Arteriographic Patterns of Atherosclerosis and the Association between Diabetes Mellitus and Ethnicity in Chronic Critical Limb Ischemia.

ABSTRACT

BACKGROUND: Disparate outcomes in critical limb ischemia (CLI) persist between ethnicities. The contribution of modifiable factors versus intrinsic biologic differences remains unclear. Hence, we aimed to quantify the associations between ethnicity and anatomic patterns of arterial occlusive disease in CLI, adjusting for known atherosclerotic risk factors. METHODS: We performed a retrospective, single-center review of consecutive patients presenting to the vascular surgery service with CLI. Arterial lesions were defined by location (aortoiliac = aorta and iliac arteries; femoral = common, profunda, and superficial femoral arteries; and popliteal-tibial = infrapopliteal and tibial arteries). Stenoses ≥50% were deemed hemodynamically significant. Associations between the patients' baseline arteriographic patterns, demographics, and medical comorbidities were defined using Kruskal-Wallis, χ2, and Mantel-Haenszel χ2 tests. RESULTS: Between August 2010 and January 2014, 286 CLI patients (n = 172 male, n = 176 tissue loss) were evaluated by the Vascular Surgery service. Two hundred seventy subjects had baseline arteriograms for analysis (black n = 134, 50%; Hispanic n = 78, 29%; Caucasian n = 58, 21%.) All ethnicities presented most frequently with simultaneous disease in all infrainguinal segments (n = 124, 46%). Of Hispanics, 30% (n = 23) presented with isolated infrapopliteal disease, which was higher than any other ethnic group (P = 0.02, χ2). Caucasians (n = 8, 14%) presented more frequently with isolated aortoiliac occlusive disease than either Hispanics (n = 0, 0%) or blacks (n = 2, 1%; P = 0.06). Diabetes mellitus was most prevalent among Hispanics (n = 72, 85%) relative to blacks (n = 77, 55%) and Caucasians (n = 32, 52%; P < 0.001, χ2). Median hemoglobin A1c (HbA1c) was also highest among Hispanics (7.3%, interquartile range [IQR] 6.2-9.9) versus blacks and Caucasians (6.6%, IQR 5.8-8.2 and 6.0%, IQR 5.6-7.6; P = 0.002, Kruskal-Wallis). Tobacco abuse was most frequent among Caucasians (n = 53, 87%) and blacks (n = 113, 81%). Forty-eight (57%) of Hispanics abused tobacco (P = 0.001, χ2.) Subgroup analysis of subjects stratified by baseline HbA1c revealed that there was no relationship between ethnicity and isolated infrapopliteal disease among subjects with HbA1c ≤8.8% (P = 0.58, Mantel-Haenszel χ2). Conversely, patients with poorer glycemic control (HbA1c ≥ 8.9%) were more frequently Hispanic and had a higher probability of having isolated infrapopliteal disease (P = 0.005, Mantel-Haenszel χ2). CONCLUSIONS: Hispanic patients present more frequently with isolated infrapopliteal arterial disease relative to other ethnicities, which may contribute to disparate CLI outcomes. Isolated infrapopliteal disease appears to be driven mostly be poorer glycemic control rather than inherent biologic differences between ethnicities. Future studies aimed at understanding disparate outcomes due to race after lower extremity revascularization may benefit from stratification by the severity of diabetes mellitus. Understanding the distribution of atherosclerotic disease may improve the ability to predict outcomes in limb-threatening ischemia.