Echocardiographic Assessment of the Left Ventricle

10 Echocardiographic Assessment of the Left Ventricle



Left ventricular Assessment and Quantification


Given the prevalence of coronary artery disease, chest pain syndromes, hypertension, valvular lesions, and need to select cases for ICD insertion, assessment of the left ventricle is the principal referral reason for echocardiography in most laboratories. However, it is also one of the more difficult applications of echocardiography, because much of the echocardiographic assessment of the ventricular function is derived from subjective visual assessment. Furthermore, the three-dimensionsal (3D) geometry of the heart (which typically distorts further in disease), and the inherent limitations of depicting a 3D structure with a limited number of two-dimensional (2D) planar tomographic views, confers a fundamental challenge.


Quantification is desirable to lessen variance and to improve comparative power. However, echocardiographic quantification tools and software, other than the relatively new real-time 3D technique, are relatively unsophisticated and unsuccessful when compared to those of cardiac magnetic resonance (CMR) and cardiac CT scanning, and unsuccessful when compared to the venerable nuclear test of blood pool scanning. The fact that quantification (other than the real-time 3D technique) improves echocardiographic estimation of ejection fraction says at least as much about the lack of reproducibility of visual assessment as it does about the ability of quantification.


No method of endocardial border delineation, by any modality, is an ideal technique to determine left ventricle (LV) volumes. The topography of the endocardial surface of the LV is so irregular that modalities with poor resolution, such as multi-gated acquisition (MUGA) scanning and contrast ventriculography, falsely simplify the task. ECG-gated cardiac CT and sometimes MRI steady-state free precession (SSFP) sequences depict the endocardium so clearly that the appearance of the actual endocardial topography is initially overwhelming.



Required Parameters to Obtain from Echocardiographic Scanning


For all patients, unless the LV is solidly normal by visual assessment



image LV mass (index)


image LV volumes and ejection fraction (EF%), unless not feasible


• Biplane Simpson’s method

• Real-time 3D volumetric assessment, if available, is preferred.

• Avoid Quinones method

For any patient with coronary artery disease, or for any patient with ≥2+ mitral regurgiation or aortic insufficiency



image Biplane Simpson’s for volumetric assessment


• End-diastolic volume (EDV)

• End-systolic volume (ESV)

• EF%

image Real-time 3D volumetric assessment, if available, is preferred.


image For body surface area normalization: height, weight


image If possible, record gender and note body habitus as “small,” “average,” or “large” to provide context for interpreting mass and volume.



Left ventricular Thickness and Mass



Left Ventricular Wall Thickness




image There are several different conventions of wall thickness measurement—the standard is to use the internal interface.


image Measurements traditionally are made at the mitral leaflet tips, where the walls usually are typically of uniform thickness, not tapered.


image Avoid the right ventricular (RV) trabeculation/band common at this site of the basal anterior septum.


image Interobserver variability of wall thickness and cavitary dimensions is considerable.1


image Autopsy versus echocardiographic determination of LV wall thickness


• Note: the formalin fixation of the heart renders it contracted (systole-like). Necropsy wall thickness is the same as systolic wall thickness by echocardopgraphy,2 not the same as standard diastolic wall thickness by echocardiography.

image Wall thickness should be viewed, in general, as a shortcut to assessing for the presence of left ventricular hypertrophy (LVH). The considerable assumption in using wall thickness as a surrogate for LV mass is that the cavitary internal dimensions are normal in size.


But even in the preharmonic era, septal wall thickness ≥12.8 mm or posterior wall thickness ≥11.3 is seen by echocardiography in 3.8% of supposed normotensives.3



Left Ventricular Mass (Hypertrophy) Assessment




image LV mass is determined by both wall thickness and cavitary dimensions. Therefore, wall thickness is a relatively poor descriptor of LVH.


image LVH, by definition, is increased LV mass, not increased wall thickness, and is defined to a pathologist as LV mass >220 g.


image LVH to an imager is defined as >2 SD above the mean value for the test and method employed.


image For an echocardiographer, LVH usually is >225 g, but gender-specific definitions are appropriate, because LV mass is very clearly influenced by gender.


image LV mass correlates with many clinical factors (height, body surface area, body mass index, lean weight, skin thickness, and blood pressure). Therefore, some attempt to “index” or “normalize” LV mass for common clinical parameters should be undertaken. Generally, LV mass is indexed to either body surface area or height.


image In small ventricles, volume normalization of LV mass (LV mass/LVEDV [left ventricular end-diastolic volume]) is reasonable to describe LVH.


• Normal is 0.80 ± 15 mL/g; LVH is <0.50 mL/g.

• LV mass/LVEDV is gender independent.

• Although volume normalization is elegant, it is not widely performed.

image Wall thickness is measurable in most patients within ±10% (1 mm); therefore, as most formulas use the cube of measurements, the expected error is considerable (realistically it is about 30%; in the best case scenario it is still at least 20%).


image Severe distortion of the LV, especially from prior infarctions, renders LV mass calculations dubious.



Left Ventricular Mass Calculation and Equations


Many techniques and equations have been developed. Either they use measurements “plugged into” equations that make assumptions, or they use volumetric techniques such as Simpson’s, which are more time-consuming and technically demanding, but are better suited to hearts with abnormal geometry. The American Society of Echocardiography (ASE) equation and the Penn Formula are the most widely quoted, with the Penn formula probably the more widely used.



image ASE short-axis method, modified



image



where IVSd is interventricular septal dimension and PWd is posterior wall dimension.


image Penn method


• r = 0.96, SD = 29 g4; validated in (only) 34 autopsy cases with LVM 101–505 g

image



image Normal* mass index (g/m2)


• Males: 93 ± 22

• Females: 76 ± 18

image 1.04 is the specific gravity of myocardium used in this equation.


image 0.8 is the correction factor for the ASE equation that would otherwise overestimate the mass by 20%.


image Every millimeter of chamber inaccuracy contributes an 8-g difference to the mass.


image These equations originally derived from M-mode (at the mitral valve tip) measurements, as M-mode at the time was more accurate than were 2D measurements.


image Major, often unsubstantiated, assumptions include the following:


• Equal wall thickness throughout

• Symmetry to the ventricular geometry

image Validation4–7


• r = 0.81–0.96; standard error of the estimate (SEE) = 30–60 g (see later discussion)


Left Ventricular Hypertrophy—Increased Left Ventricular Mass




image Using the Penn method (78 men and 55 women), 2 SD above the mean:8


• LVH in men: >134 g/m2

• LVH in women: >110 g/m2

Using these paramenters, though, results in a 3% incidence of LVH in normotensive individuals.3 Changing the definition to >140 g/m2 in men reduces the incidence to 1.3%.3 As left ventricular mass index is a linear variable, a prominent effect is imparted by establishing a cut-off to dichotomize the variable as “LVH” or “no LVH.”



image Left ventricular mass index grading (g/m2)


• Male: normal, <140; mild LVH, 140–155

• Female, normal <110; mild LVH 140–155

• Moderate LVH, 155–175

• Severe LVH, >175


Should M-Mode or Two-Dimensional Data Be Used to Calculate Left Ventricular Mass?


M-mode is still a higher-resolution modality, but is subject to misalignment, which may overestimate transverse dimensions if the misalignment is significant. Generally, M-mode–derived estimates of left ventricular mass correlate only modestly with necropsy weights (r = 0.58–0.67).9


2D is lower resolution, and harmonics increase apparent thickness, but 2D measurements can be adjusted to correct for misalignment, and end up, thereby, being more reproducible.



Reproducibility Changes in Mass by Two-Dimensional and M-Mode Echocardiography10




image Modality used to determine LV mass influences variability: 2D, 115 ± 20 g versus M-mode, 127 ± 37 g


• 2D intraobserver variability: 4.2 ± 3%

• M-mode intraobserver variability: 15 ± 10%

• 2D interobserver variability: r = 0.95, P < 0.001


Reliability of Differences in Mass


When are differences in mass reliable, given the SEE?



image 2D error in tangential imaging may be no better than M-mode, and may be displaced.11


image Echocardiographic calculations of LV mass are really suited only to detect the sort of large changes in the LV that would be seen post aortic valve surgery11 rather than the small changes that happen with lifestyle or drug interventions or with detraining. MRI is able to show changes in LV mass in much smaller study cohorts than would be required using echocardiography.


image Best case scenario: LV mass change ±35 g or ±17 g/m2 has 95% or 80% likelihood, respectively, of being true.12


image Or, in a conservative approach—a 60-g difference is needed to exceed the 95% CI. This is greater than the 20- to 30-g change seen in response to antihypertensive treatment.11



Clinical Risk of Left Ventricular Hypertrophy




image For every 39 g/m2 increase in LV mass, there is a 40% increase in cardiovascular morbidity (according to the Massa Ventricolare Sinistra nell’Ipertensione Arteriosa [MAVI] Study in Italy, in which 400 males were followed over 3 years, on average).13



Echocardiography versus Electrocardiography for the Detection of Left Ventricular Hypertrophy


As noted, the incidence of LVH by echocardiography will depend on the definition. There are a host of ECG signs of LVH and left atrial enlargement that have poor sensitivity and positive predictive value (Table 10-3).


Interestingly, echocardiographic and electrocardiographic LVH predict mortality independently of each other.14



Systolic Function



Left Ventricular Segmentation


The LV is segmented, by convention shared with nuclear imaging, CT scanning, and MRI, into 17 segments.1 Abnormalities in wall motion and systolic thickening should be seen in two, preferably orthogonal, views, to be confirmed.


There is general correlation of wall motion with coronary anatomy, but individual variations in anatomy (e.g., dominance, codominance, ramus intermedius, shorter versus longer diagonal versus margin branches, previous surgical revascularization) render assumptions about the correlation of myocardial perfusion and function with coronary anatomy uncertain in some territories.



Wall Motion Score




image Grading system for each segment


• Normal/hyperdynamic = 1

• Mild/moderate hypokinetic = 2

• Severe hypokinesis/akinetic = 3

• Dyskinetic (systolic outward motion) = 4

• Aneurysm (diastolic outward position ± systolic outward motion) = 5

image Each segment should be viewed in several angles.


• Wall motion score (WMS) = sum of segment scores

• Wall Motion Score Index (WMSI) = WMS/number of segments visualized

• % normal wall motion = number of normal segments/number of segments visualized


Left Ventricular Ejection Fraction and Volume Assessment


The laboratory assessment of ventricular volumes and ejection fraction is a complex topic, as all techniques—catheterization, echocardiography, MUGA, CT, and MRI—use conventions to resolve uncertainties inherent in delineation of ventricular contours at end-diastole and end-systole. The crux of the problem is the markedly irregular topographic variation of the endocardial surface, and the geometric complexities of both ventricular cavities. To compound difficulties for echocardiogrpahy, the quality of the appearance of the ventricular border (endocardium) differs between end diastole and end systole for some modalities. End-systolic trabecular tips are generally clearer than are end-diastolic endocardial trabecular tips. At end diastole, the volume between trabeculations is large; at end systole, however, the volume between trabeculations is small. Lastly, the LV (and RV) outflow tract volumes are poorly represented by most analysis techniques.


Echocardiography, CT, and MRI share some similarities in the identification of the endocardial surface, because they all predominantly delineate endocardium. By echocardiography, CT, and MRI, the end-diastolic inner contour appears to be at the base of trabeculations; the end-systolic inner contour of the LV appears to be at the top of trabeculations, as the intertrabeculation spaces are inapparent at end systole. Thus, the volume of the trabeculations tends not to be included in the assessment of end diastole, but does tend to be included in the assessment of end systole.


MUGA and catheter ventriculography predominantly delineate blood pool. The change in the LV topography is depicted differently with MUGA (which does not visualize endocardium) and with catheterization (which faintly visualizes myocardium), than by echodardiography, CT, or MRI.


Thus, each modality to assess the LV has its own imaging-specific issues and attempts to standardize methodology using conventions: including trabeculations or not, includng papillary muscles or not, assuming the base of the LV to be the annulus or the leaflet tips, and so on.


Currently, MRI (SSFP) offers the best means to determine ventricular (including right ventricular) volumes and EF%. It shares imaging similarities with both blood-pool and endocardium-delineating methods, but has some specific issues, and does involve convention, visual editing, and therefore does have some error and variation.


Even after body surface area normalization, men have larger EDV than do women: 58 vs. 50 mL/m2, P < 0.005. Ejection fraction is not significantly different, at 69% vs. 64%.15


As an example of the differences in technique, and the inherent problems specific to different techniques, the following list presents points concerning notable differences between echocardiographic and angiographic estimates of LV volumes.



Echocardiographic Assessment of Left Ventricular Volumes




image Echocardiographic use of the mitral annulus as the base of the LV achieves a shorter long axis than does angiographic assessment.


image Echocardiography is prone to underestimating the location of the true apex, and, therefore, underestimates the true long-axis dimension of the LV (“foreshortening” due to sampling too medially).


image Echocardiographic planimetry of the endocardial surface excludes the portion of the LV within the trabecular spaces.


image Echocardiographic assessment of the LV from apical views is achieved from the same site; however, given heart motion, the optimal long-axis depiction of the LV at end systole and at end diastole may be better obtained from different sites.


image The echocardiographic apical four-chamber view tends to underestimate LV area and volume.


image The echocardiographic apical two-chamber view is much less prone to foreshortening, and to underestimation of the LV area and volume.


image Use of both the A4CV and A2CV (“biplane”) incorporates some tendency to underestimate the LV volume because of use of the A4CV.


image Echocardiographic planimetry of the endocardium is performed at the tips of the trabeculations, which excludes the intertrabecular spaces and underestimates the true LV volumes.


image Echocardiography is better at determining LV systolic volume than diastolic volume because the trabecular endocardial definition is consistently better at end systole, because approximated trabeculation tips are visually obvious. For lack of trabecular detail in diastole, echocardiography tends to overestimate diastolic volumes.


image The echocardiographic convention of inclusion of the papillary muscle bodies in both systole and diastole offsets the error in stroke volume calculation, but confers more error to actual end-systolic volume determination, as papillary muscles may occupy 10 to 15 mL of volume.



Angiographic Assessment of Left Ventricular Volumes


Angiographic assessment of the LV tends to overestimate LV volumes because of



image The inclusion of papillary muscles and mitral components in the angiographic determination of LV volume


image Projection magnification of the LV volume.


image Planimetry is performed at the deepest extent of the LV trabeculations.


image Use of a single right anterior oblique plane is less accurate than is biplane cineangiography, and overweights the visual impression toward LAD perfused myocardium.


image Traditional catheterization determinations of normal cardiac volumes, in 53 men <60 years of age, using single-plane RAO projection cineangiography, were as follows:16


• Normal EF%: 72 ± 7%SD

• Abnormal EF%: <50%

image Note that the normal stroke volume index is 45 ± 13 mL/m2.17



Radionuclide Angiography


MUGA is a reproducible test to assess the left ventricular ejection fraction. It should be emphasized that its proven merit is intertest reproducibility. However, this should not be equated with accuracy. MRI probably is more accurate.



image MUGA correlation is excellent: r = 0.94–0.98 for two calculations of the same study or first-pass versus equilibrium.


image Interobserver agreement is very also good, but not perfect (r > 0.90).


image There are some data that suggest that first-pass technique may underestimate EF%


Different modalities would, therefore, describe the same ventricle differently.



Echocardiographic Methods



Suggested Algorithm




image If obviously normal by visual estimate, perform no quantitative techniques.


image If less than obviously normal by visual estimate


• Use Simpson’s technique

• Use real-time 3D, if available


Visual Estimates


Visual estimates predictably carry the potential for significant inter- and intraobserver variability. Visual estimates of LVEF% versus objective/measured findings have been studied and tabulated:



image Performed by attendings versus fellows: 0–6.4% SD


image 95% CI of mean EF%: 5–46%


image 95% CI of mean EF% comparison: 7–36%


Intraobserver variability of visual estimates is predictably considerable and wide ranging. Thus, visual estimates are the most variable and the least accurate method of estimation of LV function, and real-world experience strongly discourages their use. Unless a laboratory validates the accuracy of their visual estimate, it provides more opinion than fact. Large inter- and intraobserver variability are depicted in Figures 10-1 to 10-3.



The Quinones Method

Related posts:

  1. The Aortic Valve
  2. Coronary Artery Disease: Ischemia, Infarction, and Complications
  3. Coronary Artery Disease: Stress Echocardiography
  4. Mitral Insufficiency

Stay updated, free articles. Join our Telegram channel
Tags:
Jun 12, 2016 | Posted by in CARDIOLOGY | Comments Off on Echocardiographic Assessment of the Left Ventricle

Full access? Get Clinical Tree
Get Clinical Tree app for offline access