MP596INFLUENCE OF HEMODIALYSIS ASSOCIATED CARDIOVASCULAR COMPLICATIONS ON PULSE WAVE ANALYSIS: MODELING-BASED APPROACH

Abstract

INTRODUCTION AND AIMS: Cardiovascular complications are the most frequent cause of death in the group of patients on intermittent hemodialysis (HD). Pulse wave analysis (PWA) performed on a pulse wave shape recorded in peripheral arteries is well-established tool for non-invasive cardiovascular system assessment. However, due to many abnormalities, such as fluid overload, changed blood rheology and arteriovenous fistula, it is unclear how reliable is PWA for patients on hemodialysis. The aim of this study was to asses if those abnormalities have any influence on PWA. METHODS: Peripheral pressure profiles (PPPs) of 21 healthy volunteers (48% females, age 40.9 +/- 10.3 years, weight 79.2 +/- 21.3 kg) and 25 stable prevalence HD patients (60% females, age 64.8 +/- 13.3 years, weight 70.1 +/- 15.4 kg, median dialysis vintage 9.6 (1.3 - 29.8) years) were recorded and analyzed using applanation tonometry (SphygmoCor, AtCor Medical, Australia). Patient-specific stroke volumes (SVs) were additionally measured using impedance cardiography (PhysioFlow, Manatec Biomedical, France). Recorded PPPs were used to estimate five patient-specific parameters of physiology-based mathematical model of blood flow in the system of fifty-five compliant and tapering arterial segments, i.e. elasticity of arterial wall, compliances and resistances of arterioles and capillary bed, moment of peak left ventricular ejection (MLV), and SV. Correlations were tested using Pearson's linear correlation coefficient. RESULTS: There was a high agreement between the simulated and recorded PPPs in both groups (average coefficient of determination equal to 0.94 and 0.92 for control and HD group, respectively). In both groups statistically significant correlations were found between model estimated and SphygmoCor derived central systolic/diastolic (CSP/CDP) pressure and heart ejection duration, compare Fig. 1. Model predicted CSP/DSP were on average higher/lower by about 5 mmHg compared to those derived by SphygmoCor. Correlation between model predicted and measured SV was found in both cohorts. Interestingly, correlations between model predicted and SphygmoCor derived augmentation index and MLV were found only in the control group, compare Fig. 1. CONCLUSIONS: Significant agreement between modeling and PWA results in healthy cohort indicate that for general population PWA agrees with the standard physiological assumptions about pulse wave propagation. This is not the case, however, in the group of HD patients, indicating that the cardiovascular system abnormalities may have a significant impact on PWA. Having established patient-specific cardiovascular modeling framework we can further investigate how various changes in CV system affect PWA. (Figure presented).