Key result
In patients with unexplained dyspnoea and preserved ejection fraction, an abnormal increase in pulmonary capillary wedge pressure during exercise independently predicted all-cause mortality.
Why the study?
Does exercise-induced pulmonary capillary wedge pressure elevation predict long-term mortality in patients with unexplained dyspnoea and preserved ejection fraction?
Does exercise-induced pulmonary capillary wedge pressure elevation predict long-term mortality in patients with unexplained dyspnoea and preserved ejection fraction?
Answers are grounded in this paper and cite their sources.
Early HFpEF, characterized by exercise-induced PCWP elevation despite normal resting pressures, is a distinct pathological entity associated with increased long-term mortality.
This editorial refers to ‘Pulmonary capillary wedge pressure during exercise and long-term mortality in patients with suspected heart failure and preserved ejection fraction’†, by S. Dorfs et al., on page 3103. Exercise is the most commonly encountered physiological stress in everyday life. In order for the cardiovascular system to provide up to 5-fold increases in cardiac output with exertion, there must be substantial increases in heart rate and stroke volume.1 Stroke volume enhancement is achieved through decreases in left ventricular (LV) end-systolic volume coupled with increases in LV end-diastolic volume, the latter being mediated by increased venous return and enhanced diastolic suction in the setting of a normally compliant ventricle. The fact that the left ventricle can fill to a larger diastolic volume in a shorter period of time during exercise without increasing filling pressures is an awe-inspiring testament to the remarkable diastolic reserve capacity of the normal heart.2 In people with heart failure (HF), this diastolic ventricular reserve is exhausted.3–5 Haemodynamically, HF is defined as an inability of the heart to pump blood to the body at a rate commensurate with its needs, or to do so only at the cost of high pulmonary capillary wedge pressures (PCWPs).6 These haemodynamic derangements are expressed clinically as symptoms of exercise intolerance, dyspnoea, and fatigue. These are the cardinal symptoms of HF, but they can also be caused by non-cardiovascular aetiologies. The presence of elevated PCWP in a patient with typical symptoms and normal ejection fraction (EF) is diagnostic of HF with preserved EF (HFpEF).7 However, the absence of high PCWP at rest does not exclude HFpEF, because many patients with exertional dyspnoea display normal haemodynamics at rest but profound increases in PCWP during exercise, a phenotype that has been termed ‘early’ HFpEF.3 In contrast to the more advanced stages of HFpEF, there is relatively little information available regarding early-stage HFpEF. Dorfs and colleagues now present compelling new data supporting the validity and clinical importance of early HFpEF.8 The authors retrospectively analysed single-centre data recorded over a 15-year period, including 355 patients with unexplained dyspnoea and EF >50% undergoing comprehensive evaluation including rest and exercise haemodynamic assessment, coronary angiography, and echocardiography. To enhance the homogeneity of their sample, patients with significant epicardial coronary disease (>50% stenosis), valvular heart disease (>mild), prior cardiac surgery, shunts, pulmonary arterial hypertension, cardiomyopathies, and pericardial disease were excluded. Haemodynamic data were acquired at rest and during graded exercise, starting at 25–50 W and increasing in 25–50 W increments to exhaustion. All-cause mortality was the primary endpoint. PCWP response to exercise was analysed by normalizing peak PCWP to peak work performed relative to body weight (PCWL, mmHg/W/kg) and more simply as PCWP scaled to peak work (PCWP/W).4 Over a median follow-up duration of 9.3 years, there were 58 deaths (16.3%).8 Compared with survivors, non-survivors were older, had more kidney disease, hypertension, and diabetes, and greater LV mass. At rest and particularly with exercise, non-survivors had higher right and left heart filling pressures and higher pulmonary artery pressures. Resting cardiac output was lower and pulmonary vascular resistance higher in non-survivors compared with survivors. Cardiac output was assessed using the assumed Fick technique, and thus could not be measured during exercise. Resting PCWP was associated with increased mortality, both in univariate analysis [hazard raio (HR) 1.12 per 1 mmHg increase] and in multivariable analysis, accounting for other established risk factors including age and co-morbidities (HR 1.09 after adjustment). Patients with resting PCWP >12 mmHg, the recommended partition value for positively diagnosing HFpEF,7 had > 2-fold greater risk of death (HR 2.21, P = 0.018). This is an important finding in that it further supports this haemodynamic definition of HFpEF. A sensitivity analysis using other commonly used cut-offs to diagnose HFpEF, such as PCWP ≥15 mmHg, was not performed, but would be of interest. Next the authors examined the prognostic value of the increase in PCWP with exercise. In univariate analysis, each 10 mmHg/W/kg increase in PCWP divided by peak workload performed relative to body weight (PCWL) (roughly half of 1 SD) was associated with a 35% increased risk of death.8 The predictive value of PCWL was attenuated but remained significant in multivariable analysis (HR 1.16, P = 0.03). Importantly, addition of PCWL to resting PCWP provided significant incremental predictive value, with an impressive net reclassification index of 56% (P < 0.001). Indeed, outcomes in patients with normal resting PCWP but abnormal PCWP increase (defined as PCWL >25.5 mmHg/W/kg, 32% of all patients) were not strikingly different from patients with high PCWP both at rest and with exercise (10-year mortality 28% vs. 35%), and both of the abnormal PCWP groups showed much higher mortality than the normal rest–exercise PCWP group (6.6%). Normalizing exercise PCWP to peak workload in W rather than W/kg provided similar predictive value to PCWL. The authors are to be commended on this landmark contribution to the literature.8 These data provide compelling, unequivocal evidence that the early stage of HFpEF, characterized by exercise-induced PCWP elevation despite normal resting pressures, represents a pathological entity that should be identified and treated, not only for symptom relief, but also to improve long-term survival.3,8 In addition, these data raise new questions for future research. Patients with early HFpEF may respond differently to treatments compared with those with advanced HFpEF, possibly with greater plasticity compared with advanced disease where mechanical changes in the heart and vasculature are less reversible. If this is the case, identification of early HFpEF might be important to better individualize treatment. Additional natural history data would be important to define, such as the proportion of people that go on to develop congestion and elevated PCWP at rest. Further study into the processes that govern the transition from early to advanced HFpEF would certainly be of great interest. Because exercise PCWP is mechanistically related to symptoms and independently predicts outcome, it may serve as a novel intermediate endpoint in phase II clinical trials testing new HF treatments. PCWP at rest can be quite variable, while maximal reserve responses tend to be more uniform. Indeed, as pointed out by the noted physiologist Henry Barcroft nearly a century ago, these is a striking increase in variability for any organ system when comparing measures of its resting function with measures of the maximum function that an organ can achieve under stress.9 There are some limitations to the study from Dorfs and colleagues.8 Despite a median follow-up duration of nearly a decade, the mortality rate was fairly low, certainly much lower than for patients with advanced (hospitalized) HFpEF in community-based studies.10 The data come from a single centre where diagnostic practices may differ from those of other centres, and, as acknowledged by the authors, there is bias owing to the fact that patients had been referred for cardiac catheterization, which may limit generalizability. Cause of death was not available so we cannot assess cardiac vs. non-cardiovascular modes, and other important endpoints such as changes in symptom severity, quality of life, and HF hospitalizations were not evaluated. Cardiac output reserve with exercise, which is known to be impaired in HFpEF,6 could not be assessed because of the absence of expired gas analysis, and might have provided even greater predictive insight. While the association between exercise PCWP and outcome appears indisputable, the metric applied to assess exercise PCWP in clinical practice warrants further consideration. The authors' primary analysis used the peak exercise PCWP divided by peak workload performed relative to body weight (PCWL).8 The thinking behind this approach presumably was that people with greater body weight must do more metabolic work to move the pedals at the same ergometric load. However, because this was supine exercise, where gravity is not an issue, body mass probably was not as relevant. A potential limitation of this approach is the high dependence of PCWL on body weight and workload achieved. Figure 1A shows the striking increase in PCWL assuming a normal peak exercise PCWP of 15 mmHg as body weight increases at various workloads. For example, a 120 kg person (not uncommon in HFpEF) would have a PCWL >25.5 mmHg/W/kg (abnormal according to the current study) even with a PCWP of 15 mmHg at 60 W exercise. This haemodynamic picture does not seem indicative of HFpEF. The authors also analysed the exercise PCWP response without scaling to body mass but simply to external work performed. This simpler approach is appropriate because previous studies have shown that HFpEF patients develop marked PCWP elevation even with low level activity.3,4 However, there still remains a strong workload dependence of PCW/W (Figure 1B). For example, a patient that can exercise to a peak workload of only 20 W (also not uncommon in HFpEF) would have an abnormal PCW/W even with a normal exercise PCWP of 15 mmHg. In contrast, a patient with exercise PCWP of 30 mmHg, which most would agree is abnormal, would have a ‘normal’ PCW/W ratio if they were able to exercise beyond 50 W. The increase in PCWP with supine exercise in HFpEF patients is non-linear, with 80% of the peak PCWP achieved developing during the first stage (Figure 1C).3 Thus, from an operational standpoint, it might be most straightforward simply to use the PCWP achieved at 20 W exercise, with values at end expiration exceeding 20–25 mmHg defining early HFpEF. (A) The increase in peak workload performed relative to body weight (PCWL) as body weight increases at various peak workloads achieved, assuming a peak exercise pulmonary capillary wedge pressure (PCWP) of 15 mmHg. The dotted line indicates the cut-off point identified as a potential criterion to diagnose early heart failure with preserved ejection fraction (HFpEF). (B) Showing how PCWP scaled to peak workload varies with increasing work, being more likely to be abnormal at lower workloads, even with normal PCWP values. The dotted line indicates the mean PCW/W among non-survivors in the current study. (C) Showing how the majority of elevation in PCWP during supine exercise in HFpEF occurs at low workload (shaded), with little additional increase up to peak exercise. Adapted with permission from Borlaug et al.3 Dorfs and colleagues have made a significant vertical advance in our understanding of HFpEF, showing for the first time that early HFpEF is not only an important cause of exercise intolerance, but is also a disease entity associated with increased mortality. The next steps will be to determine what causes the transition from early to advanced HFpEF, how this transition can be prevented, and how we might mitigate the pathological increase in PCWP with exercise that causes symptoms of exercise intolerance in the continually expanding cohort of patients with HFpEF. Conflict of interest: none declared.
No takes yet. Share an insight, caveat, or question.
Barry A. Borlaug (2014) conducted an editorial in Unexplained dyspnoea and EF >50% (n=355). Exercise pulmonary capillary wedge pressure (PCWL) vs. Normal exercise PCWP was evaluated on All-cause mortality (HR 1.16, p=0.03). In patients with unexplained dyspnoea and preserved ejection fraction, an abnormal increase in pulmonary capillary wedge pressure during exercise independently predicted all-cause mortality.
Synapse has enriched 10 related papers. Consider them for comparative context: