IGlarLixi will be started at 10 dose steps (10 U insulin glargine plus 5 mcg lixisenatide, once daily). On the basis of prebreakfast self-monitored blood glucose measurements, doses of IGlarLixi will be titrated individually once per week to achieve a prebreakfast plasma glucose of 80-130 mg/dL by use of an algorithm (adding 2 dose steps for prebreakfast plasma glucose >130 mg/dL; no dose change for prebreakfast plasma glucose of 80-130 mg/dL; reducing 2 dose steps for prebreakfast plasma glucose < 80 mg/dL). The daily dose of IGlarLixi could be titrated to 60 dose steps (60 U insulin degludec plus 20 mcg lixisenatide).

Patients in this arm will continue the basal insulin used before the randomization, with dosage titration on the basis of the following algorithm: adding 2 units for prebreakfast plasma glucose >130 mg/dL; no dose change for prebreakfast plasma glucose of 80-130 mg/dL; reducing 2 units for prebreakfast plasma glucose < 80 mg/dL. Moreover, they will be assigned to canaglifozin, according to the current clinical practice and the drugs' data sheet. Canagliflozin will be started at 100 mg daily per oral administration, and augmented to 300 mg/per day if required (HbA1c >7.5 after 12 weeks).

Patients in this arm will continue the basal insulin used before the randomization, with dosage titration on the basis of the following algorithm: adding 2 units for prebreakfast plasma glucose >130 mg/dL; no dose change for prebreakfast plasma glucose of 80-130 mg/dL; reducing 2 units for prebreakfast plasma glucose < 80 mg/dL. Moreover, they will be assigned to dapaglifozin, according to the current clinical practice and the drugs' data sheet. Dapagliflozin will be started at 10 mg daily per oral administration

Patients in this arm will continue the basal insulin used before the randomization, with dosage titration on the basis of the following algorithm: adding 2 units for prebreakfast plasma glucose >130 mg/dL; no dose change for prebreakfast plasma glucose of 80-130 mg/dL; reducing 2 units for prebreakfast plasma glucose < 80 mg/dL. Moreover, they will be assigned to empaglifozin, according to the current clinical practice and the drugs' data sheet. Empagliflozin will be started at 10 mg daily per oral administration, and augmented to 25 mg/per day if required (HbA1c >7.5 after 12 weeks).

Patients in this arm will continue the basal insulin (glargine, degludec or glargine-300) used before the randomization. The insulin titration will be guided by the medical staff, according to the following algorithm: adding 2 units of basal insulin for prebreakfast plasma glucose >130 mg/dL; no dose change for prebreakfast plasma glucose of 80-130 mg/dL; reducing 2 units of basal insulin for prebreakfast plasma glucose < 80 mg/dL. The short acting insulin analogue (lispro, aspart or glulisine) will be started at the dosage of 4 units before meals (3 times per day) and will be titrated twice a week until achieving pre-prandial glucose values ranging from 80-130 mg/dL.

Difference between groups in total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides

In order to measure satisfaction with diabetes treatment regimens, we used the self-reported Diabetes Treatment Satisfaction Questionnaire. This instrument aims to assess levels of satisfaction in subjects using different treatment strategies. The questionnaire consists of eight questions: six questions addresses general satisfaction with a score from 0 to 6 for each question (0 = worst), that has to be computed in a total score ranging from 0 (=worst) to 36 (=best); among the remaining two questions, which has to be computed separately as two subscales, one concerns the perception of hyperglycemic events and another the perception of hypoglycemic events, both with a score from 0 (none of the time) to 6 (most of the time).