality), for thiopurine monotherapy over no treatment for maintenance of remission induced by corticosteroids (low quality), and against methotrexate monotherapy for either induction or maintenance (low quality) — directionally identical to the ACG 2025 guideline’s thiopurine/ methotrexate stance, though this guideline supplies the underlying RCT-level numbers (Tables 11–12).
• Safety differentiation between classes: vedolizumab and anti-IL agents associated with lower serious-infection risk than TNF antagonists in observational data (a prior meta-analysis found 32% lower risk with vedolizumab) — may be preferred in patients at higher infection risk. TNF antagonists carry an FDA black-box warning for malignancy (lymphoma, particularly with combination therapy in young men — matches existing wiki teaching from the Sabiston/Inflammatory Bowel Disease chapter). JAK inhibitors carry the ORAL SURVEILLANCE cardiovascular/malignancy signal (same trial discussed in ACG 2025 - Ulcerative Colitis Guideline) — both US and EMA guidance restrict first-line JAK inhibitor use, more cautiously in patients ≥65, smokers, or with cardiovascular/ cancer history.
• Pregnancy (a topic the ACG 2025 guideline explicitly excludes from scope): JAK inhibitors and S1P receptor modulators should be avoided in women of childbearing age contemplating pregnancy — limited human safety data, animal studies suggest adverse effects at supratherapeutic doses. TNF antagonists and other biologics, by contrast, have reassuring large-registry data (PIANO registry) showing no significant increase in adverse pregnancy/neonatal outcomes.

Limitations

• Explicit, wiki-relevant methodological disagreement with ACG 2025 - Ulcerative Colitis Guideline: the ACG guideline states it “determined not to use indirect or post hoc data to guide our positioning recommendations” and explicitly declines to rank therapies via network meta-analysis. This AGA guideline does the opposite — its entire efficacy-bucket framework (higher/intermediate/ lower) is built substantially on NMA p-scores from indirect, non-head-to-head comparisons, on the explicit reasoning that with 11+ approved agents and few head-to-head trials, some positioning guidance was judged more useful to clinicians than none. Both guidelines agree on the one genuine head-to-head data point (VARSITY: vedolizumab > adalimumab) and both name infliximab/upadacitinib favorably, but the AGA’s population-specific bucket-flipping (vedolizumab: higher in naive → lower in biologic-exposed) is a level of granularity the ACG guideline deliberately avoids producing, and a reader should not treat the two guidelines’ positioning tables as simply additive/reconcilable — they reflect a genuine, stated difference in how much weight indirect evidence should carry, not just a difference in which trials were newest at each guideline’s writing.
• Almost all comparative-efficacy (bucket) recommendations are conditional/low-certainty — the guideline itself repeatedly flags the underlying NMA evidence as low-to-very-low certainty even where a directional recommendation was still issued.
• Narrower scope than ACG 2025 - Ulcerative Colitis Guideline — no mild-moderate UC, no ASUC, no diagnosis/monitoring, no surgery. Use both pages together for full UC coverage; do not treat this page as comprehensive on its own.
• Living-guideline format means this snapshot (December 2024 text) may already be stale relative to the AGA’s 6-monthly update cycle by the time this is read — worth checking for a newer version before relying on specific numeric efficacy-bucket placements in high-stakes decisions.
• Cross-checked against the British IBD 2025 guideline (BSG 2025 - IBD Guidelines (Adults), ingested alongside this one) — no direct contradiction found on UC pharmacotherapy positioning; see Ulcerative Colitis for the combined synthesis.

Relevance

Adds UC-specific advanced-therapy positioning/sequencing detail to Ulcerative Colitis beyond what ACG 2025 - Ulcerative Colitis Guideline provided — particularly combination-therapy evidence (UC-SUCCESS trial specifics), de-escalation guidance (a topic ACG doesn’t address at all), step-therapy framing, and pregnancy-specific drug-class guidance. The NMA-vs-no-NMA methodological split between this guideline and the ACG one is itself worth preserving as an explicit note on Ulcerative Colitis, per this wiki’s convention of flagging contradictions between sources rather than silently merging them into one voice.