In chronic mitral regurgitation (MR), symptoms do not develop until decompensation of the left ventricle (LV). Patients with MR, who are asymptomatic with normal ventricular performance, will need valve surgery at an annual rate of 10.3% [1] and early surgery is recommended for asymptomatic severe MR because of increased postoperative LV dysfunction [2] in patients with delayed surgery. However, recent reports have emphasized another process called "watchful waiting" [3] for relatively low-risk patients.

Mortality after delayed valve surgery is significantly increased with severe MR (Effective Regurgitant Orifice Area (EROA) ≥0.4 cm²) [4] and depressed LV ejection fraction (EF < 60%). Therefore, timing of the valve surgery for asymptomatic MR should be based on quantitative grading of the regurgitation severity and assessment of LV systolic dysfunction by transthoracic echocardiography [5].

Cardiac magnetic resonance imaging (MRI) is a superior method in evaluating EFs, LA volumes, and pulmonary veins (PVs) [6, 7]. Enlarged LA volumes and PVs are important components of clinical deterioration in patients with MR; including atrial fibrillation and these could be subtle markers of early deterioration in asymptomatic patients with moderate to severe MR. In the present study, we compared cardiac MRI and transthoracic echocardiography to evaluate the concealed remodeling in LV, LA, and PV; for better definitions of patients that may benefit from early valve surgery.

Table 2 demonstrates the echocardiographic findings of the study population. None of the patients had systolic pulmonary artery pressure over 40 mmHg. Mean LA EF calculated by MRI was 0.27 ± 0.13. Figure 3 shows the agreement between echocardiography and MRI. The differences between the two methods were not statistically significant for LA volume or LV EF. LA volumes calculated by cardiac MRI were larger than LA volumes calculated by echocardiography; however the differences were not statistically significant (p > 0.05) (Figure 4). There were positive correlations between LA volumes calculated by echocardiography and cardiac MRI (r: 0.940, p < 0.0001). There was a negative relationship between EROA and LA EF, but did not reach statistically significant level (r:-0.412, p: 0.071).

Echocardiographic data	Mean	(min-max)
Left atrium (cm)	4.2 ± 0.7	(3.0-6.0)
Right ventricle (cm)	2.3 ± 0.3	(1.7-2.8)
Left ventricle end-diastole (cm)	4.9 ± 0.4	(4.4-5.9)
Left ventricle end-systole (cm)	3.0 ± 0.4	(2.4-3.8)
Interventricular septum (cm)	1.0 ± 0.1	(0.8-1.3)
Posterior wall (cm)	1.0 ± 0.1	(0.9-1.3)
Mitral E wave velocity (cm/s)	108 ± 28	(67.3-190.0)
Mitral A wave velocity (cm/s)	91.2 ± 30.2	(48.0-156.0)
E/A ratio	1.2 ± 0.3	(0.7-1.7)
Deceleration time (ms)	212 ± 51	(125-300)
Isovolumic relaxation time (ms)	87 ± 32	(55-150)
Left atrial area 1 (cm2)	21.6 ± 6.3	(13-33)
Left atrial area 2 (cm2)	20.7 ± 6.8	(12-36)
Left atrial volume (cm3)	77.3 ± 38.7	(36-169)
Pulmonary vein S wave velocity (cm/s)	58.5 ± 11.0	(45-84)
Pulmonary vein D wave velocity (cm/s)	52.8 ± 13.0	(27 -82)
Pulmonary vein A wave velocity (cm/s)	37.2 ± 9.5	(26.4-64.7)
Pulmonary vein A wave duration (ms)	107.5 ± 18.4	(85-150)
Mitral regurgitation velocity (cm/s)	524.1 ± 98.9	(266-662)
Left ventricle ejection fraction (%)	60 ± 3	(60-70)
SPAP (mmHg)	24.8 ± 5.5	(18-33)

A: Atrial, D: Diastolic, S: Systolic, Left atrial area 1: From four chamber view, Left atrial area 2: From two chamber view, SPAP: Systolic pulmonary artery pressure

There was no correlation between right upper PV indices and the severity of MR. According to LV EF, right upper PV indices were found lower in nine patients who were supposed to be late for mitral valve surgery (EF < 60%), in comparison with patients with preserved EFs (0.16 ± 0.06 vs. 0.24 ± 0.08, p: 0.024). There was no significant correlation between the velocities measured with pulse wave Doppler from the right upper PV (including systolic, diastolic velocities and the proportion of systolic-diastolic flow velocities: S/D) and right upper PV index and the other PV indices (p > 0.05).

Despite the increasing tendency for early surgery in the management of asymptomatic severe MR, watchful waiting is a notable option for some selected patients [3]. Therefore, it is important to differentiate the subgroups of asymptomatic patients with MR that early surgery should be considered. This study was designed to find out if cine MRI could give us more information to evaluate this subgroup, i.e. to define patients in this group who might have irreversible changes in LV and LA that couldn't be detected by echocardiography at the time of evaluation for surgery decision.

According to the published guidelines; moderately reduced LV EF (<60%) is an indicator of mitral valve surgery decision in MR [5]. Echocardiography could sometimes fail in calculating the LV EFs, because of the 3-dimentional enlargement of the ventricle. Similarly; in 9 patients (42%); LV EFs were moderately depressed (<60%) by MRI, while all LV EFs were preserved (≥60%) by echocardiographic evaluation. The values of LV EFs (<60%) published on guidelines [5] are based on echocardiographic results and not on MRI results and for that reason maybe specific LV EF threshold values by MRI should be determined as indicators for mitral valve surgery timing. Lower LA EFs in these 9 patients could also indicate the ongoing remodeling including both LV and LA, despite watchful waiting. Additionally, reduced right upper PV indices might be the indicators of clinical deterioration including atrial fibrillation. Therefore, MRI should be preferred when small changes in functional parameters are of clinical importance to disease management like asymptomatic MR that we follow up for appropriate surgery timing [7, 16].

The advantages of MRI over echocardiography are emphasized by Bellenger at al. as better image quality and quantification possibilities with high reproducibility that requires smaller sample sizes to prove statistical significance [17]. MRI was suggested to be an accurate and reproducible method in patients with heart failure that yields serial assessment of LV remodeling [18]. Westenberg et al. searched restrictive annuloplasty results by MRI. They demonstrated significant LA and LV reverse remodeling over time using MRI [19]. Since follow-up studies performed by echocardiography are not optimal for precise assessment of LA and LV volumes, MRI was considered to be the gold standard method for assessment of LV function and volumes [20]. On the other hand, LA volume measurements in MR images were also shown to be reproducible in clinical practice according to previous studies [21] and MRI has become the gold standard for LA volume determination [22–25].

History of mitral valve disease duration sometimes does not reflect the exact period, because of the long asymptomatic clinical course of chronic MR. Hence we defined a new index; PV index that could help to define both mitral valve disease duration and the severity. All four PVs could be displayed by MRI and their indices could be calculated as the components of LA remodeling. In our study, right upper PV diameters were measured both in diastole and systole by MRI and right upper PV indices for each patient were calculated. In nine patients who were supposed to be late for mitral valve surgery according to LV EFs (LV EF < 60%), right upper PV indices were found lower by MRI. Although PV index does not give information about MR severity; it might reflect the remodeling of LA. LA remodeling is a consequence of the duration of MR and its severity, therefore it is an important determinant of hemodynamic and clinical picture in patients with MR. In a published study, the comparison of 3 D magnetic resonance angiography with 2 D cine MRI for characterizing anatomy and size of PV revealed that these two methods were similar with regard to the evaluation of PVs. In that study, reduced difference between the PV diameter at systole and diastole was observed in patients with atrial fibrillation [26]. Accordingly, reduced right upper PV index may alert us for a new atrial fibrillation that originates from PVs and points out a relatively high-risk subgroup of asymptomatic MR. It is important for watchful waiting process with MR. After atrial fibrillation, patients become symptomatic and a new stage begins with a rapid deterioration of the disease. As this index is defined for the first time, its thresholds are not clear yet, but we believe that, randomized clinical trials with large study populations can determine the thresholds for PV index.

The most important limitation of our study is the small sample size because of using too many exclusion criteria that could affect our study results. Age was not considered as eligibility criteria; however age could be an important factor for PV indices and LA remodeling. PV index is a novel parameter and a preliminary validation study could be helpful for this index. Our study could be improved by following these patients for long-term mortality and morbidity, but it was not ethical to observe patients with severe MR and reduced EF determined by MRI. Therefore, mitral valve surgery was recommended for these 9 patients with depressed ejection fractions by MRI.

In asymptomatic patients with MR, the most important issue is the decision of valve surgery. Despite the new approaches like watchful waiting process, early surgery is recommended when valve repair is easy and feasible. Evaluating asymptomatic MR by MRI is important for more favorable calculation of LV EF. Reduced LA EFs and right upper PV indices by MRI, as signs of LA remodeling might be the predictors of worsening clinical outcome of asymptomatic MR; including atrial fibrillation. We believe that, MRI should be preferred when small changes in functional parameters like LV EF, LA EF, and PV index are of clinical importance to disease management like asymptomatic MR that we follow up for appropriate surgery timing [7, 16].

All patients provided written informed consent to participate in the research.