A drug that fixes diabetes in mice, and neither label contains the word glucose
also called glucose absence menin inhibitor label · no glycaemic endpoint menin · REVUFORJ KOMZIFTI glucose
Menin blockers lower blood sugar and raise the number of insulin-making cells in diabetic mice. We read both approved labels word by word: neither contains glucose, insulin, diabetes, HbA1c, hyperglycaemia or hypoglycaemia even once.
The same thing, for a scientist
A verified, machine-checkable absence. Both US prescribing informations were downloaded from DailyMed on 2026-09-19, HTML-stripped to plain text, and searched case-insensitively. In **REVUFORJ** (87,856 characters, SPL v5, 20 Feb 2026) and **KOMZIFTI** (59,081 characters, SPL v1, 16 Dec 2025), the following term counts are **zero in both**: glucose, glycem*, glycaem*, diabet*, insulin, HbA1c, hyperglyc*, hypoglyc*, islet, GLP [E1][E2]. The selected laboratory-abnormality tables confirm it: revumenib's table runs phosphate, AST, ALT, creatinine, potassium, intact PTH, ALP, triglycerides, cholesterol and corrected calcium; ziftomenib's runs AST, potassium, albumin, ALT, sodium, creatinine, ALP, bilirubin and potassium increased [E1][E2]. No glycaemic parameter appears in either. This is the exact endpoint the animal pharmacology predicts would move, and it is the one endpoint neither programme reports.
The two tables, side by side
REVUFORJ, Table 8 (denominators 139-240) 1: Phosphate increased 51; AST increased 44/6; ALT increased 40/8; Creatinine increased 38/2; Potassium decreased 34/12; Parathyroid hormone, intact increased 34; ALP increased 33/<1; Triglycerides increased 27/3; Phosphate decreased 25; Cholesterol increased 17/0; Calcium corrected increased 15/0.
KOMZIFTI, Table 4 (denominators 108-110) 2: AST increased 53/4; Potassium decreased 52/22; Albumin decreased 51/5; ALT increased 50/6; Sodium decreased 49/3; Creatinine increased 45/4; ALP increased 41/5; Bilirubin increased 27/6; Potassium increased 26/6.
Ziftomenib's table has no calcium, phosphate or PTH row at all - and neither table has a glucose row.
Why the absence is not innocent
Serum glucose is on every routine chemistry panel run on every one of these patients, every visit. It is not an extra assay, an extra tube or an extra dollar. The data exist in both sponsors' databases already; they were simply not selected for the label's table. The same is true of albumin-corrected calcium for ziftomenib.
So this is not a gap in knowledge, it is a gap in reporting, and it is a gap in the one direction that would test the most interesting mechanistic prediction the preclinical literature makes about this drug class in normal tissue. concept-endocrine-endpoint-absence-menin-trial-registries established that no endocrine outcome measure exists in the three lead registries; this entry extends that to the labels, where PTH did survive into print and glucose did not.
What a glucose row would be worth
If fasting glucose fell in treated patients, it would be the first human confirmation of the beta-cell pharmacology and the first demonstration that a clinical menin inhibitor engages the endocrine axis at all, which would change how the whole class is monitored in children. If it did not move, the entire benefit literature would be shown not to translate at clinical exposures - which would simplify this beat enormously and strengthen, not weaken, the toxicity case by removing the offsetting argument. Either result is decisive; the measurement is free; nobody has reported it.
| method | DailyMed SPL HTML downloaded 2026-09-19, tag-stripped, case-insensitive term counts |
|---|---|
| revuforj chars | 87856 |
| komzifti chars | 59081 |
| revuforj setid | 6eb3cdbc-0e74-477d-82d6-3bb172d3f63f |
| komzifti setid | b650f696-3391-4274-8b55-a5f5e9d04769 |
zero terms
[ "glucose", "glycem", "glycaem", "diabet", "insulin", "HbA1c", "hyperglyc", "hypoglyc", "islet", "GLP" ]
- A class that raises beta-cell mass in animals is being given to children with no glycaemic monitoring requirement of any kind.
- If beta-cell mass rises on treatment, the hazard is on withdrawal - hypoglycaemia in a child nobody is watching for it in.
- The absence cuts against the benefit story too: with no human glucose data, the offsetting argument to the toxicity case is entirely preclinical and entirely tool-compound.
- Do the sponsors' safety databases contain post-baseline glucose values, and would a request under FDA's transparency provisions or a data-sharing platform surface them?
- Did the FDA multi-discipline review for either NDA tabulate glucose? The review documents are public and have not been read for this question.
- Is there a single published menin-inhibitor case report with a glucose time course?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The REVUFORJ prescribing information contains no occurrence of glucose, insulin, diabetes, HbA1c, hyperglycaemia, hypoglycaemia, islet or GLP, and its selected laboratory table has no glycaemic 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 HTML rendering on 2026-09-19 after tag stripping; a term embedded only in an image or a PDF-only appendix would be missed, though the SPL is the authoritative electronic text. The absence concerns reporting, not necessarily measurement.
Evidence that argues against this (1)
- Absence from the label does not mean absence from the dataset: glucose is on every routine chemistry panel, so the data almost certainly exist and were simply not selected for the table, which makes the absence easier to remedy but no less real. Syndax Pharmaceuticals et al. 2026
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The KOMZIFTI prescribing information likewise contains none of those terms, and its laboratory table has no glucose, calcium, phosphate or PTH row; ALP increased in 41% (grade 3-4, 5%).
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 and same limitation. Denominators are 108-110, smaller than revumenib's, and the population is adults with NPM1-mutant AML only.
Evidence that argues against this (1)
- Ziftomenib's paediatric section states safety and effectiveness in children have not been established, so the absence of paediatric-relevant endocrine measurement is at least internally consistent for that product. Kura Oncology et al. 2025
- supports →A documented absence: not one endocrine outcome measure across the three lead menin-inhibitor trials
Extends the registry absence to the labels: PTH survived into the REVUFORJ table, glucose did not survive into either, despite being on every routine chemistry panel
- related to →The same drug target, pursued for benefit: menin inhibition is a diabetes strategy in the literature
The one endpoint the benefit literature predicts would move is the one endpoint neither label reports; a glucose row would be decisive in either direction and costs nothing
Fills the verified absence: adds the glycaemic panel neither label reports, with the post-treatment draws as the primary comparison