Aortic Stiffness in a Mortality Risk Calculator for Kidney Transplant Recipients

BACKGROUND: The association between aortic stiffness and all-cause mortality in kidney transplant recipients (KTRs) is uncertain, and aortic stiffness has not yet been incorporated into risk prediction tools. METHODS: During 2007 to 2012, we measured carotid-femoral pulse wave velocity (PWV; SphygmoCor apparatus) 8 weeks after transplantation. The association between PWV and mortality was assessed in a Cox regression analysis adjusting for seven risk factors from a previously validated model. Internal validation was performed by bootstrap resampling, and discrimination and overfitting evaluated by Harrell's C and the calibration slope. RESULTS: Of 1497 KTRs, 1040 (69%) had a valid PWV measurement. During a median follow-up of 4.2 years, 82 patients died. The association between PWV and mortality showed a ceiling effect, and PWV was truncated at 12 m/sec. Each 1 m/sec increase in PWV, up to 12 m/sec, was associated with mortality, hazard ratio (HR) 1.36 (95% CI, 1.14-1.62; P = 0.001). An interquartile range increase (3.8 m/sec) tripled the hazard of mortality, HR, 3.21 (95% CI, 1.63-6.31), similar to the effect of being approximately 20 years older (interquartile range increase (21.6 years); HR, 3.06 [95% CI, 1.87-5.29]). The PWV improved model discrimination with an increase in Harrell's C from 0.76 to 0.78; C difference, 0.024 (95% CI, 0.005-0.043; P = 0.01). Overfitting was moderate with a calibration slope of 0.89, and the final model was adjusted accordingly. A spreadsheet version is presented to estimate expected 5-year survival. CONCLUSIONS: The PWV is a strong risk factor for mortality in KTRs.