Evaluation of Lower Extremity Calcium Score as a Measure of Peripheral Arterial Disease Burden and Amputation Risk.

ABSTRACT

BACKGROUND: The ankle-brachial pressure index (ABI) and toe-brachial pressure index (TBI) are commonly used diagnostic tools for peripheral artery disease (PAD) that are unreliable in the presence of calcified vessels. In this study, we aimed to demonstrate the utility of the lower extremity calcium score (LECS) in addition to ABI and TBI in measuring disease burden and predicting the risk of amputation in patients with PAD. METHODS: Patients who were evaluated in the vascular surgery clinic at Emory University for PAD and who underwent noncontrast computed tomography of the aorta and lower extremities were included in the study. Aortoiliac, femoral-popliteal, and tibial calcium scores were measured using the Agatston method. ABI and TBI that were obtained within 6 months of the computed tomography scan were noted and divided into categories of PAD severity. Associations between ABI, TBI, and LECS of each anatomic segment were evaluated. Univariate and multivariate ordinal regression analyses were performed to predict the outcome of amputation. Receiver operating characteristic analysis was performed to compare LECS with other variables in its ability to predict amputation. RESULTS: Fifty patients included in the study cohort were divided into LECS quartiles, with 12-13 patients in each quartile. The highest quartile tended to be older (P = 0.016), had a higher percentage of diabetics (P = 0.034), and had a higher frequency of major amputations (P = 0.004) compared to the other quartiles. Patients in the highest quartile of tibial calcium score were more likely to have stage 3 chronic kidney disease (CKD) or greater (P = 0.011) and also had a higher frequency of amputation (P < 0.005) and mortality (P = 0.041). We found no significant association between each anatomic LECS and ABI/TBI categories. On univariate analysis, CKD (Odds Ratio [OR] 12.92 (95% CI 2.01 to 82.83), P = 0.007), diabetes mellitus (OR 5.47 (95% CI 1.27 to 23.64), P = 0.023), tibial calcium score (OR 6.62 (95% CI 1.79 to 24.54), P = 0.005), and total bilateral calcium score (OR 6.32 (95% CI 1.18 to 33.78), P = 0.031) were associated with increased risk of amputation. On multivariate stepwise ordinal regression, TBI and tibial calcium score were identified as important predictors of amputation, with hyperlipidemia and CKD increasing the overall prediction of the model. On Receiver operating characteristic analysis, the addition of the tibial calcium score (area under the curve 0.94, standard error 0.048) significantly improved the prediction of amputation compared to hyperlipidemia, CKD, and TBI alone (area under the curve 0.82, standard error 0.071, P = 0.022). CONCLUSIONS: The addition of tibial calcium score to other known PAD risk factors may improve the prediction of amputation in patients with PAD.