Letter Regarding Article by Baicu et al, “Left Ventricular Systolic Performance, Function, and Contractility in Patients With Diastolic Heart Failure”
To the Editor:
This timely article by Baicu et al addresses the controversial topic of patients with diastolic heart failure (DHF), a term currently defined as heart failure in the presence of a normal ejection fraction >50%.1 Forty-seven DHF patients underwent catheterization in association with echocardiography.2 Ten patients who were catheterized had no evidence of cardiovascular disease and served as controls. From their analyses, the authors concluded that a majority of DHF patients displayed left ventricular normal systolic performance, function, and contractility.
The article’s Figures 2 and 4 appear to indicate that both endocardial and midwall fractional shortening are load independent. A reevaluation was therefore conducted and yielded the following regression equations: (1) YU=33.4–.0214 XU (n=28, NS (0.11)), (2) YL =32.7–.0600 XL (n=15, NS (0.6)), (3) Ymid=14.9−0.00314 Xmid (n=26, NS (0.76)), and (4) Ymidout=12.8–0.0092 Xmidout (n=15, NS (0.72)). Here, YU versus XU and YL versus XL refer to the upper and lower endocardial fractional shortening versus stress relationships, respectively (Figure 2), and subscripts mid and midout refer to the midwall fractional shortening versus stress relationships (Figure 4). Note the marked differences in the statistics. Furthermore, the authors need to address the following questions: Why were the control data omitted from Figures 2 and 4 and what was the significance of these control regression equations? What was the rationale for assuming no gender differences? This is an important question that needs further study.
The authors are to be commended for introducing the normalization factors that resulted in dimensionless units for Ees. However, stroke work/end-diastolic volume (SW/EDV; see Figure 1) and the end-systolic pressure/end-systolic volume ratio appeared to indicate differences between the control and DHF groups (employing data in Tables 1 and 2). No comments could be made on Ea and the Ea/Ees ratio for lack of data. Figure 3 is incomplete because SW/EDV could be load dependent and, more important, normalization of preload recruitable stroke work to mean arterial pressure leads to conflicting results.
Finally, it is hoped that these comments are interpreted as constructive in nature and that young investigators, in particular, use this article as a guideline for future studies.
We were pleased to receive Dr Mirsky’s commendation for our research.1 It is worth noting that Dr Mirsky was one of the first scientists to recognize the importance of normalizing indices of systolic and diastolic function. We simply followed his example.
Dr Mirsky correctly emphasizes the fact that all “ejection phase” indices of systolic properties are afterload dependent. Of particular importance, this fact applies to all of the more recently developed Doppler and tissue Doppler techniques used to examine regional systolic properties as well as previously defined global measurements. The mean (solid line in Figures 2 and 4) relationship between fractional shortening and systolic stress in control subjects was linear and inverse such that as stress increased, fractional shortening fell. The prediction intervals for this normal relationship (dashed lines in Figures 2 and 4) were presented to provide the “boundaries” of a normal relationship against which the data from the diastolic heart failure (DHF) patients could be compared.
As pointed out in our article and as noted by Dr Mirskyy, the mean values of the SW/EDV and ESP/ESV ratios and Ees all appear higher in patients with DHF than in controls. However, we do not think that this represents an increase in contractility in the DHF patients compared with controls, but rather reflects the presence of chronic LV remodeling. When these indices are normalized, our data support the conclusion that there are no differences in contractility between controls and DHF patients. Therefore, in patients with chronic heart disease, all indices of LV function must be “normalized” for both preload and afterload and for remodeling (both volume and mass). The resulting data provide robust evidence that LV systolic properties are normal in patients with DHF and that abnormalities in systolic properties do not contribute significantly to the pathophysiology of DHF.
SW and PRSW are useful indices of LV systolic performance and function in part because they “credit” the left ventricle for both pressure development and ejection. Therefore, appropriate components of both preload and afterload are incorporated into their calculation. However, because LV diastolic pressures are elevated in patients with DHF, it is not appropriate to use mean arterial pressure alone in the calculation of SW. Indices of “developed” not “total” pressure should be used to calculate stroke volume; in other words, mean arterial pressure should be replaced by systolic pressure minus diastolic pressure.