Afterload in MR

The LV is capable of maintaining stroke volume in chronic MR until advanced stages of the disease because, following the Frank-Starling law, diastolic lengthening increases shortening of the sarcomere. If the regurgitant leak were unrestrictive, this compensating mechanism would be insufficient to maintain forward flow because the preload-recruited stroke volume would be mostly pumped towards the left atrium. In chronic MR, the regurgitant lesion always imposes a significant opposition to backflow. Although its area may vary during systole, in fluid-mechanical terms, the regurgitant lesion in MR behaves as a small, flat and restrictive orifice. Very large free regurgitant lesions, as sometimes found in severe tricuspid insufficiency—showing a laminar regurgitant flow and a very low transtricuspid pressure gradient—are virtually never seen in chronic MR. Furthermore, even in advanced stages of the disease, any decline in cardiac output is compensated by vasomotor reflexes. Consequently, systolic blood pressure remains normal in chronic MR.1

Ventricular pressure translates to afterload by means of wall-stress. Governed by Laplace's law, systolic wall-stress is directly proportional to systolic LV pressure. But systolic wall-stress is also directly proportional to the degree of wall curvature (radius), and inversely proportional to wall thickness. Eccentric remodelling in chronic MR dilates the ventricle and makes the walls thin, and this is why LV afterload is actually above normal values in chronic disease.

In their Heart manuscript, Gaasch et al.3 provide some supporting numbers to banish the misconception of low ventricular afterload in MR. The authors elaborate on a previous theoretical framework2 and take the concept of ‘impedance’ from other systems to apply it to the mitral valve. Using simple arithmetic, they calculate the opposition to flow imposed by the MR and compare it with ‘aortic input impedance’. Their major conclusion is that impedance to retrograde flow in MR is similar or even larger to impedance to forward flow, rejecting the unsupported idea that MR acts as an easy ventricular outlet opposing little resistance to flow. Their measurements confirm normal values of LV systolic stress in patients, even in the presence of low values of total impedance (calculated as an equivalent of aortic and MR parallel resistances). The merit of this study is to provide a simple mathematical context to show that the opposition to backflow towards the atrium is at least as high as the opposition to forward flow towards the aorta.

Several issues need to be taken into account when considering the accuracy of these results and, more importantly, when extrapolating them to the clinical setting. First, left atrial pressure must be taken into consideration in the impedance model, because the haemodynamic burden depends on the area of the regurgitant office and on atrial pressure (figure 1B). Second, and more clinically relevant, using constant values of regurgitant times and a fixed value of LV ‘systolic’ pressure for their simulations is an oversimplification. The regurgitant period of MR spans the ejection period as well as the isovolumic (albeit not isovolumic in the presence of MR, see below) contraction and relaxation periods (See figure 1A,C). Thus, mean LV pressure during the regurgitation period is markedly lower than the value of mean LV pressure measured exclusively during ejection (see figure 1A, B, E). In fact, equalisation of regurgitant and ejection mean LV systolic pressures is the mechanism by which percutaneous edge-to-edge mitral valve repair abruptly rises systolic wall-stress (immediately after closing the regurgitation),4 potentially causing an acute afterload mismatch in very poor and dilated ventricles already supporting an extreme wall-stress at baseline.5

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Haemodynamic changes induced by progressive effective regurgitant orifices (ERO) of mitral regurgitation (MR). (A) Left ventricular (LV) pressure-volume loops. The disappearance of isovolumic phases is visualised as the progressive angulation of the vertical segments of the loop of ERO=0.0 cm². Bullets depict the onset and end of the regurgitant period. (B) LV (solid lines) and atrial (dashed lines) pressure evolution with time. (C) Transmitral flow velocity, showing progressive shortening of the regurgitant time with increasing degrees of MR. (D) Sarcomere strain; dashed lines account for isovolumic chamber periods. When ERO=0 cm², isometric contraction takes place with a flat strain and relaxation takes place during sarcomere lengthening (strain is increasing). MR causes both phases to take place while the sarcomere is shortening (strain is decreasing). (E) Values of mean LV pressure during the ejection and regurgitation phases. Simulations were performed using CircLab software,12 at 75 bpm, and set to autoregulate for a mean systolic arterial pressure of 92 mm Hg in all cases.

Another limitation of the study by Gaasch et al.3 is that support as their study on what they define as mitral valve impedance. Although useful for obtaining numbers in the same units aortic input impedance, impedance is an inappropriate metric to describe MR. Flow impedance is used in fluid dynamics to characterise long tubes. In tubular systems, the pressure/flow relationship is linear because the major source of energy loss is friction due to viscosity. Longitudinal impedance accounts for total energy losses along the tubular system (considering the pressure difference between the tubular inlet and outlet), whereas input impedance designates the opposition to flow the tubular system imposes at its entry (dismissing outlet pressure). Moreover, when applied to a pulsatile system as the arterial tree, impedance must be measured in the frequency domain by obtaining the full-band spectra of pulsatile pressure and flow. The calculation of the average pressure flow in the time domain should not be denominated impedance, but flow resistance instead as it only accounts for the continuous (non-pulsatile) component.

Physical principles that govern flow along tubes are not applicable to heart valve stenoses or regurgitations. In planar orifices, it is the spatial acceleration of flow component that governs energy losses. The source of the pressure drop at both sides of the lesion is the acceleration of flow as the jet crosses the restrictive orifice, being friction losses negligible. Governed by Bernoulli's equation, the pressure/flow relationship in this scenario is quadratic instead of linear. The fact that the pressure gradient is proportional to the square of velocity is the basis of the simplified Bernoulli's and the Gorlin's formulae used in every day clinical practice.

Although unsound, using impedance and resistance to characterise valve lesions is not new. A potential of aortic valve resistance was suggested in the 90s,6 ,7 but time has proved there is no clinical advantage over conventional indices such as pressure gradient or valve area to describe aortic valve stenosis. More recently, the concept of valvular-vascular impedance (Zva) was introduced to account for the total ‘load’ imposed to the LV by aortic stenosis and the vascular system as two resistances in series.8 A number of studies have reported a potential of this index to predict outcomes in particular subgroups of patients. Zva has the simplicity of providing a single number, merging the haemodynamic significance of valvular and vascular compartments in a single index. However, the value of Zva for clinical decision making is hindered due to conceptual limitations. First, it again misleadingly defines a linear pressure/flow relationship governing aortic valve haemodynamics. Second, Zva adds little light to the complex interaction between valvular and vascular compartments, because interventions that lower the load of one of them reciprocally increase the load of the other one, without modifying Zva at all.9 Finally, although both meaningful in the pathophysiology of aortic stenosis,10 vascular and valvular dysfunctions have different aetiological bases and require different therapeutic approaches (eg, antihypertensive drugs or valve replacement).

In summary, the biomechanics of valvular lesions and their haemodynamic consequences are best characterised by orifice area, pressure gradient, regurgitant volume and LV wall stresses. As in the study from Gaasch et al., ‘valve impedance’ is useful to didactically compare ‘loads’ in generic terms. Beyond this purpose, the use valve impedance is not recommended because it is not a physically meaningful metric neither of valve stenosis nor regurgitation.

The loss of isovolumic phases in MR

Defining MR as ‘a pure volume overload’ condition also incompletely accounts for all its consequences on LV wall mechanics. In the normal heart, isovolumic contraction allows the ventricle to pressurise while performing no external work. At the cellular level, force is built at constant sarcomere length during this phase by rapidly recruiting actin-myosin bridges.11 It is during ejection that the LV accomplishes external work as the sarcomere shortens by sliding its actin-myosin complexes. During diastole, physiological relaxation (detachment of actin-myosin complexes) takes place either at constant (isovolumic relaxation phase) or increasing (rapid-filling phase) sarcomere length (figure 1D).

A paradigmatic feature of MR not present in any other volume-load condition is the abolishment of both isovolumic phases from the mechanical cycle of the LV. Without an isovolumic phase during early-systole, the LV loses its ‘functional skeleton’ that endures wall contraction. At the onset of systole, MR precludes isovolumic pressurisation, forcing the sarcomeric motor to build actin-myosin bridges and shorten simultaneously. Ventricular emptying into the left atrium also abolishes the isovolumic relaxation phase, forcing the sarcomere to relax while it is still shortening. How these drastic alterations impact the mechanical efficiency of myocyte remains unknown. However, it is unlikely that its highly developed mechanoreceptor systems remain insensitive to these disturbances. As learned from the resynchronisation field, it is highly plausible that the passive changes MR induces in sarcomeric strain impact a number of metabolic routes in the myocyte.

In conclusion, the concept of chronic MR as a disease of pure high-preload and low-afterload is misleading. Afterload is actually normal or above normal in chronic phases of the disease. In addition to increased preload, the abolishment of isovolumic phases is the source of singular biomechanical consequences on the LV. These issues need to be taken into account for a deeper understanding of the ventricular compensatory mechanisms and for making the right clinical decisions in a time of rapidly evolving therapeutic opportunities.