The Predictive Ability of the Renal Resistive Index and Its Relationship to Duplex Ultrasound Waveform Propagation in the Aorta and Renal Arteries

Background: The objective is to investigate whether calculating the PPI (Pulse Pressure Index) and the RRI (Renal Resistive Index) using routinely collected Duplex ultrasound waveforms data obtained from the aorta and renal artery correlates and predicts renal function, and determine whether RRI is affected by the presence of a renal artery stenosis.

Methods: The records of 965 patients were evaluated. The RRI or pulsatility index of the aorta, renal artery, hilum, cortex, and medulla were measured with concurrent glomerular filtration rate GFR, Cr, PPI, and HR measurements, among which 75 patients had a 24-hour urine measured for CrCl, and 32 patients had aortic pulse pressure index (API) calculated from the central aortic pressure measured with applanation tonometry. The propagation of the pulsatility was evaluated by Analysis of Variance (ANOVA). The correlation coefficient (r) and the linear regression coefficient of determination R-squared (R²) were determined. The effects of a renal artery stenosis were evaluated with a paired t-Test comparing the RRI in 192 patients where only one side had a renal artery stenosis greater than 60%.

Results: The pulsatility indexes and RRIs progressively decreases and are statistically distinct by ANOVA from the aorta to the renal cortex (P = 7.26 × 10^-125). CrCl correlates with the PPI, cortex RRI and medulla RRI with r equal to -0.34, -0.23 and -0.42 (P < 0.05). GFR correlates with the PPI, cortex RRI and medulla RRI with r equal to -0.15, -0.12, and -0.20 (P < 0.0001). Cr correlates with the PPI, cortex RRI and medulla RRI with r equal to 0.09, 0.12, and 0.14 (P < 0.005). The CrCl, GFR and Cr were not statistically correlated with the HR. On univariate and multivariate analysis, the R² predictive value for PPI, cortex RRI and medulla RRI for CrCl, GFR and Cr were all less than 0.2 (P < 0.05). The cortex and medulla RRI were correlated with the API with r = 0.63 (P < 0.001). The R² predictive value of the PPI for the cortex and medulla RRI was 0.41 and 0.28 (P < 0.001), respectively. On paired t-Test analysis renal artery stenosis had no effect on the RRI (P = 0.78).

Conclusions: The RRI is calculated based on velocity waveform propagation where pulsatility slowly decreases in a series of elastic vessels. While CrCl, GFR and Cr do correlate with the PPI, cortex RRI and medulla RRI, the R² coefficient of determination for these correlations demonstrate that they are poor predictors of renal function. Renal artery stenosis did not have any effect on the RRI.