Ratios of echocardiographic surrogates of LVFP to LV size
The left ventricle, left atrium and pulmonary veins are a common conduit and abnormal diastolic function can all lead to an increase in LV end-diastolic pressure (LVEDP), mean left atrial pressure (MLAP) and pulmonary capillary wedge pressure (PCWP). All these pressures are commonly referred to as LVFP(15). However, there are important pathophysiological differences between these pressures, and different pressure profiles can be drawn by invasive catheter manometry. Invasive monitoring shows three discontinuous pressure changes in LV diastole (Figure 1): the lowest or minimum pressure in early diastole, the second before the ventricular A wave, and the third at the end of diastole. LAP also changes during diastole. LVFP decreases during isovolumic diastole until it falls below LAP, prompting mitral valve opening and subsequent blood flow from the atria to the ventricle. When the two chambers reach pressure equilibrium, blood flow is minimal and LAP matches LVFP until the atria contract, producing a further pressure difference, and blood flows through the mitral valve again until pressure equilibrium is reached and the valve closes. LAP is equal to LVFP until atrial contraction and the generation of an A wave, but is lower than LVEDP after atrial contraction is complete.
The assessment of LVFP is an important component for the evaluation of LV diastolic function and chamber stiffness. Although the use of the Swan-Ganz catheter is the gold standard, several studies in recent years have found that non-invasive cardiac ultrasound measurements of filling pressures do not differ significantly from the gold standard (16, 17). Substantial evidences have proved that the ratio of early diastolic transmitral flow velocity to early diastolic myocardial velocity using tissue Doppler tracing (E/e’) correlates well with LVFP (18-24). Meta-analysis study by Rachel Jones et al. demonstrated a moderate correlation between E/e’ and invasive LVEDP (r=0.55, 95% CI 0.46-0.62, P=0.01) (25).In addition, with the addition of indices from speckle tracking technology (STE), a number of new integration metrics have been created. The pooled meta-analysis by Lassen MCH et al. showed a significant correlation between E/early diastolic strain rate (SRe) and LVEDP measured invasively (Cohen’s d=5.30 95% CI [2.83–9.96], p<0.001)(26).
According to the definition of LV chamber stiffness, researchers hypothesized that LV chamber stiffness may be indirectly evaluated by applying the ratios of echocardiographyic surrogates of LVFP to LV size (Figure 2A). In line with the hypothesis, Chowdhury SM et al.(27) found in a population of pediatric heart transplant recipients (n=18) that lateral E/e’/LVEDV (r=0.59, P<0.01), septal E/e’/LVEDV(r =0.57, P<0 .01), and (E/circumferential SRe)/LVEDV (r=0.54, P<0.01) significantly correlated with the chamber stiffness constant β, and lateral E/e’/LVEDV displayed a C statistic of 0.93 in detecting patients with abnormal LV stiffness(β> 0.015mL-1). Furthermore, A lateral E/e’/LVEDV of >0.15 mL-1 had 89% sensitivity and 93% specificity in detecting an abnormal β. Thereafter, the DPVQ, whose principle is similar to E/e’/LVEDV, is also a non-invasive parameter obtained by three-dimensional echocardiography (3DE) and Doppler tissue imaging (DTI). The LV volume was measured by 3DE and E/e’ was measured by DTI, after which DPVQ was obtained. Kasner et al.(28) applied this index to the HFpEF population and compared it with LV chamber stiffness calculated by invasive cardiac catheterization. Significant differences for DPVQ were found between 23 HFpEF patients and normal controls [0.14(0.12–0.17) vs. 0.07(0.06–0.0.09), P<0.001) and there was a significant correlation between DPVQ and LV chamber stiffness (r = 0.91, P < 0.001).
Although the ratios of echocardiographic surrogates of LVFP to LV size are currently used to evaluate LV chamber stiffness, it still remains an issue of concern that all above mentioned methods(E/e’/LVEDV, E/SRe/LVEDV and DPVQ) are surrogate and lumped measurements, relying on E/e’ or E/SRe for evaluating LVFP. However, there are several limitations and controversies for the validities of E/e’ in the assessment of LVFP. Firstly, in terms of measurement, both E and e’ are strictly limited by the location of the sample and e’ is also dependent on the angle of measurement (<20°). Secondly, the ratio is susceptible to a number of factors such as hemodynamics, myocardial synchronization, and ventricular wall segmental motion(29). Park JH et al. also suggested in their review that the use of E/e’ may be unreliable in situations such as tachycardia with fusion of E and A velocities, significant mitral regurgitation (>2+), mitral valve repair or replacement, severe mitral annular calcification, significant mitral stenosis and presence of left bundle branch block(30). Most importantly, in terms of diagnostic accuracy, a systematic review and meta-analytic analysis from Sharifov OF et al. pointed out that there was insufficient evidence to support that E/e’ could reliably estimate LVFP in patients with preserved LVEF(31). The summary sensitivities and specificities for lateral E/e’, mean E/e’, and septal E/e’ in detecting elevated LVFP were 30% and 92%, 37% and 91%, and 24% and 98%, respectively. Additionally, we reviewed 37 literatures on the correlation between E/e’ and each LVFP (LVEDP, M-LVDP, Pre-A LVP, LAP and PCWP) (supplemental material 1). In agreement with previous studies, it was found that in patients with HFrEF and HFpER, the correlation between E/e’ and each pressure varied considerably: LVEDP (0.03-0.84 vs 0.11-0.80), M-LVDP (0.40-0.52 vs. 0.49-0.60), Pre-A LVP (0.02-0.63 vs. 0.19-0.76), LAP (0.46-0.52) and PCWP (0.19-0.91 vs. 0.083-0.78)(25, 31, 32). All these disadvantages may limit these indices in clinical use. Other parameters which could reliably and accurately assess LVFP warrant further uncovered.
Encouragingly, the ratio of early filling rate derived from the time derivative of LV volume to SRe (FRe/SRe) (33) has the potential to be a surrogate marker of LVFP. It was reported that in nondilated hearts, FRe/circumferential-SRe and FRe/ area-SRe may be more useful to accurately assess LVFP than E/e’. In addition, LA longitudinal strain derived from STE is also sensitive in estimating intracavitary pressures. It is angle-independent, thus overcomes Doppler limitations and provides highly reproducible measures. Cameli M et al. found that the E/e’ correlated poorly with invasive LVFP in a group of patients with advanced systolic heart failure (r=0.15). However, the LA longitudinal deformation (PALS) correlated well with PCWP (r=-0.81, p<0.0001), and a cut-off value of less than 15.1% had a high sensitivity and specificity of 100% and 93% in predicting elevated LVFP(34). Similarly, Cameli M et al. noted in their study that both the PALS and mean E/e’ correlated well with LVEDP in patients with preserved (r=-0.79 vs. r=0.72) or mildly reduced LVEF (r=-0.75 vs. r =0.73). However, compared to mean E/e’, PALS displayed a better performance in assessment of LVEDP in patients with moderately (r=-0.78 vs. r=0.47) or severely (r =-0.74 vs. r=0.19) reduced LVEF(35). More invasive studies are yet to be performed to further explore and confirm the relationship between the ratio of FRe/SRe or PALS to LV size and LV chamber stiffness.