We have demonstrated that a simple, geometric-free measurement of the ratio between the pulsed wave Doppler VTI of the mitral inflow at the leaflet tips and the LVOT VTI at the aortic annulus is a highly reproducible and accurate metric for severe MR. This marker has a better diagnostic performance than traditional quantitative echocardiographic parameters of vena contracta width, PISA-based EROA and regurgitant volume. A threshold of > 1.3 demonstrated a high sensitivity and specificity for mitral regurgitation and use of this cut off would accurately reclassify 37% of patients to a diagnosis of severe MR compared to the combination of EROA and vena contracta width. We have also demonstrated that the mitral-pulmonary ratio (using the RVOT continuous wave Doppler VTI) in place of the LVOT VTI, produced similar discrimination for severe MR.

Quantitative MR grading methods especially PISA based methods and vena contracta are often regarded as ‘reference standards’ when using echocardiography and are superior to qualitative assessment. However significant limitations of the PISA method make it a less desirable metric. This includes the requirement for a dedicated imaging acquisition with optimisation of the Nyquist limit that cannot be performed in post-processing; inaccuracy when there are multiple jets, eccentric jets or non-circular regurgitant orifice shape [6]; dynamic changes in estimation related to selection of the appropriate timing in the cardiac cycle [7]; magnification of misclassification if small errors in radius measurement due to squaring of this error. Despite all these limitations, EROA and RVol remain highly specific for the diagnosis of severe MR, but their limited sensitivity does not exclude the diagnosis [4]. Often MR jets can be eccentric or there may be multiple jets which makes assessment of severity challenging [3]. Our data demonstrates high specificity for these markers, but their overall accuracy is hampered by suboptimal sensitivity, and an adequate trade-off between these measures of diagnostic performance is needed with diagnostic tests.

The mitral-aortic VTI ratio while subject to errors in tracing and variable sample volume due to translation with the cardiac cycle, may be a useful additional or alternative marker in the presence of some of these PISA limitations. It does not require additional imaging and can be performed in post-processing as it is derived from the core minimum dataset of images required in standard transthoracic echocardiography. It overcomes errors in judgement for maximal velocity by incorporating complete cardiac cycle data and given mitral measurement is in diastole, is not influenced by multiple or eccentric jet morphology. Finally, magnification error is eliminated given the relatively simple formula for calculation that relies on the assumption of proportional mitral and aortic valve area. The measurements are also easily reproducible.

The mitral-aortic VTI ratio is not reliable in clinical situations where the LVOT VTI is increased such as aortic regurgitation or dynamic obstruction. In this situation the substitution of the LVOT VTI with RVOT VTI may be useful and we demonstrated that this produced similar results with a high level of discrimination for severe MR. The use of this ratio has to our knowledge not been described previously in the literature. Pulmonary flow should be equal to aortic flow in the absence of pulmonary regurgitation or intra-cardiac shunts; therefore, the results of our study are consistent with what would be expected physiologically. Although we did not include patients with significant aortic regurgitation or mixed mitral valve disease, the mitral-pulmonary ratio has important applications in these clinical situations. The RVOT continuous wave VTI is measured routinely as part of a standard echocardiogram dataset. Mixed aortic or mitral valve disease is difficult in clinical practice to assess severity and we did not include this group of patients in our study. Further validation of the mitral-pulmonary ratio in a sub-group of patients with mixed mitral valve disease or aortic regurgitation will be required.

Other scenarios where further validation of this metric is required are in patients with multi-valvular pathologies. This parameter also is affected by elevated mitral inflow velocities not attributable to MR. In these instances, assessment of this ratio may be affected by elevation in aortic or mitral velocity-time integrals, which may confound assessment of MR severity. Despite this, we suggest that this metric has a potential utility in other situations in which MR is difficult to assess, such as detecting severe MR post mitral valve intervention. Palmiero et al. demonstrated in a small sample of patients with severe MR that there was a reduction in mean MAVIR from 1.20 to 1.01 [8]. This measure may also have a role in assessment of mitral paravalvular leak post mitral valve replacement, for which echocardiographic assessment is commonly affected by acoustic shadowing.

We noted a small difference in the cut-off for severe MR using the mitral-aortic (1.26) and mitral-pulmonary ratio (1.14). There are two potential reasons for this difference. One potential cause is tricuspid regurgitation. We excluded patients with moderate or greater tricuspid regurgitation. It is likely that mild or mild-moderate tricuspid regurgitation could have contributed to a lower cut-off point noted. Secondly these measures are ratios of mitral valve pulsed-wave Doppler VTI to RVOT/LVOT VTI. These are not equivalent to ratios of stroke volume. Based on the continuity equation, if there are differences noted in LVOT and RVOT diameter, there will be differences in LVOT and RVOT VTI despite the same stroke volume. This could also account for the differences in the VTI ratio noted.

A challenge of mitral regurgitation severity assessment using echocardiography is the high rate of discordance amongst the different parameters underlining the need for an integrated approach. A recent study by Uretsky et al. reported only 8% concordance amongst the ASE recommended parameters with severe MR and although concordance markedly improved with PISA-based EROA, RVol and vena contracta, 38% of patients still had discordant agreement of these parameters [2]. Despite their importance to decision-making, these parameters are reported in less than half of echocardiogram reports by the general medical community. While we do not suggest that the mitral-aortic VTI ratio as a sole marker for use, the ease of its calculation and high diagnostic performance compared to these traditional metrics may improve clinician distinction for severe MR and this feasibility was demonstrated by the high reproducibility in our cohort comparing expert and trainee echo readers.

Few studies have investigated this ratio but have reported similarly high diagnostic accuracy [8,9,10]. Some of these studies utilised angiographic severity of mitral regurgitation as the reference standard and/or excluded patients in non-sinus rhythm. One study by Afonso et al. in sinus rhythm patients used continuous wave Doppler signal through the mitral inflow. This similarly demonstrated high diagnostic performance with a higher ratio cut-off, however of our method of the VTI ratio to preclude geometric assumptions assumes constant mitral and aortic valve areas and hence we favoured the use of pulse wave Doppler at the mitral leaflet tips which provides assurance of flow at this point only. Similarly, a previous report by Ascione et al. sampled mitral inflow at the annulus and as expected had a lower ratio cut off to predict severe MR. Triboully et al. demonstrated a threshold of 1.3 with sampling at the valve leaflet tips similar to our findings. We build on this initial work by inclusion of a larger sample of severe MR patients, inclusion of consecutive patients irrespective of rhythm, and involvement of trainee echocardiographers to assess the feasibility and ease of this measurement holding an advantage over more complex calculations.

Our study did not utilise cardiac MRI for validation of the VTI ratios; instead the reference standard of expert echocardiographers utilising current echocardiographic guidelines was used. Cardiac MRI to evaluate mitral regurgitation severity is emerging as an important reference standard, and future validation of mitral-aortic VTI ratio with this modality using a prospective study design and a larger number of patients would further strengthen the use of this parameter as an arbiter for severity in echocardiography. The objective of our study was to evaluate a simple and easily accessible parameter to allow rapid disease discrimination. All patients with severe MR were accepted for urgent or emergent intervention by a multi-disciplinary independent Heart Team, with the degree of regurgitation and consequent management being unlikely to change with additional MRI assessment. Cardiac MRI was also not widely available during the study period and is less reliable when there are arrhythmias.

We have also demonstrated the high reproducibility of this metric, not just within observers, but between junior and expert echocardiography readers. Quantitative parameters are limited by significant inter-observer variability [11] which may have implications for treatment decisions and prognosis. The lack of dedicated quantitative measurements as standard requirements in a typical echocardiographic study may also contribute to hesitation in both acquisition and reporting of these metrics with subspecialty imaging cardiologists at best routinely reporting vena contracta width in 75%, EROA 58% and RVol 29% [12]. As mitral tip VTI and LVOT VTI form part of a routine minimum dataset [13], this metric does not require additional clips and can be easily performed in post-processing. The feasibility amongst differing levels of experience is an important advantage over PISA based techniques.

We recognise several significant limitations to our study. Our sample size, while larger than other studies in this area, is still small and larger multi-centre evaluation, with use of MRI as validation will be required to ensure results remain reproducible. We did not evaluate annular changes, or serial changes in measurements which is necessary for test-retest verification as well as usefulness for successive studies and this will be the subject of future study. We did not include patients with mixed mitral valve disease or aortic regurgitation so further validation of the mitral-pulmonary ratio in this group would be required. The association between the mitral-aortic and mitral-pulmonary ratios with remodelling parameters such as LV volume, LV mass and LA volume was not fully investigated in this study. Future research is required to demonstrate that these ratios are independently associated with remodelling parameters.