Heart Failure With a Normal Ejection Fraction
To the Editor:
In a recently published article in Circulation, Dr Zile and colleagues1 evaluated patients with a history of heart failure (HF) and a normal left ventricular (LV) systolic function. Invasive hemodynamic and/or Doppler echocardiographic indices of LV filling were abnormal in all patients, suggesting that diastolic HF can be diagnosed without the need for objective evidence of LV diastolic dysfunction.
The hemodynamic correlate of pulmonary congestive symptoms is elevated pulmonary capillary pressure (PCP) (or mean left atrial [LA] pressure). Impaired LV relaxation, by itself, is not necessarily associated with elevated LA pressure. By LV catheterization, mean LA pressure is better predicted by preA pressure than by LV end-diastolic pressure,2 the focus of the current analysis. In fact, LV end-diastolic pressures are frequently elevated in hypertrophic ventricles without elevation of mean LA pressures. Although, on average, preA pressures were elevated in this study (16±8 mm Hg), obviously many patients had preA pressures that were not elevated and, therefore, not explanatory of HF symptoms. Moreover, an “abnormal relaxation” mitral inflow pattern, the most frequent Doppler finding in this study, is typically associated with LA pressures that are not elevated.3 Thus, although various parameters of LV filling were abnormal in all patients in this study, primarily indices indicating abnormal LV relaxation, many patients did not have clear invasive or noninvasive evidence of elevated LA pressure or PCP.
Notably, there was no control group in this study (ie, patients with similar echocardiographic characteristics but without a history of heart failure). Abnormalities of LV relaxation are the rule in the presence of LV hypertrophy (present in a significant proportion of study patients). Thus, it is plausible that similar abnormalities of LV diastolic parameters could have been detected in an echocardiographically comparable control group without clinical HF.
Finally, the authors do not provide any data on the results of the patients’ coronary angiograms. It is highly likely that obstructive coronary artery disease was a common finding and may have been the cause of transient HF in many of the study patients.
In summary, although various parameters of LV filling were abnormal in all study participants, the data presented by Zile et al1 do not provide evidence that diastolic dysfunction was uniformly responsible for the clinical symptoms of HF in these patients.
The major conclusion drawn from our study1 is that the diagnosis of diastolic heart failure can be made in patients with a history of heart failure and echocardiographic evidence of a normal chamber size, evidence of left ventricular (LV) hypertrophy, and a normal LV ejection fraction; diastolic dysfunction is virtually always present in such patients, and therefore, the clinical diagnosis of diastolic heart failure does not require measurement of indices that reflect the diastolic properties of the ventricle. This information should reassure the clinician and allow investigators to proceed with therapeutic trials without dissecting out the multiple hemodynamic and other factors that influence the indices of diastolic function.
Agmon and associates correctly point out the well-known correlation between LV diastolic mean or preA pressure and left atrial pressure, and they state that “obviously many patients had preA pressures that were not elevated and, therefore, not explanatory of heart failure symptoms.” In fact, over 90% of our patients exhibited an elevated preA pressure. While we agree that these diastolic pressures (16±8 mm Hg) were not high enough to produce pulmonary venous hypertension and congestion, we emphasize that the patients had been treated and, as we noted in our article, were stable (and supine in the catheterization laboratory) at the time of the study. Likewise, the E/A ratio and other echocardiographic data at the time of participation in the research study should not be expected to parallel symptoms that were previously present.
Although a control group of asymptomatic patients with LV hypertrophy might be expected to exhibit echo-Doppler evidence of diastolic dysfunction, the abnormal diastolic pressures that we observed would not be expected in a control group. Such a group was not relevant to our research that was designed simply to assess the prevalence of abnormal diastolic function in a group of patients with the clinical diagnosis of diastolic heart failure. We found that the indices of diastolic function were virtually always abnormal in patients with the clinical diagnosis of diastolic heart failure, and we concluded that the diagnosis can be made without measuring these indices. A “control group” would not affect this conclusion.
Agmon and colleagues also ask about the results of coronary angiography. Of the 47 patients who underwent the micromanometer-echocardiographic studies, none had significant coronary artery disease. Only 8 of the remaining 16 had angiographic evidence of significant coronary disease (>70% obstruction). Thus, in most of our patients, coronary disease could not be implicated as a cause of heart failure.
In summary, the precise mechanisms leading to the congestive state in patients with diastolic heart failure are incompletely understood. However, our data indicate that the diagnosis of diastolic heart failure can be made on clinical grounds without calculating the indices of diastolic function.
- ↵Zile MR, Gaasch WH, Carroll JD, et al. Heart failure with a normal ejection fraction: is measurement of diastolic function necessary to make the diagnosis of diastolic heart failure? Circulation. 2001; 104: 779–782.