Nonresponse to Renal Denervation in a Simulated Cohort of Resistant Hypertensive Men

In all three clinical trials of the Symplicity renal denervation device, a significant fraction of participants had no response to the treatment. Defining nonresponders as having a fall of less than 10 mmHg in both systolic and mean arterial pressures, we investigated the question of what mechanisms might produce resistance to renal denervation in a virtual patient cohort. The study was conducted using HumMod, an integrative physiological simulator, by varying parameters influential to resting MAP to produce a patient base. Hypertension was induced in each patient by utilizing a mixed etiology model, combining unilateral stenosis, loss of renal function, central increase in sympathetic outflow, and decreased systemic arterial compliance. Individuals were tested for resistance to thiazide diuretics, 75% renal denervation, and diuretic + denervation. Sampling resulted in patients with variable responses to each treatment, including patients that were responsive to either denervation or diuretic, but not both. Hypertensives that were non‐responders to thiazide tended to have lower systolic pressure, left atrial pressure, epinephrine, ANP, and stroke volume, with higher nerve activity, AII and plasma sodium concentrations. Non‐responders to denervation had increased systolic and diastolic pressures, along with increased LAP and ANP, but displayed reduced sympathetic outflow. Individual p‐values were low, but machine learning algorithms indicate that the multivariable distribution including systolic and diastolic pressures was highly predictive of response to denervation. Response to each intervention was independent of the other, which we attribute to the independence of the antihypertensive mechanisms of the interventions.