In patients with type 2 diabetes inadequately controlled with metformin, two main therapeutic options are equally plausible: addition of a sulphonylurea (SU) or addition of a thiazolidinedione (TZD). Since the two classes of drugs clearly differ in mechanisms of action, side effects, economic costs and cardiovascular (CV) risk factors profile, a direct comparison of the two strategies would be most appropriate. Aims: 1) To evaluate the effects of addition of pioglitazone as compared with a SU on the incidence of CV events in type 2 diabetic patients inadequately controlled with metformin; 2) To compare the two treatments in terms of glycemic control, safety, and economic costs. Methods: Randomised, open label, parallel group trial of at least 48 months duration involving several Clinical Units throughout Italy. After a 8-week run-in period, participants will be randomized to add-on either a SU: glibenclamide (5-15 mg/die), gliclazide (30-120 mg/die), glimepiride (2-6 mg/die), chosen according to local practice, or: pioglitazone (15-45 mg/die), to metformin (2 gr/die). A HbA1c value > 8.0 % on two consecutive occasions will lead to addition of insulin to ongoing oral therapy.

Primary efficacy outcome: a composite endpoint of all-cause mortality plus non fatal MI (including silent MI), non fatal stroke, and unplanned coronary revascularization.

Secondary outcomes: Principal secondary outcome: a composite ischemic endpoint of sudden death, fatal and non fatal acute MI (including silent MI), fatal and non fatal stroke, major amputations (above ankle), endovascular or surgical intervention on the coronary, leg or carotid arteries, Other secondary outcomes: - a composite CV end point of the primary end point above plus heart failure, endovascular or surgical intervention on the coronary, leg or carotid arteries, silent MI, angina, intermittent claudication with an ankle/brachial index lower than 0.90; a microvascular endpoint including: plasma creatinine increase of 2 times above the baseline value or creatinine clearance reduction of 20ml/min/1. 73m2 or development of microalbuminuria or overt nephropathy (dialysis or plasma creatinine >3,3 mg/dl); glycemic control (changes from baseline in HBA1c, time to failure of glycemic control, i.e., HBA1c >8.0% on two consecutive occasions three months apart); major cardiovascular risk factors (lipids, blood pressure, microalbuminuria, inflammation markers, waist circumference). Data regarding CV endpoints, safety, tolerability, and study conduct will be monitored and analyzed by an independent committee, and will be not available to the study investigators until the closing of data collection. Efficacy end points will be analysed on an intention-to-treat basis.