Abatacept (CTLA4-Ig)
Bristol Myers Squibb (Orencia; repurposed)
Stage-1 primary outcome not significant.
What it is
Abatacept modulates T-cell activation. A 212-person stage-1 trial did not establish delayed progression to abnormal glucose tolerance or clinical T1D. A separate 112-person trial after diagnosis found greater C-peptide preservation; that is not evidence of delaying diagnosis. The current US Orencia label has no T1D indication.
Editorial review: .
Most recent recorded citation date: 2026-09-01. Only explicit date metadata is included; an undated citation may be newer. This does not mean every claim was reviewed on that date.
Trial status, labels and access can change between reviews. How we review the evidence · How to read the evidence
Evidence behind this assessment
Key evidence notes. Study results, product eligibility and access answer different questions.
- Who was studied?
- Study populations and analysis groups vary. Product age limits alone do not describe who was studied.See the linked studies and their populations →
- Benefit or performance
- Delay of onset: In the 212-person stage-1 trial, progression to abnormal glucose tolerance or diabetes was not significantly reduced (hazard ratio 0.702; 95% CI 0.452–1.09; P=0.11). This does not establish a delay in clinical diagnosis.[1]
- Important harms and treatment burden
- Safety: The US label warns about serious infections, hypersensitivity and immunosuppression; live vaccines should not be given during treatment or within three months after stopping. Safety in arthritis or transplantation cannot establish the benefit–risk balance for presymptomatic T1D.[4]
- Approval and country access
- No T1D indication in the current US label. Country-specific off-label access, supply, cost and reimbursement have not been established here.Approval, trial recruitment, local supply and funding are separate. Check the cited label or access source.
- Follow-up and remaining uncertainty
- Durability: After two years of treatment in recent-onset T1D, follow-up one year after stopping still found higher adjusted C-peptide and a model-estimated 9.5-month lag. Both groups continued to lose function; this is postdiagnosis preservation, not prevention durability.[3]
Research status alone does not establish approval, clinical benefit or local availability.
Editorial score: calculation and evidence
A weighted editorial judgment on a 0–100 scale, not a probability of success or a measured treatment effect. Higher criterion scores mean more favorable assessments.
Default calculation: 40 × 30 + 45 × 20 + 70 × 20 + 35 × 15 + 20 × 15 = 4325; divide by total weight 100. Unrounded weighted result: 43.25.
In the 212-person stage-1 trial, progression to abnormal glucose tolerance or diabetes was not significantly reduced (hazard ratio 0.702; 95% CI 0.452–1.09; P=0.11). This does not establish a delay in clinical diagnosis.[1]
After two years of treatment in recent-onset T1D, follow-up one year after stopping still found higher adjusted C-peptide and a model-estimated 9.5-month lag. Both groups continued to lose function; this is postdiagnosis preservation, not prevention durability.[3]
The US label warns about serious infections, hypersensitivity and immunosuppression; live vaccines should not be given during treatment or within three months after stopping. Safety in arthritis or transplantation cannot establish the benefit–risk balance for presymptomatic T1D.[4]
The prevention trial studied stage-1 relatives with normal glucose and at least two qualifying autoantibodies; people positive for insulin autoantibodies were excluded (to avoid overlap with the oral-insulin trial). Eligibility was ages 6–45 at screening; baseline ages extended into the 50s. Its null primary result cannot be generalized into a proven stage-2 benefit.[1]
Current US indications include rheumatoid arthritis, specified juvenile and psoriatic arthritis, and prevention of acute graft-versus-host disease in a specified transplant setting. They do not include T1D. This label does not establish T1D supply, reimbursement or price.[4]
Editor’s take
The stage-1 result remains uncertain rather than proof of no effect: its confidence interval includes benefit and no benefit. Postdiagnosis C-peptide preservation is a separate finding. The retained numerical ratings are provisional editorial inputs: prevention-category ratings have not been calibrated separately for prediagnosis delay versus postdiagnosis C-peptide preservation. They are not probabilities or validated effect sizes, and no numerical access estimate is established.
The full picture
What was studied
Abatacept (Orencia) is a fusion protein that binds CD80/CD86 and blocks a co-stimulation signal involved in T-cell activation. It can affect immune defenses; describing it as sparing the rest of the immune system would overstate its selectivity. The current US label covers specified arthritis indications and prevention of acute graft-versus-host disease in a specified transplant setting, not T1D.4
Stage 1: progression was not significantly delayed
The stage-1 TrialNet trial randomized 212 relatives of people with T1D to abatacept (101) or placebo (111). Participants had normal glucose tolerance and at least two qualifying autoantibodies; people positive for insulin autoantibodies were excluded to avoid overlap with the oral-insulin trial. Ages 6–45 applied at screening: Table 1 reports baseline ranges of 6.8–52.8 and 7.2–53.7 years. Fourteen infusions were planned over one year; 58.4% and 44.7% respectively completed all 14. Progression to abnormal glucose tolerance or diabetes occurred in 35 versus 46 participants (hazard ratio 0.702; 95% CI 0.452–1.09; P=0.11). This failed to establish the primary benefit; it is not proof of no possible effect. C-peptide at month 12 favored abatacept (P=0.03), but that laboratory finding does not establish delayed clinical diagnosis.1
After diagnosis: a different outcome
The earlier phase-2 trial randomized 112 people with recent-onset T1D (77 abatacept, 35 placebo). Eligibility was ages 6–45, but actual enrolled ages were 6–36. The protocol planned 27 infusions over two years; 83% of all possible infusions were administered, and 103 participants completed the two-year mixed-meal test. Adjusted mean stimulated C-peptide was 0.378 versus 0.238 nmol/L, 59% higher with abatacept (95% CI 6.1%–112%; two-sided P=0.0029). A model estimated a 9.6-month lag in C-peptide decline (95% CI 3.47–15.6); it did not measure a delay in diagnosis. Decline was approximately parallel after six months. Insulin doses did not differ at 24 months.2
One year after treatment stopped, the same cohort's follow-up reported adjusted mean C-peptide of 0.217 versus 0.141 nmol/L (one-sided P=0.046) and an estimated 9.5-month lag (95% CI 3.44–15.7). HbA1c remained lower, while insulin doses were nearly the same. The original 112 randomized participants are not a claim that all 112 contributed measurements at 36 months. Continued C-peptide decline and the short follow-up after withdrawal limit any claim of permanent preservation.3
Harms and interpretation
In the postdiagnosis trial, infusion reactions affected 17/77 abatacept and 6/35 placebo participants; infections affected 32/77 and 15/35, with no excess detected in that small study. Its adverse-event table also records one accidental death judged unrelated to study treatment and one secondary malignancy, without established causality.2 The stage-1 report lists five serious adverse events (two abatacept, three placebo), more skin/connective-tissue events with abatacept, and two cancers in abatacept recipients during observation (breast and thyroid). These observations do not establish that abatacept caused the cancers. The US label separately warns about serious infections, hypersensitivity and immunosuppression, and excludes live vaccines during treatment and for three months after stopping. Trial-specific reassuring comparisons do not remove those warnings.214
Research access and remaining questions
RELAY (NCT03929601) is an actual ongoing phase-2 study after diagnosis, not merely a proposed combination. Participants receive rituximab followed by subcutaneous abatacept or placebo. The registry update posted 1 September 2026 reports 74 actual participants, active—not recruiting status, estimated primary completion in March 2027 and no posted results. That plan does not demonstrate combination efficacy or stage-1/2 prevention.5
T1D is absent from the reviewed US label. This review does not establish access, prices, coverage or supply for off-label T1D treatment in the US or other countries. TrialNet studies received public/foundation support and Bristol Myers Squibb supplied study drug; the stage-1 report also describes company manuscript review but no company role in study conduct or data analysis. Score limitation: the retained numerical inputs have not been calibrated separately for delaying diagnosis and preserving C-peptide after diagnosis; they must not be read as validated prevention effects.413
Coming soon
ETA · Investigational for T1D; no approval date established by the reviewed sources.
- →RELAY (NCT03929601) tests rituximab followed by subcutaneous abatacept or placebo after diagnosis; 74 enrolled and no results posted. · Active, not recruiting; registry update posted 1 September 2026; primary completion estimated March 2027.
Sources
- [1]Abatacept for Delay of Type 1 Diabetes Progression in Stage 1 Relatives at Risk · Peer-reviewed study · 2023-03-15 — Original full report: 212 randomized (101 abatacept, 111 placebo); primary endpoint not significant. Safety and age-at-screening limits are discussed below.
Russell et al. Original stage-1 randomized trial, Diabetes Care (2023).
- [2]Co-stimulation modulation with abatacept in patients with recent-onset type 1 diabetes · Peer-reviewed study · 2011-06-28 — Original randomized trial: 112 randomized; 103 completed the two-year mixed-meal test. C-peptide endpoint after diagnosis, not time to diabetes.
Orban et al. Original recent-onset randomized trial, Lancet (2011).
- [3]Costimulation modulation with abatacept: follow-up 1 year after cessation of treatment · Peer-reviewed study — Original follow-up of the same trial, not a second independent randomized cohort. Three-year C-peptide comparison used a one-sided P value.
Orban et al. Follow-up after treatment cessation, Diabetes Care (2014).
- [4]Orencia (abatacept): US prescribing information · Regulatory decision — Revised November 2025. Sections 1, 5.2–5.6 and 12.1; US indications and warnings, not a T1D approval or global-access source.
Current US Orencia prescribing information, revised November 2025, sections 1, 5 and 12.1.
- [5]RELAY: rituximab followed by abatacept or placebo in new-onset T1D (NCT03929601) · Trial registry · 2026-09-01 — Current registry reports active, not recruiting, 74 actual participants, and no posted results; estimated dates may change.
ClinicalTrials.gov NCT03929601, update posted 1 September 2026; checked 17 September 2026.