Summary: In 40 autoantibody-positive relatives at high risk for type 1 diabetes with low first-phase insulin release, this post hoc analysis of the TrialNet Oral Insulin trial found that oral insulin (compared with placebo) favourably shifted a combined glucose and C-peptide surrogate over a median 2.7 years, with the centroid ratio differing between arms (p=0.037, adjusted). The parent trial showed no overall effect on diabetes onset, so this is an exploratory metabolic signal, not proven prevention.
PICO Summary
| Element | Detail |
|---|---|
| Population | n=40 relatives of people with type 1 diabetes who had multiple islet autoantibodies and low first-phase insulin release (secondary stratum 1), drawn from the multicentre TrialNet Oral Insulin prevention trial; post hoc analysis, median follow-up 2.7 years (USA-led, international). |
| Intervention | Oral insulin 10 mg daily. |
| Comparison | Matched placebo. |
| Outcome | Combined glucose and C-peptide surrogate over the OGTT (30-120 min). Glucose rose and C-peptide declined on placebo, whereas glucose rose minimally and C-peptide increased on oral insulin. The change in the centroid ratio of these markers differed between groups (p=0.037, adjusted for age, BMI, and baseline C-peptide and glucose). No 95% CI, ARR, or NNT was reported, and the parent trial showed no overall effect on the diagnosis of type 1 diabetes. |
Expert Commentary
This analysis should be read as hypothesis-generating rather than practice-changing. The parent TrialNet Oral Insulin trial was negative for its prespecified endpoint, the clinical diagnosis of type 1 diabetes, and the only earlier positive signal was confined to a low first-phase-insulin-release subgroup. Here the investigators went a step further with a post hoc, surrogate-endpoint analysis in just 40 high-risk relatives, showing that a combined glucose and C-peptide metric was preserved on oral insulin while it deteriorated on placebo, with a between-group difference in the centroid ratio at p=0.037. The headline value of the work is methodological: a combined metabolic marker may detect a treatment signal earlier than waiting years for overt diabetes, which could shorten future prevention trials. The principal limitation is that this is an unplanned analysis of a tiny stratum from a trial that was already negative overall, so the finding is fragile, susceptible to multiplicity, and at clear risk of false-positive inference; it demonstrates a metabolic association, not delayed or prevented disease. Can I use this with my patients? Not yet. There is no high-risk relative in whom these data justify starting oral insulin outside a trial, and the result does not change screening or prevention practice. The sensible path is to treat this as a candidate surrogate endpoint that needs prospective validation in an adequately powered prevention study before it informs any clinical decision.
References
Triolo TM, Jacobsen LM, Cuthbertson D, Sims EK, Ismail HM, Redondo MJ, et al. Effect of oral insulin on early combined glucose and C-peptide endpoints in individuals at high-risk for type 1 diabetes. Pediatr Diabetes. 2024;2024:8343868. doi:10.1155/2024/8343868