Ketone Monitoring in T1D: Effect of SGLT2i During Usual Care and With Insulin Deficiency

The study will consist of 4 distinct segments that include routine testing of ketones during usual care, during insulin deficiency, and while taking a SGLT2i in addition to usual care, then with insulin deficiency after the SGLT2i has reached steady state and ingested that morning, in a crossover design. We will study 20 adults with T1D who are using Dexcom G6 continuous glucose monitor (CGM), plus any form of insulin delivery. Simultaneous ketone monitoring using a the Biosense® breath ketone analyzer (BKA) and capillary blood beta hydroxybutyrate (BOHB) measurement will be done 2-3 times daily for 2 weeks during each of the outpatient segments. This will generate > 800 data points comparing breath acetone, measured in units called ACEs, versus capillary blood BOHB measurements for each 2 week period. On days of insulin withdrawal, which will take place on the Washington University Intensive Research Unit (IRU), frequent monitoring with capillary blood measurements of BOHB (Precision Xtra®, Abbott) will be compared with breath BrAce measurements (Biosense®, Readout) and laboratory measurements of BOHB and acetoacetate. Electrolytes and glucose will also be measured.

Inclusion criteria

• Type 1 diabetes for >1 year
• Age 18 - 75, any gender, race or ethnicity
• HbA1c <10%
• Stable insulin delivery method for the past 30 days
• Vision of 20/40 or better, including ability to read all device instructions and insulin pump settings
• Use of an insulin pump and ability to make adjustments to pump settings
• Insulin delivery by MDI with basal insulin (glargine U100) given in the morning OR willingness to transition to basal glargine given in the morning for at least 48 hours prior to insulin withdrawal visits
• Use of personal CGM, only Dexcom G6 will be permitted for data consistency
• Use of cellular phone (iOS11 and above or Android) with data capability with connectivity to Dexcom Clarity app and Biosense Ketone Monitoring app
• Willing to share CGM, capillary ketone results and BrACE data with the study group
• Able to understand the study requirements, risks and benefits and able to provide written informed consent
• Able to schedule the visits and perform study related tasks

Exclusion criteria

• History of DKA in the past 6 months, or more than 1 episode of DKA in the past 2 years
• Use of insulin degludec, insulin glargine U300 or insulin detemir as basal insulin and unwilling to transition to glargine given in the morning for at least 48 hours before the insulin withdrawal visits
• Use of an SGLT2i in the past 30 days or intolerance to SGLT2i use in the past for any reason
• Unstable heart disease, including hospitalization for any cardiac or vascular event in the past 6 months.
• eGFR <30 ml/min/1.73m2 or hemoglobin <11.0 g/dL measured in the past year
• Cancer that has been under treatment in the past 6 months, or treatment is likely in the 3 months after signing the consent, not including skin cancer or cancers under long-term hormonal reduction treatment (breast or prostate, no other active treatment)
• Psychiatric condition that interferes with daily activities or diabetes self-care, including substance abuse
• Any illness or condition that may interfere with study measurements, or contraindications for use of SGLT2i, based on investigator discretion (hemoglobinopathy, orthostatic hypotension)
• Unwilling to avoid alcohol during the active study periods, specifically to avoid alcohol for 4 hours prior to any breath ketone measurement
• Currently following or planning to adopt a diet that is low in carbohydrates (defined as <90 grams of carbohydrate per day) or is expected to be ketogenic
• Use of oral or injected corticosteroids within the past 30 days or planned during the study period
• Taking disulfiram (due to interference with breath ketone measurements)
• History of vomiting episodes for any reason in the past 30 days, or hospitalization for cyclic vomiting in the past year
• History of urinary tract infection in the past 3 months
• Pregnancy, breast-feeding or intention of becoming pregnant during the study time period and up to 30 days after study completion
• Pre-menopausal women not using acceptable birth control defined as abstinence, surgical sterilization, hysterectomy, hormonal contraception, IUD or proven effective barrier methods