Evaluation of Intracardiac Filling Pressures

Definition of Filling Pressures

The term left ventricular filling pressure (LVFP) refers to the LV pressures during diastole. They are illustrated in Fig. 13.3 and include LV minimal pressure, pre-A wave pressure, and LV end-diastolic pressure (LVEDP). In the absence of mitral inflow obstruction, mean left atrial pressure (LAP) and mean pulmonary capillary wedge pressure (PCWP) are surrogates of LVFP.

LV minimal pressure occurs very early in diastole and is greatly influenced by the relaxation properties of the left ventricle. As relaxation worsens, LV minimal pressure rises. To have an accurate measurement of this pressure, the recording must be made with a high-fidelity micromanometer catheter, and for this reason, this measurement is rarely used clinically.

In the presence of sinus rhythm, LVEDP occurs after atrial contraction. It is influenced by the strength and effectiveness of the LA contraction (i.e., the amount of volume entering with the atrial kick), and the end-diastolic stiffness of the LV. With an effective atrial contraction, patients with diastolic dysfunction can maintain a normal mean LA pressure despite having an elevated LVEDP (see Fig. 13.3 ).

The pre-A pressure reflects the diastolic pressure in the LV during the passive phase of filling and is within a few mmHg of the mean LAP or PCWP. These are the pressures responsible for passive pulmonary congestion. Therefore when evaluating patients with dyspnea, the noninvasive estimation of these pressures is of greater importance than the estimation of LVEDP. Consequently, most of the discussion in this chapter will focus on ways to estimate mean LAP or PCWP. Because the pre-A LV pressure relates closely to mean LAP, this pressure should be reported when one is performing a left heart catheterization. This is discussed in greater detail in Chapter 8 .

Mechanisms of Diastolic Pressure Elevation

The role of the left ventricle during diastole is to provide an adequate inflow of volume (i.e., preload) to generate the stroke volume and cardiac output necessary to support body functions at rest and during a wide range of physical activies and heart rates, and to accomplish this while maintaining low filling pressures. The mechanisms responsible for normal diastolic function and the abnormalities that result in elevations of filling pressures are discussed in detail in Chapter 2 ; Chapters 4 and 5 discuss the determinants of diastolic flow and role of LA function, respectively. In simplified terms, early LV filling depends on a rapid and complete relaxation of the ventricle, which is an active, energy-consuming process, while filling during mid- to late diastole depends greatly on the passive elastic properties of the ventricle (i.e., stiffness) and the effectiveness of atrial contraction.

Impaired LV relaxation has been shown to be present in the vast majority of patients with conditions leading to or associated with diastolic heart failure or heart failure with preserved ejection fraction (HFpEF). These conditions are also associated with increased diastolic stiffness so that both factors become instrumental in the elevation of filling pressures either at rest or with exercise. As a rule, it can be stated that when diastolic function is normal (i.e., normal relaxation and stiffness), heart failure is highly unlikely to occur. Mean LAP may be elevated in normally relaxing hearts with pericardial constriction, obstruction to mitral inflow, or a rapid and severe increase in preload such as with acute aortic or mitral regurgitation.

If we accept the premise that patients with HFpEF have abnormalities of LV relaxation and stiffness (i.e., diastolic dysfunction), we can derive an algorithm for the estimation of LVFP by echocardiography/Doppler (two-dimensional [2-D]/Doppler) that is based on first determining if they have diastolic dysfunction with impaired relaxation and then looking for findings indicative of elevated LVFP. This chapter will review the algorithm currently recommended by the American Society of Echocardiography (ASE) that integrates findings derived from 2-D, PW and CW Doppler, and tissue Doppler recordings. We will also present an approach to managing indeterminate findings and evaluating challenging clinical scenarios. A limited discussion of new developments and recommendations on how to create a clinically meaningful diastolic function report will be presented.

Evaluating LV Relaxation

LV relaxation is an active process; consequently conditions affecting the myocardium such as pathologic hypertrophy, ischemia or infarction, excessive afterload, and any form of cardiomyopathy will result in impaired relaxation. Therefore one may have a high index of suspicion for the presence of impaired relaxation in patients with known coronary artery disease (CAD) and in those with chronic systemic hypertension (HTN). In addition, relaxation gradually declines with age so that with very advanced aging (≥80 years) the decline may be significant enough to impair diastolic filling (see Chapter 2 ).

Several 2-D findings are highly indicative of diastolic dysfunction; these include reduced EF, concentric left ventricular hypertrophy (LVH) not caused by athletic training, and regional wall motion abnormalities ( Fig. 13.4 ). In addition, LA enlargement is a sensitive marker of impaired relaxation, given that patients with diastolic dysfunction often have transient increases in LAP that serve as a stimulus for LA dilation. However, this finding is less specific as it can be seen with high cardiac output states, athletic hearts, primary mitral regurgitation (MR), and atrial fibrillation (AFib)/flutter. Eccentric LVH occurring with athletic training or with the early stages of chronic mitral or aortic regurgitations is usually not associated with diastolic dysfunction.

A detail discussion of the relation of PW and tissue Doppler velocities, and color flow Doppler recordings to diastolic filling dynamics and filling pressures, can be found in Chapters 9, 10, and 11 . Doppler measurements suggestive of impaired relaxation include reduced transmitral E/A ratio (<0.8), prolonged isovolumic relaxation time (IVRT >90 msec), reduced flow propagation velocity (Vp) by color Doppler, and reduced early mitral annular velocity (e′) by tissue Doppler (lateral e′ <10 cm/sec; septal e′ <7 cm/sec). As it has been discussed in previous chapters, these measurements are imperfect markers of relaxation, and many are influenced by loading conditions. In addition, Vp may remain normal in patients with small end-systolic volumes despite having impaired relaxation and is therefore less valuable in HFpEF. For this reason, the ASE guidelines recommend the use of multiple 2-D/Doppler findings in patients who lack clear-cut evidence by 2-D of impaired relaxation ( Fig. 13.5 ; also see Fig. 13.4 ). These are usually the ones with preserved EF and without regional wall motion abnormalities or pathologic LVH.

As can be seen in Fig. 13.5 , the proposed ASE algorithm for detecting diastolic dysfunction incorporates four parameters: septal and lateral e′ as an index of relaxation, average E/e′ ratio (E/ average of septal and lateral e′) as an index of LAP, peak velocity of TR as surrogate for pulmonary artery systolic pressure [PASP]), and LA volume. Patients with three or more normal values are classified as having normal relaxation, while those with three or more abnormal values are diagnosed with diastolic dysfunction. Patients with only two normal/abnormal parameters are classified as indeterminate in recognition that it may not always be possible to correctly assess diastolic function. This new algorithm was proposed to reduce the number of false positive diagnoses of diastolic dysfunction created by using only a reduced e′. However, by doing this, the algorithm can miss patients with impaired relaxation. A good example of such is a 72-year-old woman with chronic HTN (a clinical setting that often results in impaired relaxation), normal EF without concentric LVH or LA enlargement, a TR velocity of 2.7 m/sec, an average E/e′ of 10, and a lateral e′ of 5 cm/sec. The patient will be classified by the new algorithm as normal (i.e., 3/4 negative and 1 positive), although it is possible that she could have early diastolic dysfunction. Likewise, a patient with dyspnea who has an average E/e′ greater than 14 and LA enlargement with a normal right ventricular systolic pressure (RVSP) and low-normal e′, although classified as indeterminate, is likely to have diastolic dysfunction with elevated LVFP and may warrant application of the algorithm for estimating LVFP (see upcoming discussion). We have found that applying these concepts in routine clinical practice allows an assessment of diastolic function in over 90% of patients examined. The algorithm is limited in the presence of sinus tachycardia, marked bradycardia, or AFib.

Estimating Filling Pressures in Sinus Rhythm

Patients with HFpEF may have normal or elevated LVFP at rest depending on the severity of the relaxation abnormality, degree of chamber stiffness of the ventricle, loading conditions, and effectiveness of atrial contraction. Their diastolic pressure-volume (P-V) relation is typically shifted to the left in contrast to the marked shift to the right seen in patients with heart failure with reduced ejection fraction (HFrEF) ( Fig. 13.6 ). Thus it is not uncommon to see patients with impaired relaxation that are compensated with normal mean LAP, even though they might have elevated LVEDP. However, many of these patients experience an increase in LV diastolic pressure with exercise or excessive fluid intake (shifting from point A to B or C in Fig. 13.6 ), and as such are at risk for developing clinical heart failure. Likewise, with appropriate therapy, their P-V relation can shift from point B or C back to point A (see Fig. 13.6 ).

Once we have determined that a patient is likely to have diastolic dysfunction, the estimation of filling pressures is made by integrating several echo/Doppler findings given that no single parameter is perfect for estimating filling pressures. In the past, several equations were proposed in the literature to derive mean PCWP. Many of them incorporated multiple Doppler and 2-D measurements, and some utilized the E/e′ ratio alone or combined with other measurements. They all performed better in patients with depressed LV function and were less accurate in those with normal EF. Rather than trying to make a precise estimate of LAP or PCWP, the current ASE guidelines recommend combining Doppler findings with LA volume to classify diastolic dysfunction into three grades: grade I, which is generally associated with normal mean LAP (<15 mmHg); grade II, which is generally associated with a mild to moderate elevation (15–25 mmHg); and grade III, which is generally associated with marked elevation of mean LAP (>25 mmHg). The primary Doppler findings include the transmitral E velocity and E/A ratio, average E/e′ ratio, and peak TR velocity; other measurements such as deceleration time of the E velocity (DT), IVRT, and pulmonary vein velocity measurements can be used in selected situations (see upcoming discussion).

Pulmonary Vein Velocities

Pulmonary vein velocities are useful in estimating the LVEDP. During atrial contraction with an elevated LVEDP, LV pressure may rise over LA pressure (see Fig. 13.3 ) producing a reverse gradient that shortens the duration of the transmitral A velocity (AMV) relative to the pulmonary vein retrograde A velocity (APV) ( Fig. 13.7 ). The interval APV-AMV has been found to relate directly with LVEDP; whenever APV-AMV exceeds 20 msec, LVEDP is likely to be higher than 20 mmHg (see Chapter 9 ). This finding is fairly reliable as long as the PR interval is within a normal range (i.e., neither too short nor too long).

The value of pulmonary vein velocities in estimating mean LAP is limited. With an elevated V wave as seen in patients with grade III (and sometimes with grade II) diastolic dysfunction, systolic antegrade flow velocity (S) in the pulmonary vein is diminished, while diastolic flow velocity (D) is augmented (S/D ratio <1) ( Fig. 13.8 ). However, these changes are often accompanied by increasing E/A and E/e′ ratios so that the contribution of the pulmonary veins is diminished. Furthermore, in patients with normal EF, S/D often remains above 1 despite elevated LVFP.