Diabetic ketoacidosis (DKA) is increasingly recognized in adults with "ketone-prone" type 2 diabetes. In many of these patients, the pancreas can still make insulin but becomes temporarily "stunned" during severe, prolonged high blood sugar. Arginine is a naturally occurring amino acid that can trigger the pancreas to release its own insulin when glucose is high. It is FDA-approved for other uses and has been given intravenously for decades with a strong safety record. Whether a single arginine infusion given early during DKA can safely boost the body's insulin and speed recovery has not been tested.

This randomized, double-blind, placebo-controlled, phase 1/2 trial will enroll 60 adults who present to one of four Detroit-area emergency departments with DKA consistent with ketone-prone type 2 diabetes (high glucose and significant ketones). Participants will receive standard DKA care ordered by their clinicians. In addition, under blinded conditions they will receive either arginine hydrochloride 30 grams (in 300 mL) or placebo (normal saline), infused intravenously over 30 minutes as early as feasible after DKA is recognized.

The main question is whether arginine increases endogenous (self-made) insulin soon after infusion. We will measure C-peptide (a marker released in equal amounts with insulin) and glucose at 10, 30, and 90 minutes after the start of the infusion and calculate the C-peptide/glucose ratio. Secondary measures include the rate of ketone (β-hydroxybutyrate) clearance and the total insulin dose required in the first 24 hours. Additional blood tests will examine arginine and related amino acids, and a small sample of platelets will be used to explore mitochondrial function. Safety will be closely monitored during and after the infusion, and participants will be contacted at 90 days to assess for any delayed problems.

Potential risks include temporary flushing, nausea, or headache; the infusion can be stopped at any time if needed. Potential benefits include faster resolution of ketosis and reduced insulin needs, but benefits cannot be guaranteed for individual participants.

Background and Rationale. A substantial subset of adults with type 2 diabetes develop DKA ("ketone-prone" type 2 diabetes). Many of these patients lack diabetes autoantibodies (A-) and retain recoverable β-cell function (β+). The pathophysiology appears to involve reversible β-cell dysfunction from prolonged hyperglycemia. Arginine, a cationic amino acid and physiologic insulin secretagogue, depolarizes β-cells and enhances glucose-dependent insulin release within minutes. Prior human studies show prompt rises in insulin and C-peptide after arginine in individuals with preserved β-cell reserve. Arginine may also improve nitric-oxide bioavailability and mitigate mitochondrial dysfunction-processes implicated in hyperglycemic crises. We therefore hypothesize that early intravenous arginine, added to usual DKA care, will acutely increase endogenous insulin secretion and speed clearance of ketosis without added safety concerns.

Design and Setting. Prospective, phase 1/2, randomized, double-blind, placebo-controlled, parallel-group trial at four emergency departments in the Detroit metropolitan area. Target enrollment is 60 participants (1:1 allocation).

Participants. Adults (>17 years) with hyperglycemia (generally ≥250 mg/dL) and significant ketonemia/ketosis consistent with DKA and a clinical phenotype of ketone-prone type 2 diabetes. Key exclusions include known type 1 diabetes or positive GAD65 antibodies, chronic dialysis, cirrhosis, pregnancy, allergy to arginine, or features of moderate/greater alcohol intoxication. Screening may use point-of-care capillary β-hydroxybutyrate (BHB) or breath acetone to expedite identification; confirmatory laboratory thresholds will be used per protocol.

Intervention. As early as logistics allow after recognition of DKA, participants will receive a blinded 30-minute intravenous infusion of either arginine hydrochloride 30 g in 300 mL (R-Gene® 10) or matching placebo (normal saline 500 mL). Treating teams will manage DKA per standard institutional protocols; the study does not restrict clinical care (fluids, insulin, electrolytes).

Outcomes.

Primary endpoint: Endogenous insulin secretion quantified by the C-peptide/glucose ratio at 10, 30 (end of infusion), and 90 minutes after infusion start.

Key secondary endpoints: (a) Rate of BHB decline (ketone clearance), and (b) total insulin administered in the first 24 hours.

Mechanistic/biomarker endpoints: Plasma amino acids-especially arginine, citrulline, and ornithine-to compute the Global Arginine Bioavailability Ratio (arginine/[ornithine+citrulline]) at baseline and 90 minutes; insulin, proinsulin, and diabetes autoantibodies. Exploratory endpoint: Platelet mitochondrial complex IV and V activities as markers of oxidative phosphorylation.

Safety Monitoring. Participants will be observed closely during and after infusion with serial vital signs and symptom checks (e.g., flushing, nausea, headache). Pre-specified stop criteria allow immediate interruption of the infusion. Adverse events will be assessed to 90 days (telephone/medical record review) and reported per regulatory guidance. Because of the small size and favorable safety profile of intravenous arginine, the study qualifies as early-phase with streamlined oversight; an independent Data and Safety Monitoring Board has nonetheless been chartered.

Sample Size and Analysis. Based on prior insulinotropic responses to arginine, we powered the study to detect a moderate-to-large effect (standardized mean difference ~0.8) in the primary endpoint with 80% power and α=0.05, yielding ~26 per group. Allowing for attrition, we plan to enroll 60 participants. The C-peptide/glucose ratio will be assessed for normality and compared between groups with appropriate parametric or transformed analyses; ketone clearance and insulin dose will be analyzed similarly. Prespecified sensitivity analyses will account for baseline severity and timing of insulin initiation.

Operational Notes. To facilitate early enrollment, research staff will screen patients with suspected DKA using point-of-care BHB or breath acetone while confirmatory labs are pending. Blood draws for study assays occur at baseline (pre-infusion), 10, 30, and 90 minutes, with additional clinically obtained labs used for outcomes (e.g., BHB over the first 24 hours). Platelet-rich plasma will be stored for exploratory mitochondrial testing using validated methods.

Impact. If early arginine infusion safely augments endogenous insulin and accelerates ketone clearance, this pragmatic, low-cost, readily available therapy could reduce reliance on prolonged insulin infusions, shorten ICU/ED resource use, and improve patient experience during DKA. Findings will inform the design of a larger phase 2/3 trial focused on clinical effectiveness and health-system outcomes in ketone-prone type 2 diabetes.

Inclusion Criteria:

• Age >17 years.
• Unscheduled presentation to a participating emergency department with hyperglycemia (serum glucose >250 mg/dL) and significant ketonemia consistent with DKA, defined as laboratory serum/plasma β-hydroxybutyrate (BHB) >20 mg/dL (≈≥1.9 mmol/L).

Note: point-of-care capillary BHB ≥1.5 mmol/L and/or breath acetone ≥0.01% may be used for screening while confirmatory labs are pending; if confirmatory BHB ≤20 mg/dL, the participant is a screen failure.

• Clinical phenotype consistent with ketosis-prone type 2 diabetes (no known prior diagnosis of type 1 diabetes).
• Able to provide written informed consent and comply with study procedures in the ED.

Exclusion Criteria:

• Current renal replacement therapy for chronic kidney disease (hemodialysis or peritoneal dialysis).
• Known history of type 1 diabetes mellitus or known GAD65 autoantibody positivity.
• Diagnosed cirrhosis/advanced chronic liver disease.
• Pregnancy (known pregnancy or positive test at screening).
• Known allergy or hypersensitivity to arginine or its components.
• Features of at least moderate acute alcohol intoxication at screening, per treating team.

We will share de-identified individual participant data (IPD) underlying the primary and secondary outcomes, including: demographics (age, sex), baseline clinical values, all C-peptide and glucose measurements at 0/10/30/90 minutes, β-hydroxybutyrate values used to derive ketone-clearance metrics, total insulin administered in the first 24 hours, adverse events, and (when available) amino-acid measures and platelet mitochondrial assays. A data dictionary will accompany the dataset. All IPD will be de-identified consistent with HIPAA Safe Harbor (direct identifiers removed; dates converted to study day/time; site identifiers suppressed).

Beginning after publication of the primary results manuscript (or within 24 months of last-patient last-visit, whichever occurs first) and available for at least 5 years thereafter.

Access is provided upon reasonable request to the Study Contact listed on this record. Requestors must submit a brief proposal with analysis aims and statistical plan and provide evidence of IRB/ethics review or exemption. A Data Use Agreement is required (no re-identification, no onward sharing, secure storage, and acknowledgment of the source in resulting works). Approved requests will receive the de-identified dataset and data dictionary via secure transfer. Aggregated results will also be made available on reasonable request.

Quadruple-blind with centralized 1:1 allocation concealment. A site research pharmacist (unblinded) prepares study drug; infusion containers are covered to mask volume/appearance and both arms are run over 30 minutes using identical pump settings. Participants, bedside clinicians, investigators, and outcomes assessors remain blinded through database lock. Emergency unblinding is permitted for medical necessity via pharmacy access to allocation codes; all unblindings are logged. Laboratory personnel analyze specimens using coded IDs.