Are we getting closer to the HEART of Hypertensive Heart Disease?

Hypertension affects nearly half of US adults, and elevations in blood pressure are associated with an increased risk of heart failure, myocardial infarction, stroke, arrhythmia, and death.¹ Hypertensive heart disease (HHD), the development of alterations in cardiac structure in the absence of other cardiovascular disease (CVD), has traditionally been thought of as one of the major mediators of this increased risk (Figure). HHD develops as a result of chronic pressure and volume overload, which in turn lead to increased afterload and myocardial wall stress, as well as upregulation of the renin angiotensin aldosterone (RAAS) and sympathetic nervous systems (SNS). These alterations cause cardiomyocyte hypertrophy and apoptosis, and fibroblast hyperplasia which ultimately result in the development of left ventricular hypertrophy (LVH) the pathological hallmark of HHD. Concurrently, the conversion of fibroblasts into myofibroblasts results in the production of cross-linked type I collagen fibers that accumulate in the extracellular matrix in interstitial and perivascular fibrils. This increase in fibrosis leads to abnormalities of diastolic function as well an increased risk of arrhythmia and ultimately systolic dysfunction as the concentration of collagen fibrils increases.

Figure.

Pathways through which hypertension leads to hypertensive heart disease. Abbreviations: HHD = hypertensive heart disease; LVH = left ventricular hypertrophy; SNS = sympathetic nervous system; RAAS = renin-angiotensin-aldosterone-system

In observational studies, LVH has been shown to increase CVD risk and mortality independent of blood pressure,^{2, 3} however, it remains to be shown whether the phenotypic reversal of HHD through the treatment of hypertension improves outcomes. Reductions in blood pressure lead to a concurrent decrease in LV mass, but there are class effects of medications and not all antihypertensive agents result in a similar reduction in left ventricular mass for a given change in blood pressure; ACE/ARBs, angiotensin II receptor antagonists and calcium antagonists perform better than diuretics.⁴ Different antihypertensive classes also have differing effects on the extracellular space and fibrosis. Losartan⁵ and lisinopril⁶ have been shown to reduce fibrosis and improve diastolic function, but diuretics have not. Recent work in heart failure has shown that antifibrotic therapies that target myofibroblasts may also help to reverse interstitial fibrosis.⁷ Although decreasing blood pressure results in reductions in CVD events and mortality as well as the regression of LVH and fibrosis, the contribution of HHD regression to these improved outcomes had not been clearly demonstrated.

In this issue of Hypertension, Upadhya et al. report the results of SPRINT-HEART, a cardiac magnetic resonance imaging (CMRI) ancillary of the SPRINT trial designed to examine the role intensive blood pressure lowering plays on alterations in cardiac structure and function and their subsequent effect on the clinical outcomes.⁸ Despite a reduction in concentric remodeling and trend towards a greater reduction in left ventricular mass in the intensive therapy arm after 18 months, there was no significant improvement in cardiac function or change in native T1 times. The authors conclude that regression of HHD is not the mechanism through which intensive blood pressure lowering results in improved outcomes in SPRINT. Their finding is concordant with the SPRINT electrocardiography (ECG) analysis of LVH regression by Solimon et al.⁹

In order to ascertain the contribution of LVH regression and/or fibrosis reduction to improvements in morbidity and mortality associated with blood pressure lowering, one must be able to accurately quantify these abnormalities of cardiac structure. Endomyocardial biopsy is the gold standard for assessing HHD and quantifying fibrosis and myocyte hypertrophy, however its invasive nature and associated risks make it an impractical solution for longitudinal assessments of HHD. The sometimes non-diffuse, patchy nature of fibrosis is another limitation of endomyocardial biopsy. Thus, a non-invasive method of quantifying myocyte hypertrophy/LVH and fibrosis is preferable in particular to study the effects of hypertension and blood pressure lowering on the progression or regression of HHD over time. There are several non-invasive ways to diagnose LVH including ECG, echocardiography, and CMRI. Of these three modalities, CMRI has the advantage of the highest precision in the quantitation of ventricular volumes and mass, superior reproducibility, and the added benefit of tissue characterization. Late gadolinium enhancement has traditionally been used to localize fibrosis because gadolinium collects in the extracellular space. With greater replacement of extracellular space by fibrosis there is a larger concentration of gadolinium which is visualized as hyperenhancement. However, there are some limitations to the use of late gadolinium imaging, in HHD fibrosis is diffuse and hyperenhancement is not well visualized or easily quantified, and gadolinium cannot be safely administered to certain individuals with advanced kidney disease (eGFR <30 mL/min/1.73 m²), a common co-morbidity in HHD. To overcome these limitations, T1 mapping was developed. T1 mapping is a reflection of the relaxation of water in the intra- and extracellular spaces and can be thought of as a virtual biopsy. Abnormalities of native T1 can reflect cellular hypertrophy and/or extracellular fibrosis, though without the use of gadolinium contrast one cannot discern which space is affected. One of the first studies of CMR in HHD found that adults with hypertension and LVH had an increase in extracellular volume reflected by a longer native T1 time compared to hypertensive individuals without LVH as well as normotensive controls.¹⁰ However T1 mapping also has limitations in that there are only slight differences and some overlap between T1 scores in normal versus hypertensive individuals, and T1 timing can be affected by other alterations in the extracellular volume including infiltration and inflammation.¹¹ Thus the use of CMRI with T1 mapping in SPRINT had the potential to explore changes in HHD associated with intensive and standard blood pressure lowering. However, of the 337 participants, only 9.1% of men and 13.3% of women had LVH and T1 times displayed a narrow distribution, and as such only a relatively small proportion of individuals with HHD was included in this sub-study, which may in part explain the null findings.

If alterations in LV mass and fibrosis do not explain the reductions in CVD and mortality observed with intensive BP lowering, how can we explain these findings? One possible explanation is that there are other mediators such as alterations in the RAAS or SNS, aside from HHD reversal that result in the improved outcomes, or that the aggressive use of diuretics in SPRINT resulted in the significant reduction in heart failure that was seen. Perhaps it is not just the BP lowering effect, but class actions of medications that alter cardiac structure as has been previously demonstrated. Unfortunately, the authors do not provide data on the medication class use and associated alterations in LV mass and native T1. Though we know that at the time of randomization participants should have been on 2 or 3 drugs including a thiazide diuretic, ACE/ARB or calcium antagonist. These medications, through pleiotropic effects as well as their resultant reductions in blood pressure also reduce mechanical and sheer stresses on the glomerular apparatus and vasculature. Additional reductions in oxidative stress and inflammation may also potentially contribute beneficial downstream effects on the vasculature as well as end-organs such as the brain and kidney. Other alternative explanations include the possibility that HHD is irreversible or the 18 months duration of follow-up was too short to see significant differences between groups, or that the range of native T1 values seen in the intensive and standard arms were too close to for native T1 mapping to be a useful measure to follow between groups. Prior studies have shown it can take up to 3 years for full resolution of LVH, normalization of LV mass and diastolic function.¹² Additionally, the authors did not perform post-gadolinium T1 mapping to calculate extracellular volume, a more robust and specific measure of fibrosis. Indeed, a recent study of symptomatic patients with severe aortic stenosis, a higher-afterload condition compared to HHD, showed regression of both extracellular matrix volume and cellular hypertrophy after aortic valve replacement using pre and post-gadolinium T1 mapping calculation.¹³

In conclusion, the findings by Upadhya et al. support those of Soliman and suggest that the regression of HHD has little effect on the improved outcomes which result from intensive blood pressure lowering in SPRINT. Based on these findings there seems to be no clinical role for the serial assessment of cardiac structure to estimate the effect of blood pressure lowering. The use of CMRI with gadolinium to better quantify the changes in intra- versus extracellular spaces that result from the use of antifibrotic antihypertensive agents such as ACEI/ARBs may prove beneficial in the assessment of regression of fibrosis in conjunction with functional assessments of diastolic function.