A closed-loop insulin system, also referred to as the "artificial pancreas" (AP), is made up of an insulin pump, a continuous glucose monitor, and an application communicating between the two to adjust insulin administration based on glucose control. This is meant for the treatment of type 1 diabetes. The McGill Artificial Pancreas (MAP) has been used previously in type 1 diabetes with significant benefits. Though prior studies have shown significant benefit with this system, some challenges still exist.

Empagliflozin is used in type 2 diabetes; it allows for glucose to be removed through the urine. Though its use is not approved in type 1 diabetes in North America, it (along with similar drugs) has been used in studies as adjunctive therapy with insulin with benefits on blood sugar control.

The purpose of our study is to see if a small dose of empagliflozin (2.5 mg and 5 mg) is enough to help those who cannot achieve adequate glucose control on a closed-loop insulin system.

The primary hypotheses of the study are the following:

1. The use of empagliflozin 2.5 mg daily will increase time in range compared to placebo for those on the closed-loop system.
2. The use of empagliflozin 5 mg daily will increase time in range compared to placebo for those on the closed-loop system.

Some of the novel advances for type 1 diabetes treatment include continuous subcutaneous insulin pump therapy (CSII) and continuous glucose monitoring (CGM). A step further is the combination of both technologies, i.e. closed-loop insulin system, also referred to as the "artificial pancreas" (AP), where insulin delivery is modified based on glucose readings from CGM. Multiple studies have shown the benefit of closed-loop insulin systems on glycemic control, with their strength being nocturnal euglycemia, while post-prandial hyperglycemia remains a challenge due to decreased insulin absorption.

Novel medication therapies previously used for type 2 diabetes are also under investigation for type 1 diabetes. Sodium glucose-linked cotransporter inhibitors (SGLT2i's) increase urinary glucose excretion and, in doing so, euglycemia. Their use is predominantly in type 2 diabetes for glucose control and for cardiovascular and renal protection, though various studies have demonstrated improved glycemic control with SGLT2i use as adjunct to insulin therapy in the treatment of type 1 diabetes. The use of both closed-loop insulin therapy in conjunction with SGLT2i therapy is thus a novel treatment for optimizing glycemic control in type 1 diabetes, in particular those who do not yet reach target glucose levels despite being on a closed-loop system. Our research group has previously studied the impact of a hybrid closed-loop insulin system with empagliflozin 25 mg, which increased the time in target glucose range of 3.9 to 10.0 from 70% to 84%. Notably, the EASE-3 (Empagliflozin as Adjunctive to inSulin thErapy) trial had previously shown a beneficial glucose effect with empagliflozin 2.5 mg without increasing the risk of diabetic ketoacidosis.

The objective of the study is to assess the effectiveness of low-dose empagliflozin (2.5 mg and 5 mg) on glucose control in adult patients with type 1 diabetes on a closed-loop insulin pump system, who would not otherwise be within a percentage of time in range (TIR) by blood glucose of 3.9 to 10 mmol/L of > 70% on the system.

Hypothesis

The primary hypotheses of the study are the following:

1. The use of empagliflozin 2.5 mg daily will increase time in range compared to placebo for those on the closed-loop system.
2. The use of empagliflozin 5 mg daily will increase time in range compared to placebo for those on the closed-loop system.

There will also be secondary hypotheses, based on quality of life measures, the effect of both doses of empagliflozin on time spent in various ranges of blood glucose, average total daily insulin dose, and average change in ketone levels. It is hypothesized that either dose will increase the time spent in a favourable glucose range, with less time spent hypo- and hyper-glycemic, with improved quality of life measures and lower total daily doses of insulin without increase in ketone levels.

Methods This is a three-way, randomized, cross-over double-blinded outpatient trial to compare the effect of glycemic control for adult participants with type 1 diabetes taking placebo, empagliflozin 2.5 mg daily, and empagliflozin 5 mg daily while using a hybrid closed-loop insulin system (i.e. the McGill Artificial Pancreas (MAP), used in over 20 studies thus far). Adult participants meeting inclusion criteria after baseline laboratory investigations and training will use the MAP during a run-in period of 14 days to identify those with TIR < 70%. The chosen participants will have a randomized sequence of interventions (placebo, empagliflozin 2.5 mg, or 5 mg), each intervention lasting 14 days, with a washout period of 7 - 21 days in between. Optimization of MAP insulin parameters will be adjusted on approximately Day 4 so that the last 10 days of CGM with each intervention will be used for data analysis. During each intervention, ketone levels will be measured daily. Questionnaires assessing quality of life will be filled out after each intervention. There will be a visit at the end of the study to assess participant satisfaction. Data analysis will be conducted to compare the differences in each intervention for each participant of the following : CGM data of the last 10 days of each intervention, daily ketone levels, total daily dose of insulin, and scores of quality of life questionnaires.

Inclusion Criteria:

• A clinical diagnosis of type 1 diabetes for at least one year, as per their treating diabetes physician in agreement with the primary investigator's clinical judgment (confirmatory C-peptide and antibodies will not be required)
• HbA1c between 7 and 10.5% (both extremes inclusive), performed within the last 6 months prior to study inclusion
• Insulin pump use (of any modality) for minimum 3 months
• Agreement to the use of highly effective method of birth control in women of child-bearing age and active avoidance of pregnancy during the trial. Child-bearing potential refers to participants of the female sex post-menarche and have not reached menopause or have a disclosed medical condition causing sterility (e.g. hysterectomy). Post-menopausal state refers to the absence of menses for 12 months without any alternative cause.
• Time in range (3.9 - 10.0 mmol/L) < 70% as per CGM readings of the last 10 days during a 2-week run-in period on the closed-loop insulin delivery system.

Exclusion Criteria:

• Current or < 2 week use of any other anti-hyperglycemic agent other than insulin
• Current or \