Three extra tubes: the glucose study that answers the inversion in children
also called paediatric glucose substudy menin inhibitor · C-peptide menin inhibitor children
Add fasting glucose, insulin, C-peptide and glucagon to blood that is already being drawn from children on menin inhibitors, and take the same tests again a few months after the drug stops. It would show whether the effect seen in diabetic mice happens in a child, in either direction.
The same thing, for a scientist
The complement to `exp-endocrine-surveillance-panel-menin-treated-children`, which covers the mineral and growth axes and deliberately excludes the glycaemic one. This substudy adds fasting glucose, insulin, C-peptide, glucagon, HbA1c and, where feasible, a fasting GLP-1, at baseline, monthly on treatment, and at 3, 6 and 12 months **after** the last dose. It uses blood already drawn. It exists because a nine-paper preclinical literature predicts a glycaemic effect [E1][E2], because the single strongest in-vivo result is a fall in fasting glucose with a rise in beta-cell mass over 12 weeks [E1], and because neither approved label contains a single glycaemic parameter [E3][E4].
Why C-peptide and glucagon and not just glucose
Glucose alone cannot distinguish the three live outcomes. Glucose plus C-peptide separates increased insulin secretory capacity from improved insulin sensitivity; glucagon tests the specific counter-direction reported in germline Men1(+/-) males, where fasting hyperglycaemia is driven by raised basal glucagon and hepatic glucose output rather than by insulin resistance 5; HbA1c integrates over the treatment period and is robust to the corticosteroid and nutrition noise that dominates single fasting values in this population. A fasting GLP-1 would test the gut arm directly 2.
The confounds, which are severe and must be designed around
Children on menin inhibitors are frequently also receiving corticosteroids, asparaginase, parenteral nutrition and antifungals, all of which move glucose hard. The design handles this by (a) requiring a steroid-free interval of at least 7 days for the protocol-defined fasting draw where clinically possible, or recording steroid exposure as a covariate; (b) using HbA1c as a steroid-robust integrator; (c) making the post-treatment draws the primary comparison, because by 3-12 months off drug the leukaemia-directed confounders have largely resolved and any persistent glycaemic shift is attributable.
The post-treatment arm is again the part no existing requirement compels - FDA's PMR 4730-1 concerns patients on treatment for at least three years.
What each outcome would mean
| Result | Interpretation |
|---|---|
| Fasting glucose falls, C-peptide rises | The beta-cell pharmacology translates; the inversion is real in humans; monitor for post-withdrawal hypoglycaemia |
| No change in anything | The benefit literature does not translate at clinical exposures; hyp-menin-inhibition-is-tissue-divergent loses its benefit arm and this beat reverts to a one-sided toxicity story with better evidence than it has now |
| Glucose rises with raised glucagon | The germline Men1(+/-) phenotype is being reproduced pharmacologically; a genuinely new and unanticipated harm |
| Post-treatment hypoglycaemia | Expanded beta-cell mass persisting after drug withdrawal; the most clinically urgent possible finding and one nobody is currently watching for |
| design | Prospective observational substudy nested in existing paediatric menin-inhibitor treatment and survivorship follow-up, using blood already drawn |
|---|---|
| readout | fasting glucose, insulin, C-peptide, glucagon, HbA1c, fasting GLP-1 where feasible; baseline, monthly on treatment, and 3, 6 and 12 months after last dose |
| controls | steroid exposure recorded as covariate; steroid-free interval of >=7 days for protocol draws where clinically possible; post-treatment draws as the primary comparison |
| cost tier | low |
| timeline | 24 months to first readout on an existing cohort |
- A null result would remove the benefit arm of this beat's central hypothesis and leave the harms standing alone, so the study is not a search for reassurance.
- If beta-cell mass rises on treatment, withdrawal hypoglycaemia is a foreseeable harm that current follow-up would miss entirely.
- Measuring glucose without measuring the mineral and growth axes at the same visits would waste the draw; this substudy should be run with exp-endocrine-surveillance-panel-menin-treated-children, not instead of it.
- Would COG or the Innovative Therapies for Children with Cancer consortium adopt a six-analyte add-on to an existing follow-up schedule?
- Is there already a banked paediatric sample set from AUGMENT-101 on which fasting insulin and C-peptide could be run retrospectively?
- What is the smallest cohort that could detect a clinically meaningful HbA1c shift given the confounding in this population?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A menin inhibitor lowered fasting glucose and raised beta-cell mass over 12 weeks in diabetic mice.
Menin-regulated Pbk controls high fat diet-induced compensatory beta cell proliferation. · Ma J, Xing B, Cao Y, He X, et al. · EMBO Molecular Medicine 2021What this evidence cannot show
Tool compound, n=4-5, adult mice.
Evidence that argues against this (1)
- No clinical menin inhibitor has been shown to have any glycaemic effect in any species. Syndax Pharmaceuticals et al. 2026
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Menin inhibition also raises gut-derived GLP-1 and improves glucose tolerance in obese mice.
Targeted inhibition of menin promotes beta-catenin-mediated GLP-1 expression and improves glucose tolerance in high-fat diet-induced obese mice. · Cao X, Yu L, Zhang Q, Cheng Z, Meng H, Wang C, et al. · FASEB Journal 2025What this evidence cannot show
Tool compound; mouse; no human data.
Evidence that argues against this (1)
- The mechanism runs through intestinal beta-catenin gain, which carries its own hazard and may not be desirable to reproduce in a child. Chamberlain CE et al. 2014
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The REVUFORJ label contains no glucose, insulin, diabetes or HbA1c term and no glycaemic laboratory row.
REVUFORJ (revumenib) tablets, for oral use - US Prescribing Information (SPL version 5, published 20 Feb 2026) · Syndax Pharmaceuticals, Inc. · DailyMed / FDA Structured Product Label, setid 6eb3cdbc-0e74-477d-82d6-3bb172d3f63f 2026What this evidence cannot show
Text search of the DailyMed SPL, 2026-09-19.
Evidence that argues against this (1)
- Glucose is on every routine chemistry panel, so the values very likely exist in the sponsor database and were not selected for the label. Syndax Pharmaceuticals et al. 2026
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The KOMZIFTI label likewise contains no glycaemic term or row.
KOMZIFTI (ziftomenib) capsules, for oral use - US Prescribing Information (SPL version 1, published 16 Dec 2025) · Kura Oncology, Inc. · DailyMed / FDA Structured Product Label, setid b650f696-3391-4274-8b55-a5f5e9d04769 2025What this evidence cannot show
Same method; adults only, so less directly relevant to a paediatric substudy.
Evidence that argues against this (1)
- Ziftomenib has no paediatric indication, so no children are exposed on label. Kura Oncology et al. 2025
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Germline Men1(+/-) male mice develop fasting hyperglycaemia at 12 months driven by raised glucagon and hepatic glucose output.
Male Men1 heterozygous mice exhibit fasting hyperglycemia in the early stage of MEN1. · Gao Z, Zhang L, Xie W, Wang S, Bao X, Guo Y, et al. · Journal of Endocrinology 2016What this evidence cannot show
Sex-specific, single time point, germline model.
Evidence that argues against this (1)
- Acute and pharmacological menin loss improves rather than worsens glucose handling at shorter durations. Yang Y et al. 2010
P1 and P3: does fasting glucose fall while PTH rises in the same patients, and does either persist after the last dose?
Fills the verified absence: adds the glycaemic panel neither label reports, with the post-treatment draws as the primary comparison
- co-occurs with →The cheapest study that would settle this: nine assays and a hand radiograph, on children already being followed
Companion substudy: that panel covers the mineral and growth axes and deliberately excludes the glycaemic one; both should run on the same draws
The glucagon arm tests the counter-direction: pharmacological reproduction of the germline Men1(+/-) fasting-hyperglycaemia phenotype