Author + information
- Received January 7, 2017
- Revision received April 19, 2017
- Accepted April 19, 2017
- Published online July 12, 2017.
- aDepartment of Cardiology and Clinical Research, Inselspital, Bern University Hospital, Bern, Switzerland
- bMount Sinai Health Medical Center, Icahn School of Medicine, New York, New York
- cJagiellonian University, Krakow, Poland
- dLeon H. Charney Division of Cardiology, New York University School of Medicine, New York, New York
- ↵∗Address for correspondence:
Dr. Franz H. Messerli, Department of Cardiology and Clinical Research, University Hospital, Bern, Freiburgstrasse, CH-3010 Bern, Switzerland.
Longstanding hypertension ultimately leads to heart failure (HF), and as a consequence most patients with HF have a history of hypertension. Conversely, absence of hypertension in middle age is associated with lower risks for incident HF across the remaining life course. Cardiac remodeling to a predominant pressure overload consists of diastolic dysfunction and concentric left ventricular (LV) hypertrophy. When pressure overload is sustained, diastolic dysfunction progresses, filling of the concentric remodeled LV decreases, and HF with preserved ejection fraction ensues. Diastolic dysfunction and HF with preserved ejection fraction are the most common cardiac complications of hypertension. The end stage of hypertensive heart disease results from pressure and volume overload and consists of dilated cardiomyopathy with both diastolic dysfunction and reduced ejection fraction. “Decapitated hypertension” is a term used to describe the decrease in blood pressure resulting from reduced pump function in HF. Progressive renal failure, another complication of longstanding hypertension, gives rise to the cardiorenal syndrome (HF and renal failure). The so-called Pickering syndrome, a clinical entity consisting of flash pulmonary edema and bilateral atheromatous renovascular disease, is a special form of the cardiorenal syndrome. Revascularization of renal arteries is the treatment of choice. Most antihypertensive drug classes when used as initial therapy decelerate the transition from hypertension to HF, although not all of them are equally efficacious. Low-dose, once daily hydrochlorothiazide should be avoided, but long-acting thiazide-like diuretics chlorthalidone and indapamide seem to have an edge over other antihypertensive drugs in preventing HF.
- antihypertensive therapy
- cardio-renal syndrome
- hypertensive heart disease
- left ventricular hypertrophy
- Pickering syndrome
Most longstanding hypertension ultimately leads to heart failure (HF) unless this sequence of events is otherwise interrupted by other outcome and as a consequence, patients with HF very commonly have a history of hypertension. In the Framingham Heart Study cohort in a total population of 5,143 subjects, hypertension antedated the development of HF in 91% of all newly diagnosed HF patients during up to 20 years of follow-up (mean 14.1 years) (1). Adjusting for age and HF risk factors, the hazard for developing HF in hypertensive compared with normotensive subjects in the Framingham Heart Study data was about 2-fold in men and 3-fold in women. Multivariable analyses revealed that hypertension had a high population-attributable risk for HF, accounting for 39% of cases in men and 59% in women. Among hypertensive subjects, myocardial infarction, diabetes, left ventricular (LV) hypertrophy, and valvular heart disease predicted an increased risk for HF in both sexes. At 80 years of age, the lifetime risk of HF was about 20% in the Framingham cohort, and this risk doubled for patients with blood pressure (BP) of 160/100 mm Hg compared with those with 140/90 mm Hg (2).
Not surprisingly, prevention of hypertension and other HF risk factors such as obesity and diabetes during middle age substantially prolongs HF-free survival. Men and women without hypertension, obesity, or diabetes at 45 years of age lived on average 34.7 years and 38.0 years without incident HF, and they lived on average an additional 3 to 15 years longer free of HF than did those with 1, 2, or 3 risk factors (3). Thus, the absence of hypertension, obesity, and diabetes by 45 and 55 years of age is associated with up to 86% lower risks for incident HF in men and women across the remaining life course. Importantly, the 22-year follow up of the SHEP (Systolic Hypertension in the Elderly Program) trial documented that compared with placebo, each month of active chlorthalidone-based antihypertensive therapy during the trial period of 4.5 years was associated with 1-day prolongation of life expectancy free from cardiovascular death (4).
In most hypertensive patients LV diastolic dysfunction is the first discernible manifestation of heart disease (Figure 1). Cardiac remodeling to a predominant pressure overload consists of concentric LV hypertrophy (increase in cardiac mass at the expense of chamber volume). In contrast, cardiac remodeling to a predominant volume overload (e.g., obesity, chronic kidney disease, anemia) consists of eccentric hypertrophy (increase in cardiac mass and chamber volume) (5). When pressure overload is sustained, diastolic dysfunction progresses, the concentric remodeled LV decompensates, and hypertensive HF with preserved ejection fraction (HFpEF) ensues. In contrast, when volume overload is sustained, LV dilatation progresses, the eccentric remodeled LV decompensates, and HF with reduced ejection fraction (HFrEF) ensues. The combination of LV hypertrophy with increased levels of biomarkers of subclinical myocardial injury (high-sensitivity cardiac troponin T, N-terminal pro–B-type natriuretic peptide) identifies patients at highest risk for developing symptomatic HF, especially HFrEF (6). The end stage of hypertensive heart disease, usually the result of longstanding pressure and volume overload, consists of dilated cardiomyopathy with both diastolic dysfunction and reduced ejection fraction.
From a clinical point of view hypertensive heart disease can be divided into 4 ascending categories, based on the pathophysiologic and clinical impact of hypertension on the heart:
Degree I: Isolated LV diastolic dysfunction with no LV hypertrophy
Degree II: LV diastolic dysfunction with concentric LV hypertrophy
Degree III: Clinical HF (dyspnea and pulmonary edema with preserved ejection fraction)
Degree IV: Dilated cardiomyopathy with HF and reduced ejection fraction (7)
The categories would indicate that diastolic dysfunction is a much more common complication of longstanding hypertension than is systolic dysfunction. Patients with HFpEF have more LV hypertrophy, epicardial coronary artery lesions, coronary microvascular rarefaction, and myocardial fibrosis than do control subjects. Coronary microvascular dysfunction may conceivably be the result of a systemic inflammatory state and oxidative stress accelerated by comorbidities of HFpEF (8,9). Importantly, isolated diastolic dysfunction also can trigger pulmonary edema, as was documented by Gandhi et al. (10). They found LV ejection fraction during an episode of acute hypertensive pulmonary edema to be similar to the one measured after treatment, when the BP had been controlled. In these patients systolic BP (SBP) was 200 ± 26 mm Hg during the initial echocardiographic examination and reduced to 139 ± 17 mm Hg at the time of the follow-up examination. Thus, a normal LV ejection fraction after the treatment of a patient with hypertensive pulmonary edema allows us to conclude that the pulmonary congestion was due to isolated, transient diastolic dysfunction. Transient systolic dysfunction with or without mitral regurgitation seemed to be infrequent during acute episodes in these patients (10).
In advanced HF SBP is usually low, even in patients who were previously hypertensive. This phenomenon is termed “decapitated hypertension”: patients who are hypertensive to begin with progressively develop normal and even low BP as HF becomes more severe. Severe LV dysfunction can be a powerful antihypertensive mechanism. The decrease in SBP results from reduced pump function and fall in cardiac output despite the presence of compensatory mechanisms such as peripheral vasoconstriction. Patients with decapitated hypertension are difficult to manage because of their inability to tolerate HF medications, most of which tend to lower BP, such as angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), diuretics, and beta-blockers.
This interplay among high BP, hypertensive HF, and dilated cardiomyopathy (i.e., the phenomenon of decapitated hypertension) was lucidly discussed by Celia Oakley (11) about 4 decades ago: “The development of left ventricular failure because of hypertension determines a decrease of the previously raised BP to normal levels and, since the failure usually persists, the BP remains normal (11). If the patient recovers from HF, then the BP rises and the diagnosis is likely to be ‘hypertension.' In other words, the ‘diagnosis” varies between dilated cardiomyopathy and hypertension according to left ventricular function and only if a patient with dilated cardiomyopathy, HF actually recovers and develops high BP can the causal or conditioning role of high BP be proved.”
We evaluated this hypothesis in a meta-analysis and were able to document that in patients with HF who underwent cardiac resynchronization therapy (CRT) an increase in BP occurred. (12). CRT improves cardiac function by reverting asynchronous mechanical events, especially in patients with a wide QRS complex or with echocardiographic dyssynchrony. In analyzing 18 studies we showed that CRT resulted in an increase in SBP by about 4 mm Hg, no change in diastolic blood pressure and an increase in pulse pressure compared with baseline or medical therapy. In a recent study by Tanaka et al. (13), an increase in SBP after CRT was associated with a decrease in the combined endpoint of HF hospitalization and all-cause mortality.
BP in HF Patients
Although hypertension is well known to trigger incident HF, higher SBP in patients with established HF seems to paradoxically have a protective effect on survival. Numerous studies have shown that in populations with HF, high SBP is associated with improved, not adverse, outcomes (14–20) and this holds true for both acute (14–17) and chronic (18,19) HF. A retrospective study examining 2,289 patients from the COPERNICUS (Carvedilol Prospective Randomized Cumulative Survival) trial by Rouleau et al. (21) found that the lower the pretreatment SBP of patients in the cohort is, the higher the risk is of a major clinical event. For each 10-mm Hg decrease in the pretreatment SBP there was an 18% increase in the risk of death, 11% increase in the combined risk of death or hospitalization for HF, and 9% increase in the combined risk of death or hospitalization for any reason.
The increase in central BP occurring with beta-blockade may be an additional reason for the beneficial effect of this drug class in HF. McAlister et al. (22) reported that the magnitude of heart rate reduction (which is indirectly proportional to the increase in central pressure) is associated with the survival benefit of beta-blockers in HF. The same phenomenon may be responsible for the benefits of heart rate lowering with ivabradine in patients with HF in the SHIFT (Systolic Heart failure treatment with the lf inhibitor ivabradine) trial (23). As discussed, in many patients with HF SBP is low, often critically so. In this situation, an increase in central pressure secondary to heart rate lowering as has been documented with ivabradine (24) may turn out to be beneficial. The decrease in cardiac workload with ivabradine is likely to override a potentially detrimental effect, if any, of a higher central systolic pressure. In contrast to hypertension, where heart rate lowering has been documented to progressively increase cardiovascular mortality, in HF there is an inverse correlation between resting heart rate and outcome (22,25).
Both the heart and the kidney are target organs in hypertension and their function and structure becomes progressively impaired with longstanding hypertensive cardiovascular disease. Not surprisingly, therefore HF and renal failure commonly occur in the same patient as a sequence of hypertension. Cardiorenal interactions occur in both directions and by a variety of mechanisms (26). The different interactions that can take place between heart and kidney have been thoroughly classified by Ronco et al. (27). In HF the most common types of cardiorenal syndrome (CRS) are:
Type 1 (acute): Acute HF leading to acute kidney injury (renal failure)
Type 2 (chronic): Chronic HF leading to progressive chronic kidney disease
Type 3: Acute, worsening of kidney function leading to HF
Type 4: Progressive primary chronic kidney disease leading to HF
From a clinical point of view the occurrence of renal failure in HF, regardless of the exact CRS type, greatly complicates all therapeutic strategies. As chronic kidney disease progresses the prevalence of HF increases and vice versa. The increase in comorbidities, as well as on the need for incremental therapies, enhances the risk of hyperkalemia. From a clinical point of view the presence of CRS drastically limits the therapeutic armamentarium in HF patients. Conceivably a treatment strategy with the newly available K+ binders in conjunction with a mineralocorticoid antagonist will help to reduce CV mortality and morbidity in HF (28).
In 1988 Pickering et al. (29) reported a series of 11 hypertensive patients with bilateral atheromatous renal artery stenosis (ARAS) who presented with a history of multiple episodes of flash pulmonary edema (FPE). The clinical entity of FPE and bilateral ARAS, now known as Pickering syndrome (30), has to be classified as CRS type 3. Patients with Pickering syndrome most commonly present with severely impaired LV filling and LV hypertrophy but may have normal or only mildly impaired LV systolic function. This, together with defective natriuresis secondary to bilateral ARAS, is the main pathogenetic mechanism leading to FPE (Figure 2). The fact that flooding of the alveolar space can occur within minutes resulting in an acute life-threatening emergency is what distinguishes FPE from other forms of decompensated HF (31). Regardless of its etiology, an acute increase of the LV end-diastolic pressure is the common denominator for the development of FPE (Figure 2). The presence of recurrent FPE, lack of typical angina, increased BP, and elevated creatinine should raise the suspicion of bilateral ARAS and hence Pickering syndrome as a possible etiology for FPE. In the series described by Pickering et al., FPE occurred on average 2.3 times before the diagnosis of ARAS was made, and in our study (21) three-quarters of all patients had more than 1 episode of FPE. Classically, these patients have some degree of renal failure and present with sudden onset of severe unprovoked dyspnea (“flash” pulmonary edema). The frequent nocturnal appearance of FPE may be related to reverse nocturnal BP dipping, a phenomenon that has been well documented in patients with ARAS.
Acute management of FPE is aimed at maintaining adequate oxygenation, diuresis, decrease of pulmonary capillary pressure, and treatment of underlying cause. However, aggressive treatment of FPE in patients with bilateral ARAS usually will cause a further decrease of glomerular filtration rate and hyperkalemia may ensue. Renal revascularization is the treatment of choice once the patient is stable and FPE has been resolved. The American Heart Association/American College of Cardiology guidelines endorse revascularization as a class 1 recommendation for patients with hemodynamic significant ARAS and recurrent unexplained congestive HF or pulmonary edema (32). In our analysis, 92% of all patients had no further FPE after revascularization. (30) Patients with recurrent FPE after renal artery stenting should have repeat Doppler ultrasound of the kidneys to rule out recurrent ARAS due to restenosis.
Antihypertensive Therapy for HF Prevention
By definition, all antihypertensive drugs lower BP. However, scrutinizing the literature reveals that not all antihypertensive drugs are created equal in their propensity to prevent HF. β-Blockers remain a cornerstone in the treatment of HF and recent review concluded that “irrespective of age or sex, patients with HFrEF in sinus rhythm should receive beta-blockers to reduce the risk of death and admission to hospital” (33). Most disappointing it is therefore that β-blockers have no better preventive effect on HF than do other antihypertensives. Among the 12 randomized controlled trials we analyzed, evaluating 112,177 patients with hypertension, beta-blockers reduced BP by 12.6/6.1 mm Hg when compared with placebo, resulting in a 23% (trend) reduction in HF risk (p = 0.055). (34) Compared with other antihypertensives, beta-blockers conferred similar but no incremental benefit for the outcomes of all-cause mortality, cardiovascular mortality, and myocardial infarction, but increased stroke risk by 19% in the elderly. Given this increased risk of stroke, beta-blockers should not be considered as first-line agents for prevention of HF.
In a Cochrane meta-analysis, calcium-channel blockers (CCBs) increased the risk of HF events (risk ratio [RR]: 1.37; 95% confidence interval [CI]: 1.25 to 1.51) as compared with diuretics. Although CCBs reduced stroke as compared with ACE inhibitors and reduced stroke and myocardial infarction as compared with ARBs, CCBs also increased HF events as compared with ACE inhibitors (RR: 1.16; 95% CI: 1.06 to 1.27) and ARBs (RR: 1.20; 95% CI: 1.06 to 1.36) (35). Stroke remains the most devastating complication of hypertension, but prima vista CCBs are probably not the antihypertensive drug class of choice for the prevention of HF.
However, a recent meta-analysis concluded that BP lowering by calcium antagonists is as efficacious as BP lowering by other antihypertensive drugs in preventing new onset HF (36). Thomopoulos et al. (36) reassessed the previously reported inferiority of CCBs in preventing HF and documented that the inferiority of CCBs may depend on, at least to a large extent, an unequal use of accompanying drugs. The more aggressive use of drugs known to reduce HF symptoms (diuretics, β-blockers, and renin-angiotensin system blockers) in the control arms might have masked onset of HF symptoms to a greater extent, thereby creating a bias against CCBs.
In ALLHAT (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial), the alpha-blocker doxazosin arm, compared with the chlorthalidone arm, conferred a higher risk of stroke and combined cardiovascular disease. Importantly, HF risk was doubled with doxazosin (4-year rates, 8.13% vs. 4.45%; RR: 2.04; 95% CI: 1.79 to 2.32; p < 0.001) (34). This would indicate that for the treatment of hypertension alpha-blockers should be avoided in patients at risk or with HF. Although the data in aggregate are less convincing for alpha-blockers than for CCBs, specifically amlodipine, a very similar point as was made previously for CCBs (35) could be made for alpha-blockers. In ASCOT (Anglo-Scandinavian Cardiac Outcomes Trial), doxazosin gastrointestinal therapeutic system given as a third-line add-on drug did not increase the risk of HF and was well tolerated (35).
Blockers of the renin-angiotensin system are efficacious drugs to treat hypertension and to prevent HF. Contrary to the common clinical notion and to some guidelines, no meaningful difference in efficacy has been documented between ACE inhibitors and ARBs (37,38). Of note, the first-in-class angiotensin II receptor neprilysin inhibitor valsartan/sacubitril not only is a novel drug for the treatment of HF, but also is likely to become a useful antihypertensive drug. Recent data have indicated that it may have a preferential effect on systolic pressure (39). A meta-analysis has shown better BP lowering with valsartan/sacubitril than with ARBs. In the PARADIGM (Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure) trial, valsartan/sacubitril showed a striking reduction in cardiovascular mortality and morbidity in patients with HFrEF (40). Of note, the long-term use of neprilysin inhibitors may compromise beta-amyloid peptide degradation in the brain, thereby possibly accelerating progression of Alzheimer's disease in patients at risk. Thus, it remains to be seen whether the risk–benefit ratio of valsartan/sacubitril confers incremental long-term prognostic benefits in patients with hypertension.
Finally, the thiazide-like diuretics chlorthalidone and indapamide are outstanding agents when used as antihypertensive drugs to prevent HF. In the SHEP trial (41) as well as in HYVET (Hypertension in the Very Elderly Trial) (42) there was a highly significant reduction of HF with active therapy against placebo, amounting to an RR of 0.51 (95% CI: 0.37 to 0.71) for chlorthalidone and 0.36 (95% CI: 0.22 to 0.58) for indapamide (p < 0.001 for both). The HF preventive efficacy of diuretics as a group in 10 randomized controlled trials was clearly superior to the one of all other antihypertensives (RR: 0.84; 95% CI: 0.73 to 0.98) (36). Of note, no outcome data are available for hydrochlorothiazide, for neither HF nor any other cardiovascular endpoint. In contrast to chlorthalidone and indapamide, hydrochlorothiazide should be avoided in hypertensive patients at risk for HF.
In conclusion, most antihypertensive drug classes when used as initial therapy decelerate the transition from hypertension to HF, although not all of them are created equal in this regard. Low-dose, once daily hydrochlorothiazide should be avoided, but the thiazide-like diuretics chlorthalidone and indapamide seem to have an edge over other antihypertensive drugs in preventing HF.
Antihypertensive Therapy in HF Patients With Persisting Hypertension
In most HF patients too low a BP is a more common clinical problem than is hypertension. However, some patients with HFpEF and a few with HFrEF present with hypertension. Because no outcome data are available, our treatment recommendations are purely empirical, based on clinical and pathophysiologic considerations (Central Illustration). Also, our recommendations of antihypertensive therapy are based on the assumption that all HF patients are on baseline triple therapy consisting of an ACE inhibitor or an ARB, plus a β-blocker and a loop diuretic, and despite this still exhibit residual hypertension. The proposed antihypertensive strategy is geared at improving diastolic and microvascular dysfunction in HFpEF and at improving or at least preserving systolic function in HFrEF. As a first step we suggest to increase afterload reduction by switching to valsartan/sacubitril (38) and to a vasodilating β-blocker such as carvedilol or nebivolol. Clearly a statin should be part of the therapeutic strategy in all patients with HFpEF to reduce microvascular dysfunction.
Dr. Messerli has served as a consultant for Daiichi-Sankyo, Pfizer, Servier, WebMD, Ipca, ACC, Menarini, and Sandoz. Dr. Rimoldi has served as a consultant for Servier, Menarini, and Takeda. Dr. Bangalore has reported that he has no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atheromatous renal artery stenosis
- angiotensin-converting enzyme
- angiotensin receptor blocker
- blood pressure
- calcium-channel blocker
- confidence interval
- cardiorenal syndrome
- cardiac resynchronization therapy
- heart failure
- heart failure with preserved ejection fraction
- heart failure with reduced ejection fraction
- left ventricular
- risk ratio
- systolic blood pressure
- Received January 7, 2017.
- Revision received April 19, 2017.
- Accepted April 19, 2017.
- 2017 American College of Cardiology Foundation
- Lloyd-Jones D.M.,
- Larson M.G.,
- Leip E.P.,
- et al.
- Ahmad F.S.,
- Ning H.,
- Rich J.D.,
- Yancy C.W.,
- Lloyd-Jones D.M.,
- Wilkins J.T.
- Seliger S.L.,
- de Lemos J.,
- Neeland I.J.,
- et al.
- Paulus W.J.,
- Tschope C.
- Mohammed S.F.,
- Hussain S.,
- Mirzoyev S.A.,
- Edwards W.D.,
- Maleszewski J.J.,
- Redfield M.M.
- Oakley C.
- Rouleau J.L.,
- Roecker E.B.,
- Tendera M.,
- et al.
- Rimoldi S.F.,
- Messerli F.H.,
- Cerny D.,
- et al.
- Messerli F.H.,
- Rimoldi S.F.,
- Bangalore S.,
- Bavishi C.,
- Laurent S.
- Bock J.S.,
- Gottlieb S.S.
- Ronco C.,
- Cicoira M.,
- McCullough P.A.
- Epstein M.,
- Pitt B.C.
- Messerli F.H.,
- Bangalore S.,
- Makani H.,
- et al.
- Hirsch A.T.,
- Haskal Z.J.,
- Hertzer N.R.,
- et al.
- Kotecha D.,
- Manzano L.,
- Krum H.,
- et al.
- Bangalore S.,
- Wild D.,
- Parkar S.,
- Kukin M.,
- Messerli F.H.
- Chen N.,
- Zhou M.,
- Yang M.,
- et al.
- Bangalore S.,
- Fakheri R.,
- Toklu B.,
- Ogedegbe G.,
- Weintraub H.,
- Messerli F.H.
- Messerli F.H.
- Bavishi C.,
- Messerli F.H.,
- Kadosh B.,
- Ruilope L.M.,
- Kario K.