Modeling the implications of changes in vascular calcification in patients on hemodialysis

Abstract

The Treat-to-Goal Study found that sevelamer slowed the progression of coronary calcification in patients on hemodialysis compared to calcium-based phosphate binders. To understand the implications of this effect for cardiovascular events, risk equations are needed. Data on 179 patients on hemodialysis treated at one center in France included biochemical values during the year prior to study entry, patient characteristics, and cardiovascular events over an average of 4 years. As arterial calcification was evaluated ultrasonographically and quantified using a 0 to 4 score, an equation relating this to the electron-beam tomography (EBT)-based calcification score used in the trial was developed and applied to all patients. The estimated scores were then used in survival and Cox proportional hazards analyses of cardiovascular events in relation to the degree of calcification, controlling for other characteristics. Mean age at inclusion was 54 years, dialysis vintage 70 months, average follow-up 49 months; 32% suffered an event. The calcification score, diabetes, C-reactive protein (CRP), diastolic blood pressure, gender, smoking and hypertension are independent predictors of cardiovascular risk. The resulting equation indicates that, relative to a calcification score below 400, the risk of an initial event increases 44% for a score of 600, and more than doubles for a score of 1000. In the absence of long-term follow-up studies, these equations permit quantification of the expected long-term clinical consequences of the impact of various phosphate binders on vascular calcification. Together with resource use and cost information, these equations are key inputs for formal cost-effectiveness analyses.