LAUSR

BACKGROUND The incidence of cardiovascular and limb-specific adverse outcomes is higher in peripheral arterial disease (PAD) patients with diabetes. Metformin is associated with improved cardiovascular morbidity and mortality. However, the effect of metformin on limb-specific outcomes is unclear. The objective of this study was to assess the effect of metformin on outcomes after intervention for PAD. METHODS Patients who underwent revascularization for chronic limb ischemia (Rutherford 3-6) between June 2001 and December 2014 were retrospectively identified. Primary patency (PP), secondary patency (SP), limb salvage (LS), major adverse limb events (MALE), major adverse cardiac events (MACE), and survival rates were compared using Kaplan-Meier and Cox regression. RESULTS One thousand sixty-four limbs in 1204 patients were identified (147 metformin, 196 other hypoglycemics [OH], 216 insulin, and 645 nondiabetics (nondiabetes mellitus [DM]). Non-DM had significantly lower incidence of CAD (46%) than insulin (65%), metformin (56%), and OH groups (63%) (P < 0.001). Insulin patients (17%) had significantly higher incidence of end-stage renal disease (ESRD) than non-DM (3%), metformin (1.4%), and OH groups (8%) (P < 0.001). Ninety four percent of patients in the metformin group were on aspirin, which was significantly higher than non-DM (86%), OH (83%), and insulin groups (86%) (P = 0.02). Similarly, statin use was significantly higher in the metformin group (71%) than in OH (64%), insulin (61%), and non-DM groups (55%) (P = 0.002). Majority of patients in the insulin group presented with critical limb ischemia (CLI) (93%), which was significantly greater than the metformin (59%), OH (72%), and non-DM groups (50%) (P < 0.001). Sixty-month PP was significantly greater in non-DM group (62%) (P = 0.005) in overall comparison with no significant difference between metformin (56%), OH (60%), and insulin (51%) groups (P = 0.06). Sixty-month SP was similar in metformin (76%), OH (85%), insulin (76%), and non-DM (80%) groups (P = 0.27). LS was significantly worse in insulin group (62%) (P < 0.001) with no significant difference between metformin (84%), OH (83%), and non-DM (87%) groups (P = 0.45). Freedom from MALE at 60 months was 53% in the insulin group, which was significantly worse as compared with metformin (71%), OH (70%), and non-DM (67%) groups (P = 0.001). Sixty-month survival was significantly improved in metformin (60%) and non-DM (60%) groups as compared with that in OH (41%) and insulin groups (30%) (P < 0.001). Freedom from MACE was significantly greater in metformin (44%) and non-DM (52%) groups than that in OH (37%) and insulin groups (25%) (P < 0.001). Metformin use (HR, 0.7 [0.5-0.9]; P = 0.008) was an independent factor associated with freedom from mortality. CONCLUSIONS Metformin is associated with improved survival and decreased incidence of adverse cardiac events in PAD patients. However, it did not have an impact on patency or LS rates after open and endovascular interventions. LS was worse in diabetic patients primarily treated with insulin.