Revascularization Outcomes Stratified by Glycemic Control in Patients with Diabetes Mellitus and Chronic Limb-Threatening Ischemia

Abstract

The prevalence of chronic limb-threatening ischemia (CLTI) has increased alongside rising rates of diabetes mellitus (DM). While diabetic patients with CLTI have worse outcomes compared to patients without diabetes, conflicting data exist on the relationship between the severity of DM and CLTI outcomes. Close inspection of the relationship between DM severity and outcomes in CLTI may benefit surgical decision-making and patient education. We retrospectively reviewed patients who received endovascular intervention or surgical bypass for CLTI at our multidisciplinary Limb Preservation Program from 2013 to 2019 to collect patient characteristics using Society for Vascular Surgery (SVS) reporting standards, arterial lesion characteristics from recorded angiograms, and outcomes, including survival, amputation, wound healing, and revascularization patency. Controlled DM was defined as SVS Grade 1 (controlled, not requiring insulin) and Grade 2 (controlled, requiring insulin), while uncontrolled DM was defined as SVS Grade 3 (uncontrolled), and DM severity was assessed using preoperative hemoglobin A1c (HgbA1c) values. Product-limit Kaplan-Meier was used to estimate survival functions. Univariable Cox proportional hazards analyses guided variable selection for multivariable analyses. Our Limb Preservation Program treated 177 limbs from 141 patients with DM. Patients with uncontrolled DM were younger (60.44±10.67 vs. 65.93±10.89years old, P=0.0009) and had higher HgbA1c values (8.97±1.85% vs. 6.79±1.10%, P<0.0001). Fewer patients with uncontrolled DM were on dialysis compared to patients with controlled DM (15.6% vs. 30.9%, P=0.0278). By Kaplan-Meier analysis, DM control did not affect time to mortality, limb salvage, wound healing, or loss of patency. However, multivariable proportional hazards analysis demonstrated increased risk of limb loss in patients with increasing HgbA1C (hazard ratio (HR)=1.96 [1.42-2.80], P<0.0001) or dialysis dependence (HR=15.37 [3.44-68.73], P=0.0003), increased risk of death in patients with worsening pulmonary status (HR=1.70 [1.20-2.39], P=0.0026), and increased risk of delayed wound healing in patients who are male (HR=0.48 [0.29-0.79], P=0.0495). No independent association existed between loss of patency with any of the variables we collected. Patients with uncontrolled DM, as defined by SVS reporting standards, do not have worse outcomes following revascularization for CLTI compared to patients with controlled DM. However, increasing HgbA1c is associated with a greater risk for early amputation. Before revascularization, specific attention to the level of glycemic control in patients with DM is important, even if DM is "controlled." In addition to aggressive attempts at improved glycemic control, those with elevated HgbA1c should receive careful education regarding their increased risk of amputation despite revascularization. Future work is necessary to incorporate the severity of DM into risk models of revascularization for the CLTI population.