Insulin use is associated with poor limb salvage and survival in diabetic patients with chronic limb ischemia

The goal was to compare the outcomes in patients with disabling claudication (DC) or critical limb ischemia (CLI) to determine if diabetics (DM) have poorer patency, limb salvage (LS), and survival rates than nondiabetic patients and if the diabetic regimen affects these outcomes.

All patients who presented with DC or CLI between June 2001 and September 2008 were included. Non-DM patients were compared with those with DM who are currently managed by diet only or oral medications (D-OM), oral medications plus insulin (OM+INS), or insulin alone (INS).

Of the 746 patients (886 limbs), there were 406 patients (464 limbs) in non-DM, 96 patients (135 limbs) in D-OM, 98 patients (118 limbs) in OM+INS, and 146 patients (185 limbs) in INS groups. There were more patients with coronary artery disease, hypertension, and renal insufficiency in the DM group than non-DM, with the INS group having the highest incidence of renal insufficiency/dialysis (46%/20%). Gangrene and foot sepsis were significantly more frequent in patients in OM+INS (45%/3%) and INS (50%/6%) than non-DM (15%/0.2%) and D-OM groups (25%/1%; P < .001). More patients in the INS group (14%) and OM+INS (9%) had primary amputation than non-DM (4%) and D-OM (4%; P < .01). Mean follow-up was 26.3 +/- 20.7 months. Overall survival following revascularization was similar in D-OM and non-DM and OM+INS and INS, the latter being significantly worse (P < .001). The LS rate in D-OM and non-DM was also identical, whereas OM-INS and INS had significantly worse LS, with OM-INS marginally better than INS (P = .094). Primary patency (PP) was worse in endovascular-treated patients on insulin than non-DM and D-OM patients (P < .001), whereas PP was similar between groups in open-treated patients. Multivariate analysis showed that coronary artery disease, renal insufficiency, chronic obstructive pulmonary disease, indication for intervention, insulin use, nonambulatory status, and statin drug non-use were independently associated with decreased survival, whereas insulin use, presence of gangrene, need for infrapopliteal interventions, and nonambulatory status were independently associated with limb loss. TransAtlantic Inter-Society Consensus (TASC) classification of the treated lesions being C or D, infrapopliteal interventions, and indication of intervention (DC vs CLI) were independently associated with primary patency, whereas insulin use was not.

Diabetic patients who present with limb ischemia can be subdivided into three distinct subgroups based on their diabetic regimen. The survival and LS rates of those controlled with diet or OM are nearly identical to nondiabetics, both of which are significantly better than OM+INS or INS. The PP rate in endovascular-treated patients is worse in patients who are on insulin. Being on insulin is independently associated with decreased survival and limb loss but not PP.