Background: Left heart disease (LHD) is the most common cause of pulmonary hypertension (PH) and is associated with poor patient outcomes, especially among patients undergoing heart transplant evaluation.
Implications for clinicians: Left heart disease should be considered in all patients undergoing an evaluation for PH. Correct management of PH from LHD is to optimize treatment of LHD. Pulmonary vasodilators used to treat pulmonary arterial hypertension should not be used in patients with PH from LHD.
Conclusions: Additional research is needed to better understand how PH develops in patients with LHD and to investigate the role for treatment targeting PH in these patients.
Left heart disease (LHD) is the most common cause of pulmonary hypertension (PH), and occurs in patients with heart failure with reduced ejection fraction (HFrEF), heart failure with preserved ejection fraction (HFpEF), and valvular heart disease (Figure 1).1,2 The presence of PH in patients with LHD is associated with reduced exercise tolerance and reduced survival, especially following heart transplant.3–10 Identifying LHD as the cause of PH is critically important because it determines the correct approach to management, which is optimal treatment of the underlying LHD with evidence-based and/or standard-of-care pharmacologic or surgical therapies. In patients presenting with PH-LHD, there is currently no role for treatment with pulmonary arterial hypertension (PAH)–specific therapies and, with few exceptions, they should not be administered because they are costly, lack efficacy, and in some cases, are known to increase morbidity and mortality.
NOMENCLATURE, CLASSIFICATION, AND DEFINITIONS OF PH-LHD
The nomenclature that has emerged to categorize patients with PH-LHD attempts to describe the clinical context, pathophysiology, and hemodynamic features seen in these patients. This has resulted in a wide variety of terms used in an effort to accurately describe patients with PH-LHD. Multiple different terms, sometimes used in combination, may be appropriate to describe the unique characteristics of an individual patient with PH-LHD. Adding to this complexity is changes in terminology that have occurred over time as our understanding of PH-LHD has evolved. Recent guidelines attempted to simplify this language and to classify patients with PH-LHD according to hemodynamic characteristics (Table 1).
The hemodynamic definition of PH is a sustained elevation in mean pulmonary artery pressure (mPAP) ≥25 mm Hg. The usual hemodynamic findings in a patient with PH-LHD are mPAP ≥25 mm Hg in combination with elevated left heart filling pressures, defined as a pulmonary artery wedge pressure (PAWP) >15 mm Hg or left ventricular end diastolic pressure (LVEDP) >15 mm Hg. The clinical characteristics (ie, presence of reduced ejection fraction [EF], clinical features of HFpEF, presence of valvular heart disease) are considered together with the hemodynamic features to arrive at a final diagnosis of PH-LHD.
In patients with PH-LHD, it is important to characterize elevated mPAP as resulting only from passive transmission of elevated left heart filling pressures proximally into the pulmonary circulation vs increased pulmonary arteriolar resistance resulting from changes in the function and structure of pulmonary arterioles. The terms used to describe patients in whom PH results from transmission of elevated left heart filling pressures include pulmonary venous hypertension and passive PH. The most recent guidelines recommend using the term isolated postcapillary PH (Ipc-PH) for this group of patients. In these patients, reduction of left heart filling pressures to normal also reduces mPAP to normal.1,11
In other patients, the elevated mPAP is not fully accounted for by passive, proximal transmission of elevated left heart filling pressures. In these patients, it is believed that long-standing elevations in mPAP result in part from vasoconstriction and remodeling of the pulmonary arterioles so that mPAP is elevated out of proportion to the PAWP. Terms that have been used to describe this hemodynamic profile include mixed and out-of-proportion PH. The most recent guidelines recommend using the simple descriptive term “combined post-capillary and precapillary PH” (Cpc-PH).
Differentiating Ipc-PH from Cpc-PH is important because it has prognostic implications, especially in patients undergoing evaluation for heart transplant (Figure 2). Typically, PH-LHD has been characterized by measures of resistance and pressure difference across the pulmonary vasculature (Table 2). Patients have been considered to have Ipc-PH if the pulmonary vascular resistance (PVR) and transpulmonary gradient (TPG) are normal and Cpc-PH if the PVR and TPG are elevated. Recent guidelines support the use of the diastolic pressure gradient (DPG) to differentiate the hemodynamic subtypes of PH-LHD.11 The rationale for this recommendation is that the DPG is less dependent on stroke volume and left atrial pressure, and it was shown to be predictive of survival and correlated with pathologic changes.12,13 However, the DPG is subject to error14,15 and the association with survival is inconsistent, a reminder that a single measurement is rarely useful to characterize patients with PH.16–18
Further hemodynamic characterization of Cpc-PH is guideline-recommended in patients with LHD-PH being considered for heart transplantation because it identifies patients at risk for post-transplant right ventricular (RV) failure and death.19,20 Vasodilator studies are conducted with a right heart catheter in place. A rapidly acting vasodilator is infused, typically nitroprusside, and measurements of the PVR and TPG are made.1 Among patients in whom the PVR and TPG are reduced to normal levels while maintaining a systemic systolic blood pressure of ≥85 mm Hg, PH is considered to be reversible or reactive. In these patients post-transplant mortality is similar to patients without PH.21 Among patients in whom PVR and TPG cannot be reduced to normal, PH is considered not acutely reversible. In many of these patients, the PVR may be lowered or become reversible after prolonged reduction of PAWP with aggressive treatment with diuretics, vasodilators, inodilators, and mechanical support so that patients can become eligible for heart transplantation.
Accurate prevalence estimates for PH-LHD are limited by factors such as reliance on echocardiographic assessments of pulmonary artery pressure (PAP) to identify affected patients,22,23 and inconsistent definitions and cutoffs to diagnose PH-LHD. Studies that make use of gold-standard invasive hemodynamics may be affected by referral bias since sicker patients are likely referred for right heart catheterization (RHC). Even well done invasive studies only provide data at a single time point while the patient is at rest, fasting, and possibly sedated—all of which may affect hemodynamic measurements.24
Estimated rates of PH in LHD vary widely. The prevalence of Cpc-PH has ranged from 25% to 47% in hospitalized patients, and was 40% in a recent large ambulatory HFrEF population.10,25,26 Among patients with HFpEF, PH is present in 36% to 83%. A large community study examined echocardiograms from 244 patients with HFpEF and 719 hypertensive controls. Pulmonary hypertension, defined as pulmonary artery systolic pressure (PASP) >35 mm Hg was found in 83% of HFpEF patients compared to only 8% of controls. Using a higher PASP cutoff of 45 mm Hg would have resulted in prevalence of about 50%,7,27 a rate similar to that found in another study of 299 patients with HFpEF.28 A single-center registry of patients undergoing RHC found PH, defined as PVR >2.5 or TPG >12, in 69% of HFpEF patients evaluated.29
Left-sided valvular heart disease is also commonly associated with PH and is important to recognize because it is an indication for valve replacement or repair. Mitral stenosis is the valvular lesion most often associated with PH, at a rate of up to 73%.30 Pulmonary hypertension occurs at lower rates in patients with mitral regurgitation (23%–44%) and aortic stenosis (29%–47%).31–37
Compared to patients with LHD and no PH, patients with PH-LHD have worse outcomes, including worse survival.4,27,38–41 This is true in both HFpEF and HFrEF and in studies using both echocardiography and invasive hemodynamics to diagnose PH.41 Additionally, survival worsens as PAP increases. A study of patients undergoing endomyocardial biopsy showed a 25% increase in the risk of death for every increase of 5 mm Hg in mPAP.4 Patients with Cpc-PH generally have more severe hemodynamic impairment and worse prognosis compared to Ipc-PH.4,10,26
PATHOBIOLOGY AND PATHOPHYSIOLOGY
Our understanding of the pathophysiology of PH-LHD has improved in recent years; however, significant gaps remain. It is believed that the first event in the development of PH-LHD is increasing left heart filling pressures and pulmonary venous hypertension. Even when left ventricular (LV) systolic function is normal, diastolic filling abnormalities may result in increased PAP.42 Additionally, elevated left heart filling pressures also reduce compliance of the pulmonary vasculature and increase RV afterload by enhancing pulmonary artery wave reflections.43 Next, pulmonary arteriolar vasoconstriction occurs secondary to endothelial dysfunction characterized by decreased production of and/or decreased responsiveness to nitric oxide (NO), as well as overproduction of endothelin-1 (ET-1), activation of the reninangiotensin-aldosterone system (RAAS), and neurogenic activation. Elevated PAPs lead to injury, followed by pathologic remodeling of the pulmonary arterioles including muscularization, medial hypertrophy, and neointimal proliferation.44–49 Ultimately the increased afterload imposed on the RV leads to RV systolic dysfunction and failure.3,50
ASSESSMENT AND DIAGNOSTIC APPROACH
Making a diagnosis of PH-LHD is challenging because symptoms are nonspecific, diagnostic tests can be difficult to interpret, and PH may be multifactorial. Consequently, PH-LHD is often incorrectly diagnosed and treated as PAH, especially in elderly patients.51,52 A thoughtful and comprehensive approach to the evaluation of PH is needed so that appropriate treatment can be chosen.
History and Physical Examination
Information gathered during a comprehensive history and physical examination is important because it is used to prioritize next steps in the diagnostic evaluation and to provide context for the interpretation of diagnostic testing (Table 3). Details about congenital heart disease, murmurs, valvular disease, HFrEF and HFpEF, and coronary artery disease (CAD), as well as an assessment of risk factors for LHD should be ascertained. Special attention should be paid to factors associated with HFpEF29 such as female gender, advanced age, diabetes, hypertension, CAD, arrhythmias, and sleep-disordered breathing. Orthopnea and paroxysmal nocturnal dyspnea in a patient with PH strongly suggest the presence of LHD. Physical examination findings that point toward LHD include pulmonary crackles, left-sided S3 or S4, left-sided murmurs, or irregular heart sounds consistent with arrhythmia.
An electrocardiogram and chest x-ray should be performed in all patients undergoing evaluation for PH. Though insensitive and nonspecific, these studies may point to LHD as a cause of PH with evidence of left heart disease such as evidence for myocardial infarction, abnormal heart rhythms, cardiac chamber enlargement or wall thickness, pulmonary edema or congestion, and the absence of parenchymal lung disease.
Echocardiography is the most useful noninvasive modality for the evaluation of PH. It is easy to obtain and may immediately point to LHD as a cause of PH (Table 4, Figure 3). Echocardiographic findings of diastolic dysfunction are well described,53 but are potentially insensitive for the diagnosis of HFpEF.27,54,55 Therefore, HFpEF should be suspected when findings such as LV hypertrophy and left atrial enlargement are present.29 In a community-based study of 244 HFpEF patients and 719 hypertensive controls, elevated PAP on echocardiography was both sensitive and specific for the diagnosis of HFpEF,27 suggesting that the presence of PH on echocardiography is often itself evidence of HFpEF.
Echocardiographic estimation of PASP is the most commonly used method to assess for PH. To estimate the PASP, the tricuspid regurgitant (TR) jet is imaged and interrogated with spectral Doppler in multiple echo windows, and the peak TR jet velocity is determined. The PASP is calculated using the modified Bernoulli equation: PASP=4(V2)+right atrial (RA) pressure. This technique is limited because it cannot be utilized in patients without an adequate TR jet and spectral Doppler signal.9 Errors in estimation of PASP may lead to important misclassification of PH.23
The shape of the ventricular outflow tract Doppler signal is useful to differentiate patents with PH-LHD. Transient flow deceleration in the right ventricular outflow tract during systole is caused by early return of reflected pulmonary arterial waves, resulting in notched pattern of the Doppler signal. Early wave reflection occurs in the setting of elevated PVR, and notching occurs earlier in systole as PVR increases. Thus, the presence of PH without notching strongly favors a diagnosis of PH-LHD, specifically Ipc-PH.56
Echocardiography is also useful to assess for RV dysfunction, which is an important marker of increased mortality.3,57 Abnormalities of RV size, thickness, and function also provide evidence of clinically significant PH when the PASP cannot be estimated, or indicate that the severity of PH is worse than the estimated PASP suggests. Measurements including RV fractional area change, tricuspid annular systolic plane excursion (TAPSE), tissue Doppler imaging of the tricuspid valve annulus, and strain analysis may all be useful to assess RV systolic function and have prognostic value.57–59 These measurements have important limitations so that the overall impression of an experienced echocardiographer is important.
Magnetic Resonance Imaging
In patients with suspected PH-LHD, cardiac magnetic resonance imaging (CMR) is useful to detect structural abnormalities of the LV and left atrium, LV systolic function, presence of congenital heart disease, and presence of myocardial fibrosis or infiltrative disease. Similarly, RV enlargement, hypertrophy, and systolic function are best determined by CMR.
Right and Left Heart Catheterization
Data from an optimally performed RHC must be incorporated together with the patient's clinical characteristics and echocardiographic data to arrive at a final diagnosis of PH-LHD. In patients with PH-LHD, data from the RHC are also useful to optimize medical management and are necessary to assess risk in patients being considered for transplantation and mechanical circulatory support. It is crucial that the procedure be performed correctly and that data are properly collected and interpreted.60,61
A complete hemodynamic assessment includes measurement of RA, RV, pulmonary artery and pulmonary artery wedge pressures, and cardiac output. All pressures should be determined at end expiration during spontaneous breathing to minimize the effects of intrathoracic pressure variation on the measurement. Thermodilution cardiac output when measured in triplicate and injected during end expiration62 remains valid even in the setting of low cardiac output and severe TR.63
Incorrect determination of the PAWP may be secondary to many errors, including improper transducer position and zeroing to atmospheric pressure, waveform dampening, incompletely wedged catheter, placement in the RV, measurement that is not at end expiration, and use of the electronic mean obtained from the computer monitor. Mitral regurgitation causes large “v” waves that can be mistakenly interpreted as an elevated PAWP, confounding the calculation of the TPG and DPG. This can be accounted for by reading the PAWP at the time of the “a” wave. Several recent studies have examined the relationship between the PAWP and LVEDP and found that a substantial percentage of patients with PAWP ≤15 had LVEDP >15, which could lead to misclassification of patients with PH-LHD as PAH. Measurement of the LVEDP should be considered when a reliable PAWP tracing could not be obtained or the value of the PAWP is inconsistent with the expected value based on the clinical picture. Measurement performed manually on pressure tracings at end expiration is most tightly correlated with LVEDP.64
Procedures in the catheterization laboratory are performed while patients are at rest in a fasting state and often after sedating medications have been administered. Hemodynamic findings of LHD may not be apparent under these conditions so that a diagnosis of LHD could be missed and patients are inappropriately diagnosed with PAH.65 To uncover hemodynamic abnormalities consistent with LHD, measurement of the PAWP after provocative testing with exercise or volume loading can be considered,61 especially in the setting of a clinical history or echocardiographic findings associated with HFpEF.61 In one recent report, after fluid challenge, 46 patients (22.2%) originally classified as having PAH were reclassified as having pulmonary venous hypertension, most secondary to LHD.66
MANAGEMENT OF PH-LHD
With few exceptions, the appropriate therapy for PH-LHD is optimizing treatment of the underlying LHD. In the case of HFrEF and HFpEF, therapy should include guideline-recommended treatments with diuretics, vasodilators, and neurohormonal antagonists, as well as with device and surgical therapies when appropriate.67 The benefits of these therapies were emphasized in a recent study showing that adjustment of diuretics and vasodilator agents in response to data from continuous PAP monitoring devices reduced heart failure hospitalizations.68 Comorbidities that may contribute to PH such as sleep apnea, pulmonary embolism, and chronic obstructive pulmonary disease should also be identified and aggressively treated.
Increased morbidity and mortality associated with PH in patients with LHD makes PH an attractive therapeutic target. However, despite beneficial acute hemodynamic effects and small studies with phosphodiesterase type 5 (PDE5) inhibitors that have shown improvement in exercise capacity, no study has shown PAH therapies to be beneficial in PH-LHD, and some PAH therapies have been associated with significant adverse effects including increased mortality.1,8,69 Most studies of PAH therapies in LHD have not specifically enrolled PH-LHD, so it is possible that undetected benefits will be found in future trials.
Data demonstrating acute hemodynamic improvements including reduced PAWP, PVR, and increased cardiac output provided the rationale for the Flolan International Randomized Survival Trial (FIRST) of epoprostenol in HFrEF. However, the FIRST trial was stopped early when a trend toward increased mortality in the epoprostenol group was identified.70 Multiple trials of endothelin receptor antagonists for the treatment of HFrEF have been performed and shown either no improvement or worsening edema and hospitalization. Several of these negative studies have never been published.71–75
Phosphodiesterase type 5 inhibitors also have been shown to have beneficial acute hemodynamic benefits including improvements in gas exchange, skeletal muscle function, diastolic function, and RV function69; reduced PVR and TPG; increasing cardiac output; improvements in peak oxygen consumption and 6-minute walk; and decreased heart failure hospitalizations.76,77 Similar improvements have been shown in patients treated with the soluble guanylate cyclase stimulator riociguat.78 In clinical practice, sildenafil also decreases PVR and improves RV function after heart transplantation and LVAD implantation.79,80 Despite these encouraging findings, long-term benefits of treatment with these agents have not yet been demonstrated in a randomized controlled trial so that they should not be routinely prescribed in patients with PH-LHD. Additional trials are now underway.
Prolonged treatment with intravenous vasodilators and mechanical support may restore vasodilator response in patients with HFrEF found to have initially irreversible Cpc-PH during heart transplant evaluation.81 In several reports, LV assist device support has been shown to be effective in reversing PH permitting heart transplant without increased rates of RV failure or death.79,82,83
Much remains unknown about PH-LHD. An improved understanding of triggers and development of vascular changes in PH-LHD as well as the relationship between the LV and the pulmonary vasculature is needed. Failure of studies to demonstrate beneficial long-term outcomes in PH-LHD patients treated with pulmonary vasodilators suggests that PH in this setting may be a marker of severe or inadequately treated left heart failure. Studies that are specifically designed to focus on pulmonary vasodilators in PH-LHD patients with optimally managed LHD are needed.
Disclosure: The authors have nothing to disclose.