Adverse cardiac remodeling is absent in patients with true controlled resistant hypertension

Abstract Resistant hypertension (RHTN), defined as blood pressure (BP) that is uncontrolled with ≥3 medications, including a long‐acting thiazide diuretic, also includes a subset with BP that is controlled with ≥4 medications, so‐called controlled RHTN. This resistance is attributed to intravascular volume excess. Patients with RHTN overall have a higher prevalence of left ventricular hypertrophy (LVH) and diastolic dysfunction compared to patients with non‐RHTN. We tested the hypothesis that patients with controlled RHTN due to the intravascular volume excess have higher left ventricular mass index (LVMI), higher prevalence of LVH, larger intracardiac volumes, and more diastolic dysfunction compared to patients with controlled non‐resistant hypertension (CHTN), defined as BP controlled with ≤3 anti‐hypertensive medications. Patients with controlled RHTN (n = 69) or CHTN (n = 63) who were treated at the University of Alabama at Birmingham were offered enrollment and underwent cardiac magnetic resonance imaging. Diastolic function was assessed by peak filling rate, time needed in diastole to recover 80% of stroke volume, E:A ratios and left atrial volume. LVMI was higher in patients with controlled RHTN (64.4 ± 22.5 vs 56.9 ± 11.5; P = .017). Intracardiac volumes were similar in both groups. Diastolic function parameters were not significantly different between groups. There were no significant differences in age, gender, race, body mass index, dyslipidemia between the two groups. The findings show that patients with controlled RHTN have higher LVMI, but comparable diastolic function to those of patients with CHTN.


INTRODUCTION
Resistant hypertension (RHTN) is defined as BP which is uncontrolled with ≥3 medications-commonly a long-acting calcium channel blocker (CCB), a blocker of the renin-angiotensin system [angiotensinconverting enzyme (ACE) inhibitor or angiotensin receptor blocker (ARB)], and a diuretic, all prescribed at maximal or maximally tolerated doses and at the appropriate dosing frequency. 1 A subset of RHTN, controlled RHTN, is BP that is controlled with ≥4 medications ( Figure 1). The prevalence of RHTN among all patients with treated hypertension is 10-15%. 2,3,4 Patients with either uncontrolled or controlled RHTN have a higher risk of developing cardiovascular (CV), neurological, and renal adverse outcomes than patients with non-RHTN. 5,6 However, controlled RHTN patients have rarely been studied as a separate entity and thus the cardiac structure and function of these patients is not well characterized compared to patients with controlled non-RHTN (CHTN).
We have previously demonstrated intravascular volume excess, as indicated by increased atrial natriuretic peptide (ANP) and B-type natriuretic peptide (BNP) levels, in patients with RHTN compared to normotensives and patients with CHTN. 7 Thus, intravascular volume excess appears to be a major pathophysiologic mechanism underlying resistance to anti-hypertensive treatment. Since volume overload is a known risk factor for LVH, 8 in this study, we tested the hypothesis that RHTN patients with controlled BP would have higher LV mass, higher prevalence of LVH, larger intracardiac volumes, and more diastolic dysfunction than those with CHTN.
Diastolic dysfunction is present in approximately 50% of patients with clinical congestive heart failure, 9 and has been associated with worse CV outcomes. 10,11 Hypertension is one of the most prevalent risk factors for diastolic dysfunction. 12,13 BP control can improve diastolic function, and some ARBs and CCBs have been particularly helpful in achieving that goal. [14][15][16] This study tested whether BP control can restore and/or preserve diastolic function in patients with controlled RHTN to levels similar to those in patients with CHTN.

METHODS
Hypertensive patients were recruited from the University of Alabama at Birmingham (UAB) Hypertension Clinic. Controlled RHTN was defined as controlled office BP (≤135/85 mmHg) for at least three consecutive follow-up visits on four or more medications, including a long-acting thiazide diuretic ( Figure 1). Patients then underwent 24-h ambulatory blood pressure monitoring (ABPM) to assess BP control out of clinic ( Figure 2

RESULTS
A total of 182 patients met the inclusion criteria for the study.
After undergoing 24-hr ABPM and CMR, the final cohort included   Table S5).

Analyses based on age, dyslipidemia, and diabetes
Linear regression modelling showed that the latter association was independent of age, gender, and BMI.

Cardiac diastolic function
The similar diastolic function seen in all groups may be related to sustained BP control. Many factors other than HTN are known to increase the risk of diastolic dysfunction, such as older age, obesity, coronary artery disease, history of myocardial infarction and systolic dysfunction, were similar in both groups, suggesting that they were not confounders. 12,13 MRA use in the controlled RHTN group could also have improved diastolic function.
In this patient population 67% of patients with controlled RHTN were taking an MRA (Table 1). Spironolactone has been shown to improve diastolic function, especially in patients with LVH and heart failure with preserved ejection fraction (HFpEF). 35 Patients with older age and dyslipidemia had longer time to PRF and percentage of diastole needed to reach PRF, correlating with worse diastolic function regardless of their HTN phenotype. Age is a wellknown risk factor for diastolic dysfunction. 12,13 Dyslipidemia, however, has been reported as a risk factor in only a few studies. [39][40][41] The effects of dyslipidemia and aging were likely more pronounced in our study, given that the patients had similar blood pressure, one of the cardinal risk factors for diastolic dysfunction. The association between dyslipidemia and the PFR parameters was independent of age, gender or BMI.
This association is likely due to the effects of dyslipidemia on the myocardium. High cholesterol diet has been linked to impaired cardiac relaxation due to alterations in cellular calcium currents in rabbit models. 39 Low HDL levels were also associated with higher E:A ratios on echocardiography in humans. 40 Another study found that high triglyceride levels were positively correlated with the E:A ratios on echocardiography. 41   The wider pulse pressure seen in this study is likely due to patients with masked uncontrolled HTN as seen in the subgroup analyses (Supplementary Table S1). This difference in pulse pressure has also been shown in previous data from our group. The etiology of pulse pressure widening is not fully understood, but may be due to increased sympathetic tone, as reflected in higher out-of-clinic catecholamines and metanephrines on 24-h urine collections. 43 Although the method of calculating non-dipping nocturnal BP is not well defined, it has been shown to be associated with more adverse target organ damage in other studies. 44 It is important to note that, given the relatively low and well controlled BP in the participants in this study, non-dipping is more common in this study population likely due to their systolic and diastolic pressures being closer to the physiologic norms. The prevalence was accordingly relatively high in all groups. There is an association of non-dipping BP with LVH, but it was shown in patients which would be defined as having poorly controlled BP according to recent guidelines. 25 In addition, the association of non-dipping BP with adverse outcomes was nonsignificant in many studies when adjusted to nighttime and 24-h BP. 45,46 This shows that non-dipping is unlikely to be a major factor of cardiac remodeling in our patient population which has well controlled BP and low nocturnal BP.

Study design
The The strengths of the current study include enrolling patients after three consecutive controlled clinic BP readings obtained unattended and according to the AHA HTN measurement recommendations, use of ABPM to identify patients with masked uncontrolled hypertension, and use of CMR, which is considered the current gold standard for studying the structure and function of the heart. Limitations include the small sample size and the cross-sectional design, which does not account for changes in a baseline cardiac structure and function over time. Some participants were followed for only a short time at our institution prior to the beginning of the study, making it difficult to establish the dura-