1 Introduction

Arterial hypertension (AH) is a public health burden, with an overall prevalence in adults estimated around 30–45% [1], making it a major preventable cause of cardiovascular disease and all-cause mortality worldwide [2]. A major contribution to the morbidity and mortality related to AH derives from heart failure (HF), for which AH represents the leading risk factor [3, 4]. Hypertensive patients show a five- to six-fold higher risk of developing HF than healthy subjects, and lifetime HF risk grows directly proportional to blood pressure [5]. The presence of AH before the HF-related symptoms represents the most prevalent risk factor in heart failure with preserved ejection fraction (HFpEF), which is defined, according to current guidelines, by left ventricular ejection fraction (LV EF) ≥ 50% [5]. Indeed, hypertensive patients fall into stage A-HFpEF, according to the American College of Cardiology Foundation/American Heart Association classification (AHA/ACC), while those with AH-associated asymptomatic cardiac structural and/or function alterations fall into Stage B-HFpEF [6].

To tailor preventive and therapeutic interventions, a complete understanding of the mechanisms underlying the transition from HF stages A-B to clinical HF (AHA/ACC Stage C) would be of utmost importance. Cardiopulmonary-echocardiography stress test (CPET-ESE) is currently used to detect early alterations in patients with a definite diagnosis of HF, differentiating central from peripheral abnormalities in the cardiopulmonary response to physical exercise [7,8,9,10,11,12]. However, the opportunity to study the entire cardiovascular system under exercise makes the CPET-ESE equally effective in highlighting initial alterations induced by hypertension and shared with HF, thus identifying early and long-acting pathways that might account for the transition between the two clinical conditions [13]. As a matter of fact, CPET-ESE has recently demonstrated an additional predictive value in patients with subclinical HF compared to the two techniques taken individually, independently from bio-humoral, clinical and instrumental parameters evaluated at rest [14,15,16]. Several biomarkers of cardiac function bear prognostic significance in patients with overt HF, among which the N-terminal pro-B type natriuretic peptide (NT-proBNP) is the most used, according to current Guidelines [5]. However, these biomarkers are often normal in subjects in ACC/AHA HF stages A-B, including those with AH, and therefore cannot be used to evaluate the risk of transition towards more advanced HF Stages. Indeed, a multiparametric score including NT-proBNP and CPET-ESE derived parameters has been recently proposed as an accurate prognostic tool to predict HF hospitalization and cardiovascular death in Stage C-HFpEF, as well as to identify the subjects in Stages A and B at risk of transition towards more advanced HF Stages [13].

2 Pathophysiology

AH leads to progressive cardiovascular alterations through several different processes. Longstanding pressure overload on the LV is the first well-known mechanism of damage, which brings about structural remodeling and ultimately LV concentric hypertrophy (LVH). All of these alterations are often globally referred to as “hypertensive heart disease”, even if there is no unanimous agreement on the definition [17, 18]. Commonly, diastolic dysfunction (usually assessed by echocardiography) is the first manifestation of this pathological process, discernible in early, mild AH even before the development of LVH [19,20,21]. This notwithstanding, it is worth noting that hypertensive heart disease is also characterized by underhanded systolic dysfunction, even before the onset of clinical HF [12].

A network metanalysis by Sciarretta et al. [22] investigated the impact of different antihypertensive therapies in the prevention of HF, concluding that diuretics seem the most effective class of drugs in preventing the transition from AH to overt HF. Nevertheless, pressure overload is only one of the numerous mechanisms involved in the development of HF. Structural alteration of small coronary arteries, with increased collagen deposition and fibrosis (i.e. microvascular disease), is a key feature of cardiac chamber damage, detectable in the early stages of hypertensive heart disease [23]. Chronic hyperactivation of neurohormonal pathways (i.e. renin-angiotensin-aldosterone system) has a pivotal role in AH, as does the persistent increase of proinflammatory cytokines (tumor necrosis factor-alpha, interleukin-1, and interleukin-6), growth factors (i.e. transforming growth factor-beta) and reactive oxygen species [18, 24]. This miscellaneous humoral environment leads to myocardial fibrosis, coronary microvascular endothelial inflammation, and rarefaction [25,26,27]. This, in turn, reduces coronary flow reserve, which is also impaired by diminished vasodilatory capacity [28], elevation in LV diastolic filling pressure and compression of subendocardial microcirculation due to increased LV wall thickness [29]. In a vicious circle, the inability to satisfy increased oxygen demand (associated with LVH) predisposes to ischemia and further fibrosis and remodeling [18].

The longstanding neurohormonal and inflammatory activation also impacts considerably on the remodeling and stiffening of systemic arteries, which is a crucial feature in the pathophysiology of AH. Increased arterial stiffness causally contributes to isolated systolic hypertension [30], leading to both ventricular-arterial uncoupling and microvascular damage [31,32,33]. There is a strong correlation between LV diastolic function, assessed by tissue Doppler imaging-derived mitral annular velocity in early diastole (e’), and indices of arterial afterload (i.e. arterial elastance, arterial compliance, systemic vascular resistance) or vascular stiffness (i.e. pulse wave velocity) both in healthy ageing and in hypertensive patients with or without HF [34]. Moreover, pulse wave velocity is related to abnormal myocardial deformation in systole, assessed by speckle tracking echocardiography (STE) [35] and to increased serum levels of biomarkers of collagen turnover (suggesting excessive myocardial collagen deposition). Beyond its impact on LV function and structure, the increased pulsatile energy transmitted by stiff vessels to the microvascular system promotes remodeling of the small vessels, which might further impair coronary flow reserve at the cardiac level and oxygen extraction in the periphery [36]. Finally, the microvascular disease promoted by increased arterial stiffness is responsible for the damage of target organs such as the brain and the kidney [37]. Due to these synergetic mechanisms, and in parallel with LVH development, AH is thought to progress towards HFpEF [38,39,40,41], which is indeed characterised by impaired LV systo-diastolic function, reduced exercise capacity and a high prevalence of peripheral and coronary microvascular dysfunction [42,43,44].

On the other hand, AH can also lead to HF with LV EF < 40% (HFrEF) [5]. The Cardiovascular health study demonstrated that LVH is a strong predictor of depressed ventricular function and acts as a direct or indirect predecessor of systolic deterioration through chronic myocardial ischemia [45]. Indeed, comorbidities commonly associated with AH, such as obesity, chronic kidney disease and anemia, can lead to volume overload and thus contribute to LV dilation [20]. Therefore, end-stage hypertensive heart disease can result in dilated cardiomyopathy. Furthermore, the development of acute ischemic events, facilitated by the co-existence of AH and coronary atherosclerosis, is another frequent circumstance that can cause the progression of hypertensive heart disease towards dilated cardiomyopathy [40, 46, 47].

3 Combined Cardiopulmonary-Echocardiography Stress Test

Transthoracic echocardiography has a well-known role in the clinical assessment of hypertensive patients, both to detect LV remodeling and to evaluate the response of cardiac structural and functional alterations to antihypertensive medications [2, 40]. 2D-derived resting LV mass and geometry and left atrial dimension have been used traditionally in the stratification of cardiovascular risk, while further evaluation of LV diastolic and systolic function can be assessed by echo-Doppler and STE [48]. Noteworthy, the latter technique—namely, STE-derived global longitudinal strain (GLS)—has shown to be more reliable than LV EF in evaluating inapparent abnormalities in myocardial contractile function [49]. However, physical exercise can stress cardiopulmonary homeostasis and unmask pathological hemodynamic changes still unapparent at rest, allowing a better characterisation of the transition from AH to HF (Table 1).

Table 1. CPET, ESE and the combined CPET-ESE approach in the analyses of cardiopulmonary response to exercise in hypertensive patients with and without HFpEF.
Full size table

In hypertensive patients, ESE finds its main application in risk stratification and in the diagnosis of myocardial ischemia [50]. However, this technique has also been successfully used in the last years to gain insight into cardiovascular mechanisms underpinning effort intolerance in patients in different subsets of HF, from Stages A-B to overt HF [51,52,53]. ESE can integrate the assessment of cardiac function analysed at rest through the evaluation of chamber geometry and volumes, LV systolic and diastolic function, left atrial structure and function, and valvular function during exercise [54].

Progressive impairment of cardiac mechanics and/or coronary flow reserve can result in the absence of contractile reserve, defined as the inability to increase LVEF ≥ 7.5% from rest to peak exercise [55]. Subclinical LV contractile dysfunction can be evaluated by tissue Doppler imaging-derived systolic mitral annulus tissue velocity (s’) [12] and STE-derived GLS [55]. On the other hand, early (E) to late (A) diastolic transmitral flow velocity ratio is part of the routine assessment of diastolic function both at rest and during exercise, together with the ratio of transmitral flow velocity to mitral annular velocity in early diastole (E/e’) [56]. The latter, in particular, correlates well with LV filling pressure when assessed at rest. Resting E/e’ > 15 represents a major criterion in the diagnostic workup of HFpEF [57], as well as a powerful prognostic indicator in this population [58, 59]. However, albeit peak effort E/e’ > 15 is deemed to express a stress-induced increase in LV filling pressure, doubts have been recently raised regarding the technical feasibility and reliability of this parameter during exercise [60]. The evaluation of diastole is integrated with the estimation of systolic pulmonary arterial pressure (sPAP), which is calculated by adding estimated right atrial pressure to measured tricuspidal regurgitation velocity [61]. An inappropriate increase in sPAP indicates a pulmonary hypertensive response to physical or pharmacological stress [54], which is common in AH and is usually associated with increased LV filling pressure [9, 62].

ESE can be easily combined with lung ultrasound, which allows for the assessment of extravascular lung water, as revealed by the presence of B-lines [63, 64]. An increase in the number of B-lines during exercise, rather than their absolute number at a given moment, indicates cardiogenic extravascular lung water accumulation, which is likely due to increased filling pressures [65, 66]. In a recent study involving patients in AHA/ACC HF Stages A/B (81% of which had AH), those with higher peak-rest ΔB-lines (i.e., more severe exercise-induced pulmonary congestion) were at higher risk of transition towards manifest HFpEF after a two-year follow-up [13].

The combination of ESE with CPET can provide important additional information on the systemic and pulmonary response to exercice. CPET has been increasingly recognised as a reliable tool in assessing aerobic fitness due to the possibility to analyse in a non-invasive and multiparametric fashion the cardiovascular, respiratory and metabolic response to exercise [14]. Oxygen consumption (VO2) was the first exercise-derived parameter used to categorise HFrEF [67]; a decade later, Mancini et al. proposed peak VO2 as a prognostic risk factor for cardiovascular death in patients with advanced HFrEF [68]. However, increasing evidence suggests that effort intolerance is as common in patients with chronic, stable HFpEF as it is in subjects with HFrEF [7, 10, 12, 69]. Current Clinical Recommendations discuss the possibility of converting measured peak VO2 to percent-predicted to account for age and sex differences [70]. However, recently the reliability of this parameter compared to peak VO2 in HFpEF has been questioned [71].

CPET can detect impairment in aerobic metabolism by evaluating the anaerobic threshold (AT), which identifies the physiological switch to mainly anaerobic metabolism [72]. Earlier AT denotes a principal role of anaerobic metabolism during exercise, and this finding characterises both physical deconditioning and HF [73]. The increased amount of time spent in anaerobic conditions appears to be related to impaired oxygen extraction [74]. Indeed, VO2 depends on cardiac output (CO; i.e., heart rate times stroke volume) and oxygen extraction by peripheral tissues (i.e., arteriovenous oxygen difference [AVO2diff]), as summarised by Fick’s principle [15]. Thus, VO2 is the physiological result of the interplay between central (CO) and peripheral (AVO2diff) components. The physiological increase in AVO2diff during exercise is blunted in the whole spectrum of HF, probably due to microvascular and mitochondrial impairment in AHA/ACC A-B stages and C-HFpEF [12, 75], with the additional contribution of cardiac cachexia and sarcopenia in advanced HFrEF [73]. Noteworthy, AVO2diff can be assessed only through invasive catheterisation or estimated by CPET-ESE from Fick’s principle as VO2/CO. Indeed, although CPET allows a global analysis of VO2 during exercise, only a concurrent invasive hemodynamic evaluation by cardiac catheterisation or with imaging techniques such as ESE allows distinguishing its central and peripheral determinants [76]. For this reason, in the last years, the integrated CPET-ESE evaluation has been revealed to be a precious tool to refine the characterisation of patients with HF or at risk of developing it [12,13,14,15,16, 51] (Fig. 1).

Fig. 1.
figure 1

The combined CPET-ESE approach to evaluate the cardiopulmonary response to exercise in hypertensive patients with and without HFpEF. Bottom images depict left ventricular hypertrophy and arterial thickening, which are characterizing features of HFpEF. AVO2diff arteriovenous oxygen difference, CO cardiac output, CPET-ESE cardiopulmonary-echocardiography stress test, GLS global longitudinal strain, HFpEF heart failure with preserved ejection fraction, LV EF left ventricular ejection fraction, sPAP systolic pulmonary arterial pressure, VE/VCO2 ventilatory equivalent for carbon dioxide, VO2 oxygen consumption

Full size image

The ventilatory equivalent for carbon dioxide (VE/VCO2) is another crucial parameter evaluated by CPET, as it demonstrated robust prognostic power in HF, even in submaximal levels of effort, in contrast to other CPET parameters such as VO2 [14, 77, 78]. VE/VCO2 is an index of ventilation/perfusion matching in the lung and describes the increase in minute ventilation for any given amount of CO2 generated from cellular respiration. It is determined by CO2 production, the physiological dead space to tidal volume ratio (VD/VT) and the arterial CO2 partial pressure, with higher values (i.e., a steeper slope) indicating ventilation/perfusion mismatch. The VD/VT ratio and the end-tidal partial pressure of CO2 also reflect ventilatory control and ventilation/perfusion matching during exercise. Furthermore, the latter provides a reliable non-invasive estimation for arterial CO2 partial pressure. Recently, Salvioni et al. proposed an equation to predict VE/VCO2 slope in patients with HFrEF, seeking to determine whether the percentage of predicted VE/VCO2 slope could have a greater prognostic power compared to traditional VE/VCO2 slope [79]. Indeed, percent-predicted VE/VCO2 slope allowed for a refined prognostic stratification in patients with severe HFrEF (peak VO2 < 14 ml/min/kg) compared to the absolute VE/VCO2 slope value. However, such an equation has yet to find wider application, even in the research setting.

Notably, impaired ventilation and ventilation/perfusion mismatch correlated with increased B-lines in a cohort of patients across the whole HF spectrum [9].

Despite the promising evidence derived by the combined approach in patients with HF, this technique has some limitations worthy of being mentioned. Compared with rest echocardiographic and clinical evaluation, CPET-ESE is more time-consuming, more expensive, and requires specialized equipment and personnel.

4 The Transition from Subclinical Alterations to Heart Failure: The Role of Combined Cardiopulmonary Echocardiography Stress Test

Despite the importance of risk stratification and preventive interventions in hypertensive patients, and despite the tight pathophysiological connection between AH with HFpEF [18], only scarce attention has been given to the systematic analysis of the transition from one condition to the other. Moreover, many typical comorbidities of HFpEF (e.g., AH, obesity and diabetes mellitus) cluster together and are all involved in the pathophysiology of HF, making it difficult for the clinician to discern the contribution of each of these conditions and their associated cardiovascular and pulmonary alterations in the development and progression of the disease [15, 16].

As CPET-ESE is an expensive and time-consuming technique, patient selection is mandatory and should be driven by clinical judgment. Thus, hypertensive patients experiencing symptoms and signs suggestive of HF, such as dyspnea, fatigue and ankle swelling, should first undergo a comprehensive evaluation, including standard rest echocardiography and natriuretic peptides measurement (Fig. 2). HF-specific quality of life questionnaires (e.g., the Kansas City Cardiomyopathy Questionnaire), scores based on clinical and rest echocardiographic parameters (e.g., H2FPEF and HFA-PEFF [57, 80]), and submaximal exercise tests (e.g., the 6-minute walk test) can be used in clinical practice but are scarcely reliable for patients at intermediate probability. For those falling in this "grey zone", the combined CPET-ESE approach could offer a better pathophysiological characterization, supporting or ruling out the diagnosis of HFpEF. Noteworthy, subjects with resistant hypertension represent a subgroup at higher risk of transition toward HF [81] and likely candidates for the CPET-ESE examination when clinically indicated. However, it must be noted that exercise testing might be contraindicated in those with resting BP >200/110 mmHg [82].

Fig. 2.
figure 2

Tentative flow chart showing how combined cardiopulmonary and echocardiography stress test can help stratify risk and probability of HFpEF in the setting of AH. CPET-ESE derived parameters suggestive of HFpEF are not listed in the flow chart, but all are discussed within the text and summarized in Fig. 1. *According to current Guidelines: left ventricular mass index ≥ 95 g/m2 (female), ≥ 115 g/m2 (male); relative wall thickness > 0.42; left atrium volume index > 34 mL/m2 (sinus rhythm); E/e’ ratio > 9; pulmonary artery systolic pressure > 35 mmHg; tricuspid regurgitation velocity > 2.8 m/s. BNP brain natriuretic peptide, ESC European Society of Cardiology, HF heart failure, HFpEF heart failure with preserved ejection fraction, NT-proBNP N-terminal pro-B type natriuretic peptide

Full size image

In an interesting paper by Melenovsky et al., resting vascular (i.e. brachial and carotid pressures, total arterial resistance, arterial elastance) and conventional LV diastolic parameters (i.e. E-wave, A-wave, deceleration time, and e’ at both mitral annular insertions) were able to detect early cardiovascular maladaptations in hypertensive subjects with LVH compared to healthy controls [33]. Invasive hemodynamic assessment of arterial function at rest and during exercise confirmed those results, unmasking significant exercise-induced impairment in arterial compliance and vasodilatory response to nitrate infusion in HFpEF and AH [9, 83]. Such alterations can lead to a reduced peak VO2, an earlier AT and a steeper VCO2/VO2 in hypertensive patients [84, 85]. It has been recently observed that, despite peak VO2 appears more severely impaired in hypertensive subjects with HFpEF than in isolated AH, the two populations share similar signs of low peripheral oxygen extraction in terms of impaired AVO2diff [9]. However, there is also evidence of central alterations to exercise intolerance in AH, as demonstrated by a mildly reduced peak heart rate than healthy subjects [86]. Indeed, chronotropic incompetence (i.e., insufficient increase in HR during exercise) is a common feature in the HF spectrum and a strong predictor of cardiac and all-cause mortality [87, 88] Moreover, hypertensive patients present with preserved CO increase during exercise, but they show subtle alterations in regional myocardial deformation (e.g. STE-derived GLS) that are significant compared to healthy controls [9]. The correlation between STE-derived myocardial deformation parameters and impaired exercise capacity is well established [24, 33]. Nedeljkovic et al. studied hypertensive patients with and without HFpEF, showing similar echocardiographic parameters at rest in the two populations, except for a lower e’ and higher sPAP in subjects with HFpEF, suggesting that peak E/e’ could unmask HFpEF. Indeed, subjects with pathologically high E/e’ had significantly steeper VE/VCO2 slope and reduced partial pressure of end-tidal carbon dioxide (PETCO2), another marker of ventilatory/perfusion mismatch [62]. Similar data were observed by Belyavskiy et al., highlighting how the evaluation of tricuspid regurgitation velocity during exercise—and thus, indirectly, of pulmonary artery pressures—could better detect patients with HFpEF, compared with the sole peak E/e’ ratio [90]. Moreover, the mild elevation of sPAP observed during exercise in subjects with AH and HFpEF was revealed to be consistent with a steeper VE/VCO2 slope and increased peak B-lines [9].

These results suggest that exercise can elicit subclinical pulmonary congestion even in patients with isolated AH and before overt HFpEF occurs, providing a more accurate risk stratification and potentially a more personalised therapy.

Several studies analysed the impact of medical therapy on the CPET response in patients with HF [91, 92]. As far as we know, the combined approach has not been used for this purpose in HF nor AH, possibly due to its relatively recent introduction in the research setting. In the future, the application of CPET-ESE might also spread to the evaluation of the response to pharmacological therapy.

5 Conclusions

AH is a global burden and the leading risk factor for HFpEF. Given its capacity to detect early functional alterations, CPET-ESE could give more insight into the clinical stages towards HF and allow a better risk stratification in hypertensive patients. As identifying early cardiopulmonary alterations in AH may suggest novel preventive strategies and help clinicians tailor medical treatment, further investigations are needed to confirm the value of CPET-ESE in the clinical arena.