Baricitinib (JAK1/2 inhibitor)

Identifying eligible participants

The two BARICADE trials ask different questions. DELAY recruits people with stage 1b or stage 2 T1D and repeated confirmation of multiple islet autoantibodies; identifying candidates therefore involves testing before stage-3 diagnosis.⁵ PRESERVE, like BANDIT, recruits after diagnosis and does not depend on presymptomatic screening.⁶¹ Trial eligibility is not an approved treatment indication.

Therapy: an oral pill to slow the immune attack

Baricitinib is an oral inhibitor of JAK1 and JAK2, proteins involved in transmitting inflammatory cytokine signals. The BANDIT protocol describes preclinical JAK-inhibitor findings in non-obese diabetic mice as a rationale for testing beta-cell preservation in humans. Those animal findings are not evidence of human prevention or cure.²

BANDIT (ACTRN12620000239965; NCT04774224) was an investigator-initiated, phase-2, double-blind, randomized, placebo-controlled trial at four Australian centres. It randomized 91 people aged 10–30, diagnosed within 100 days, to baricitinib 4 mg daily (60 people) or placebo (31) for 48 weeks, alongside usual diabetes care.¹ Participants needed at least one islet autoantibody and remaining C-peptide; the results do not establish benefit in long-standing T1D or children under ten.¹

At week 48, the median mixed-meal-stimulated mean C-peptide was 0.65 versus 0.43 nmol/L/min (P=0.001). This was the primary outcome, a measure of remaining insulin secretion. Mean daily insulin doses were 0.41 versus 0.52 U/kg/day, and CGM glucose variability was 29.6% versus 33.8%; these secondary analyses were exploratory and not adjusted for multiple comparisons. Glucose variability was also lower in the baricitinib group at baseline. Average HbA1c was similar, and a time-in-range advantage was not demonstrated at the week-48 assessment.¹

A 2025 network meta-analysis included baricitinib among 11 of 42 interventions with higher 12-month C-peptide than comparator in its main analysis. Substantial heterogeneity and uncertain rankings limit comparisons; baricitinib was not statistically significant in the sensitivity analysis restricted to trials at low risk of bias. This synthesis does not establish superiority over other therapies.⁴

Durability: the peer-reviewed follow-up published in August 2026 included 88/91 participants at week 96 (58 previously assigned baricitinib, 30 placebo). The C-peptide difference remained at week 72 but was not statistically significant at week 96 (0.43 versus 0.35 nmol/L; P=0.336). Insulin dose, HbA1c and CGM measures did not differ significantly during follow-up. The results suggest waning benefit after stopping; they do not prove the groups became equivalent or establish a safe, effective indefinite regimen.³

Safety: during the initial 48-week trial, 47/60 baricitinib and 26/31 placebo recipients reported adverse events. Seven serious events occurred: two in one baricitinib recipient and five in three placebo recipients; none was attributed to study drug. A shingles case occurred in the baricitinib group. The small sample and short exposure cannot exclude uncommon or long-term harms.¹

The US label's boxed warnings cover serious infections, mortality, malignancy, major cardiovascular events and thrombosis. Some comparative warning evidence comes from another JAK inhibitor studied in rheumatoid arthritis patients aged 50 or older with cardiovascular risk factors; infections, malignancies and thrombosis have also been observed with baricitinib. Those findings are not a numerical risk estimate for children with T1D. The June 2026 label additionally warns that hypoglycemia, including severe episodes, has been reported after starting treatment in people with diabetes, and advises considering increased glucose monitoring. It also requires clinical attention to infections, blood counts and renal/hepatic function.⁷

Approval and access: the checked US label covers adults with rheumatoid arthritis after inadequate response to TNF blockers, severe alopecia areata, and hospitalized COVID-19 requiring specified respiratory support. It contains no T1D indication. BARICADE remains research, and Australian sites listed here are trial locations. Oral delivery does not establish low cost, reimbursement, routine T1D supply or authorization in any particular country.⁷⁶⁵

What's coming

Eli Lilly's randomized, placebo-controlled phase-3 BARICADE-PRESERVE (NCT07222332) studies recent stage-3 diagnosis, with change in C-peptide AUC at week 52 as the primary endpoint. BARICADE-DELAY (NCT07222137) studies stage 1b/2 and time to stage-3 diagnosis. Both list ages 1–35 and additional eligibility and safety criteria. On 17 September 2026, both registries were recruiting with no posted results. Planned enrolment is 300 for PRESERVE and 150 for DELAY; primary completion is estimated for July 2028 and July 2031 respectively. These estimates are not readout or approval promises.⁶⁵