Zucara’s ZT-01: what could taking the brake off glucagon do?

Type 1 diabetes has a glucagon timing problem. Glucagon secretion is poorly matched to the prevailing glucose concentration, insulin availability and nutrient signals.

When glucose falls, the glucagon response that should form one of our primary defences against hypoglycaemia becomes impaired. When glucose is normal or rising, particularly after food, glucagon may be inadequately suppressed and can add hepatic glucose to an existing glucose excursion.^2,6

This combination makes Zucara Therapeutics’ ZT-01 one of the more interesting approaches to hypoglycaemia currently in development.

ZT-01, or axareotide acetate, is a selective somatostatin receptor 2 antagonist. It reduces somatostatin restraint on pancreatic α-cells so that they can release glucagon as glucose falls. A recently published phase 1 study shows that this approach can restore a glucagon response during experimentally induced hypoglycaemia.¹

The same study also recorded a transient glucagon rise during euglycaemia, accompanied by a modest glucose rise and a higher intravenous insulin infusion rate within the clamp.¹ These observations show that the pharmacological effect is not confined to hypoglycaemia. They do not establish that ZT-01 will increase everyday insulin requirements or worsen free-living glucose control.

The central question is whether glucagon can be restored in the context in which it is deficient without causing clinically important effects in contexts where additional hepatic glucose is unnecessary.

Glucagon regulation in type 1 diabetes

Roger Unger and Alan Cherrington’s glucagonocentric model gave glucagon a prominent role in diabetic hyperglycaemia, ketogenesis and catabolism. Its strongest interpretation remains debated, particularly when evidence from experimental models is extended to people with type 1 diabetes. The less controversial point is that glucagon affects hepatic glucose production and ketogenesis, and that its regulation is abnormal in type 1 diabetes.²

Absolute glucagon concentration alone can therefore mislead. Its physiological appropriateness depends on the accompanying glucose concentration, insulin availability and nutrient state.

When glucose rises in somebody without diabetes, glucagon should fall. The β-cell doesn’t merely deliver insulin into the systemic circulation; it exposes neighbouring α-cells to extremely high local concentrations of insulin inside the islet. Together with signals from δ-cells and other intra-islet pathways, this helps tell the α-cell that additional hepatic glucose isn’t required.^2,5

β-cell destruction removes insulin delivery to muscle and liver and disrupts the local signalling network that governs glucagon. Subcutaneous insulin replacement cannot recreate that islet topology.

That helps explain a peculiar feature of T1D: the α-cell can become remarkably bad at doing the right thing in either direction. During hypoglycaemia, glucagon secretion fails. After food, glucagon can do almost the opposite of what is wanted.

A 2026 study investigated that latter phenomenon using labelled glucagon. Sixteen people with T1D and 16 without diabetes underwent a mixed-meal experiment in which stable-isotope-labelled glucagon allowed investigators to distinguish hormone appearance from clearance. Early post-meal glucagon concentrations were higher in T1D, and the difference arose from greater glucagon appearance and turnover rather than slower clearance. The authors concluded that reducing early postprandial glucagon flux could potentially improve postprandial glucose levels in T1D.⁶

α-cells therefore remain active in type 1 diabetes, but their responses are not reliably matched to the glucose context.

α-cell restraint

Somatostatin is part of the ordinary control system governing α-cell secretion.

Human-islet work by Elliott, Ustione and Piston showed that insulin and somatostatin work together to suppress glucagon as glucose rises. Somatostatin acts through SSTR2 to reduce α-cell cAMP production, while insulin promotes cAMP breakdown through a separate pathway. Together they inhibit glucagon secretion. When researchers specifically blocked SSTR2 at high glucose, glucagon secretion increased approximately 2.5-fold in human islets.⁵

The 2025 Pharmacological Reviews synthesis by Kosheleva and colleagues adds another part of the picture: GABA. Its model of islet endocrine signalling shows β-cell insulin, β-cell-derived GABA and δ-cell somatostatin contributing overlapping inhibitory signals to the α-cell. Somatostatin inhibits glucagon, while GABA-A receptor activation hyperpolarises the α-cell and suppresses secretion.⁹

The same review describes evidence that GABA signalling is substantially disrupted in T1D. Under normal circumstances, β-cell-derived insulin promotes trafficking of GABA-A receptors onto the α-cell membrane; extracellular GABA then allows chloride influx, hyperpolarising the α-cell and closing voltage-gated calcium channels required for glucagon secretion.⁹

In a healthy islet, these inhibitory signals can be summarised as:

β-cell insulin + β-cell GABA + δ-cell somatostatin → α-cell restraint

In established type 1 diabetes, important components of that system have already been damaged by β-cell destruction. ZT-01 then deliberately weakens another one:

somatostatin → SSTR2 blockade → reduced α-cell restraint

Reduced restraint may be useful when glucose is 46.8 mg/dL (2.6 mmol/L). The α-cell receiving ZT-01, however, sits within an already disrupted paracrine environment.

Euglycaemic clamp findings

In the human phase 1 experiment, ZT-01 was administered while plasma glucose was held at approximately 90 mg/dL (5.0 mmol/L). Investigators then used intravenous insulin to lower glucose through level 1 hypoglycaemia at 63 mg/dL (3.5 mmol/L) and level 2 hypoglycaemia at 46.8 mg/dL (2.6 mmol/L).¹

During those hypoglycaemic periods, ZT-01 did what it was designed to do. The glucagon response increased, and glucose counterregulatory events occurred substantially more frequently with ZT-01 than with placebo.¹

Before glucose was lowered, mean glucagon increased by 25.7 ± 2.4 ng/L after 3 mg ZT-01 and by 28.4 ± 2.4 ng/L after 20 mg, with values reported as least-squares mean ± standard error. These increments lack a direct physiological or clinical benchmark, so the more informative observation is their pairing with changes in glucose and insulin infusion. The effect was transient, and glucagon moved back towards its pre-dose concentration before hypoglycaemia began.¹

ZT-01 did not produce a sustained increase in glucagon throughout the experiment. The transient rise nevertheless coincided with measurable changes in glucose and insulin infusion.

Thirty minutes after dosing, mean plasma glucose was 86.4 ± 12.6 mg/dL (4.8 ± 0.7 mmol/L) with placebo, 108.0 ± 19.8 mg/dL (6.0 ± 1.1 mmol/L) with 3 mg ZT-01 and 109.8 ± 21.6 mg/dL (6.1 ± 1.2 mmol/L) with 20 mg, reported as mean ± 95 per cent confidence interval. Values of 6.0 to 6.1 mmol/L are not clinically meaningful hyperglycaemia. Their relevance is that glucose rose above the experiment’s narrow euglycaemic target after dosing.

Glucose exceeded the clamp’s 5.5 mmol/L upper target during 17 of 20 clamps after 3 mg and 14 of 18 after 20 mg, compared with 7 of 20 after placebo.

To restore the clamp target, investigators increased the variable intravenous insulin infusion. Mean rates, reported with 95 per cent confidence intervals, were 0.73 ± 0.44 U/h with placebo, 1.38 ± 0.55 U/h with 3 mg and 1.35 ± 0.30 U/h with 20 mg. Both ZT-01 rates differed from placebo at P≤0.0002.¹

The reported mean infusion rates were therefore roughly 85 to 90 per cent higher after ZT-01. This is a clamp-specific adjustment to an intravenous infusion under an artificial protocol, rather than an estimate of a person’s real-world daily insulin requirement. Free-living insulin delivery may respond differently because meals, activity, rescue carbohydrate and automated insulin delivery all alter the glucose trajectory.

Taken together, the paired glucagon, glucose and infusion-rate changes show that ZT-01’s pharmacological action is not intrinsically glucose-gated. The drug does not activate only once glucose has fallen:

glucose becomes low → ZT-01 activates → glucagon appears

Its action might be better represented as:

ZT-01 present → somatostatin restraint weakened → α-cell becomes more permissive to glucagon secretion

The prevailing metabolic context then determines what happens next. At 46.8 mg/dL (2.6 mmol/L), the effect may be useful. At approximately 90 mg/dL (5.0 mmol/L), the clamp shows a transient glucagon rise with a modest glucose rise that required more intravenous insulin to restore the target. The effect after a meal at 180 mg/dL (10 mmol/L) has not been established.

Historical somatostatin experiments

Two small experiments from the 1970s help explain why glucagon matters when insulin availability is limited. Raskin and Unger studied four people with type 1 diabetes receiving continuous insulin infusion. Systemic somatostatin reduced glucagon and was accompanied by lower glucose and glycosuria; replacing glucagon reversed part of that effect.³ Gerich and colleagues withdrew insulin from seven people with type 1 diabetes. During systemic somatostatin infusion, ketoacidosis did not develop within the 18-hour observation period and several metabolic measures were lower than under control conditions.⁴

These experiments were small, short and conducted with systemic somatostatin, which suppresses several hormones and is not equivalent to selective SSTR2 antagonism. They do not show that ZT-01 increases the risk of diabetic ketoacidosis (DKA), and there is currently no evidence that it does. They establish the narrower physiological point that glucagon can influence hepatic glucose production and ketogenesis when insulin is inadequate.

This evidence motivates a safety question rather than a prediction of harm: how does interrupted insulin delivery affect metabolism while SSTR2 antagonism is active? The question is relevant to pump users because interruption of rapid-acting insulin leaves no long-acting basal depot. An insulin-interruption study could answer it directly.

Glucagon-receptor antagonism and GABA

The phase 2 volagidemab study randomised 79 adults with T1D to placebo or an antagonistic monoclonal antibody against the glucagon receptor. Its primary endpoint was the change in total daily insulin use at 12 weeks.

Total daily insulin use changed by -7.59 U in the 35 mg group (95 per cent confidence interval -11.79 to -3.39), -6.64 U in the 70 mg group (95 per cent confidence interval -10.99 to -2.29) and -1.27 U with placebo (95 per cent confidence interval -5.4 to 2.9). Neither active comparison met the prespecified two-sided significance threshold of P<0.025 for the primary endpoint.⁷

Placebo-adjusted HbA1c at week 13 was lower by 0.53 percentage points with 35 mg (95 per cent confidence interval -0.89 to -0.17) and by 0.49 percentage points with 70 mg (95 per cent confidence interval -0.85 to -0.12), with nominal P values. Hypoglycaemia did not increase, but liver transaminases, LDL cholesterol and blood pressure did.⁷ The failed primary endpoint and these safety signals limit the therapeutic inference.

Volagidemab does not provide a mirror-image experiment for ZT-01. Receptor blockade and a transient increase in endogenous glucagon are different interventions, studied in different settings. The trial offers indirect evidence that glucagon signalling can affect glycaemia, while providing no estimate of the clinical effect of SSTR2 antagonism.^7,9

A randomised trial in 97 children with newly diagnosed T1D provides more limited mechanistic support. Oral GABA alone or GABA combined with GAD-alum did not preserve β-cell function, and there were no clinically significant differences in glycaemic control. The combination reduced fasting and meal-stimulated glucagon as a secondary outcome.⁸ This supports the biological link between GABA signalling and glucagon secretion, but it does not show that reducing glucagon improves glycaemic outcomes.

Automated insulin delivery

In somebody using an automated insulin delivery (AID) system, the algorithm cannot observe somatostatin signalling or glucagon. It responds to sensor glucose and its recent trajectory. A transient glucagon-associated glucose rise could therefore prompt additional insulin delivery.

The clamp demonstrates an acute version of this interaction, but it does not reproduce subcutaneous insulin kinetics or an adaptive AID system.¹ If a free-living system delivered more insulin in response to a transient glucose rise, some of that insulin could remain active after the glucagon signal subsided. This is a plausible control-system interaction, not an observed adverse effect.

The opposite counterfactual is equally important. Preventing an impending low may allow an AID system to suspend insulin less often, reduce rescue-carbohydrate use and avoid the glucose rebound that can follow treatment. Total overnight insulin exposure could remain unchanged or fall, while glucose variability and time in range improve.

Repeated-dose data on continuous glucose monitoring, insulin delivery and rescue carbohydrate are needed to distinguish these possible trajectories. There is no evidence that ZT-01 causes rebound hypoglycaemia.

Growth hormone

The phase 1 study also recorded higher growth-hormone concentrations during post-dose euglycaemia and level 1 hypoglycaemia with ZT-01 than with placebo. The investigators attributed this to somatostatin signalling at the anterior pituitary and considered it clinically unimportant in the acute study.¹ The finding confirms that systemic SSTR2 antagonism is not pancreas-specific. There is no evidence at present that this growth-hormone effect has clinically important metabolic consequences.

The ZONE phase 2 study

ZONE is a randomised, double-blind, placebo-controlled phase 2 study in adults with T1D who experience nocturnal hypoglycaemia. Its stated purpose is to determine whether ZT-01 reduces nocturnal hypoglycaemia and to assess its effects on glucose levels.¹⁰

The primary clinical goal is to reduce nocturnal hypoglycaemia below 54 mg/dL (3.0 mmol/L), but interpretation will require more than time below range. A reduction accompanied by stable mean glucose, time in range, time above range, variability and insulin delivery would support useful counterregulation at little metabolic cost.

All participants are restricted from using AID in this trial, so the questions asked here will not be answered.

Conflicts of interest

The phase 1 study was funded by Zucara, which participated in its design, analysis, interpretation and reporting; several authors also had employment, shareholding, patent or advisory relationships with the company.¹ Three authors of the Pharmacological Reviews synthesis disclosed links to Levicure and related patents.⁹ These disclosures matter because commercial interests can influence study questions, analytic choices and the framing of uncertain findings. They do not invalidate the data, but they increase the value of complete reporting and independent replication.

Conclusion

The loss of glucagon counterregulation during hypoglycaemia is an important defect in longstanding T1D. Restoring it could provide another defence against low glucose without rescue carbohydrate or a dual-hormone pump.

The phase 1 evidence is solid within the acute clamp setting: ZT-01 increased glucagon responsiveness during experimentally induced hypoglycaemia. It also caused a transient glucagon rise during euglycaemia, paired with a modest glucose rise and a higher intravenous insulin infusion rate. The clinical importance of that euglycaemic effect remains provisional because the clamp cannot estimate free-living insulin needs or overnight outcomes.¹

Other evidence supports the wider physiological question without settling the clinical one. Tracer experiments show increased early post-meal glucagon appearance in T1D.⁶ The volagidemab trial found nominal HbA1c improvements but missed its primary endpoint and identified safety signals, so it offers only indirect evidence about the metabolic role of glucagon signalling.⁷ The GABA/GAD trial reduced glucagon as a secondary outcome without improving glycaemic control or preserving β-cell function, which makes it mechanistic support rather than evidence of glycaemic benefit.⁸

ZT-01 removes part of the α-cell’s somatostatin restraint. In the context of glucose at 46.8 mg/dL (2.6 mmol/L), that may be beneficial. In the context of a euglycaemic clamp at approximately 90 mg/dL (5.0 mmol/L), it produced a transient glucagon rise, a modest glucose rise and a higher clamp insulin infusion rate. Its effect in the post-meal context at 180 mg/dL (10 mmol/L) is unknown.

ZONE should establish whether bedtime ZT-01 reduces nocturnal hypoglycaemia and show how that benefit relates to mean glucose, time in range, variability, insulin delivery and rescue-carbohydrate use. The central question is whether useful glucagon counterregulation can be restored in the hypoglycaemic context without a clinically important cost in euglycaemic or postprandial contexts.

The treatment may prevent nocturnal lows, but the trial must show that taking the brake off glucagon does not create a new glucose problem.

References

1. Abitbol A, Riddell MC, Peers S, Simonson E, Evans M, Knop FK, Liggins RT. Effect of somatostatin receptor 2 antagonism on glucagon counterregulation during a hyperinsulinaemic to hypoglycaemic glucose clamp in adult men and women with long-standing type 1 diabetes: a randomised crossover phase 1 study. Diabetologia. 2026;69(8):2143 to 2154. doi:10.1007/s00125-026-06748-9.
2. Unger RH, Cherrington AD. Glucagonocentric restructuring of diabetes: a pathophysiologic and therapeutic makeover. J Clin Invest. 2012;122(1):4 to 12. Full text.
3. Raskin P, Unger RH. Hyperglucagonemia and its suppression: importance in the metabolic control of diabetes. N Engl J Med. 1978;299:433 to 436. doi:10.1056/NEJM197808312990901.
4. Gerich JE, Lorenzi M, Bier DM, et al. Prevention of human diabetic ketoacidosis by somatostatin: evidence for an essential role of glucagon. N Engl J Med. 1975;292(19):985 to 989. doi:10.1056/NEJM197505082921901.
5. Elliott AD, Ustione A, Piston DW. Somatostatin and insulin mediate glucose-inhibited glucagon secretion in the pancreatic α-cell by lowering cAMP. Am J Physiol Endocrinol Metab. 2015;308:E130 to E143. doi:10.1152/ajpendo.00344.2014.
6. Ruchi FNU, Schiavon M, Pandey A, et al. Postprandial Glucagon Metabolism in Healthy and Type 1 Diabetes. Diabetes. 2026;75(2):256 to 263. doi:10.2337/db25-0587.
7. Pettus J, Boeder SC, Christiansen MP, et al. Glucagon receptor antagonist volagidemab in type 1 diabetes: a 12-week, randomized, double-blind, phase 2 trial. Nat Med. 2022;28:2092 to 2099. doi:10.1038/s41591-022-02011-x.
8. Martin A, Mick GJ, Choat HM, et al. A randomized trial of oral gamma aminobutyric acid (GABA) or the combination of GABA with glutamic acid decarboxylase (GAD) on pancreatic islet endocrine function in children with newly diagnosed type 1 diabetes. Nat Commun. 2022;13:7928. doi:10.1038/s41467-022-35544-3.
9. Kosheleva L, Koshelev D, Lagunas-Rangel FA, et al. Disease-modifying pharmacological treatments of type 1 diabetes: Molecular mechanisms, target checkpoints, and possible combinatorial treatments. Pharmacological Reviews. 2025;77:100044. doi:10.1016/j.pharmr.2025.100044.
10. ClinicalTrials.gov. A Study of the Effect of ZT-01 on Night-time Hypoglycemia in Type 1 Diabetes (ZONE). NCT05762107. Study record.