NONINVASIVE 24-HOUR HEMODYNAMICS PREDICT BLOOD PRESSURE RESPONSE TO RENAL DENERVATION IN THE SPYRAL HTN-OFF MED PILOT AND PIVOTAL TRIALS

Abstract

Objective: To investigate whether hemodynamic measures obtained during regular 24-hour blood pressure (BP) monitoring may predict BP response to renal denervation (RDN) using the SYMPLICITY SPYRAL renal denervation system. Design and method: This was a post-hoc analysis from the SPYRAL HTN-OFF MED pilot and pivotal trials. We used the raw BP waveforms, obtained at baseline and 3 months during 24-hour BP monitoring with an oscillometric sphygmomanometer. Waveforms were acquired at diastolic BP level, underwent in-device and post-hoc quality control, and were processed with validated ARCSolver algorithms to derive parameters of pulsatile hemodynamics (Augmentation Index – AIx). AIx quantifies the impact of wave reflection on the central aortic waveform. Its main hemodynamic determinants are aortic stiffness, peripheral resistance, heart rate and left ventricular function. We investigated the relationship between 24-hour AIx at baseline and the change in 24-hour systolic and diastolic BP (SBP/DBP) at 3 months in the combined RDN groups from both trials. Data were adjusted for observed trends in the sham group. We defined a favorable response as a drop in 24-hour systolic BP of at least 5 mm Hg. Results: Raw pressure waveforms were available for analysis in 103 patients undergoing RDN and 104 sham controls. Baseline characteristics and 24-hour BP changes at follow-up were comparable to the original trials. The relationship between baseline AIx and change in 24-hour SBP/DBP in the RDN group was: -8.2/-6.6 mm Hg for AIx < = 50. percentile and -0.8/-0.8 mm Hg for AIx > 50. percentile (p < 0.001). In addition, a higher heart rate at baseline was associated with a larger BP drop at 3 months. Taking AIx and heart rate into account, a favorable BP response following RDN could be predicted with an area under the curve of 0.70. Interestingly, baseline BP was no longer associated with BP response to RDN after adjustments for the BP changes in the sham group. Conclusions: Our hypothesis-generating results for the first time suggest that hemodynamic measures, obtained during regular 24-hour BP monitoring in the workup before RDN, may predict BP response. Due to sample size, further investigation is needed.