Prediction of mitral regurgitation severity one year after successful catheter ablation for lone long-standing atrial fibrillation

Yasunobu HARUKI; Hironori MURAKAMI; Seiko OONO; Seiichiro SAKURAI; Daisuke HOTTA; Masanori NOJIMA; Gregory W. RUHNKE

doi:10.14932/jamt.21-114

Abstract

Background: Long-standing atrial fibrillation (LSAF) is a risk factor for mitral regurgitation (MR). Conversion to sinus rhythm can be achieved by radiofrequency catheter ablation (RFCA), which is also expected to reduce the severity of MR through reverse remodeling of the left atrium (LA) and the mitral apparatus. However, the severity of MR one year after RFCA is often unpredictable. Methods: To create an appropriate model for the prediction of MR severity one year after successful RFCA, we assessed 50 patients with lone LSAF by transthoracic two-dimensional echocardiography prior to RFCA and one month and one year after RFCA. On the basis of the data obtained from the assessments, three prediction models for MR one year after RFCA were specified and analyzed using multivariate linear regression. Results: A prediction model for MR using variables, such as LA volume index and MR severity in pre-RFCA, and left ventricular ejection fraction, MR severity, and the ratio of the diameter of the anteroposterior annulus to that of the transverse mitral annulus one month after RFCA, best matched the demonstrated MR severity one year after RFCA (sensitivity, 1.00; specificity, 0.795). Excellent relationships between the predicted MR severity and the demonstrated MR severity (r = 0.732) and between changes in the demonstrated MR severity and changes in the predicted MR severity were found.

Translated Abstract

背景：長期持続性心房細動（LSAF）は，僧帽弁逆流（MR）の増悪因子である。高周波カテーテルアブレーション（RFCA）により洞調律に復帰することで，左心房（LA）と僧帽弁下組織の逆リモデリングを介したMR重症度の軽減が期待される。しかし，RFCAの1年後のMR重症度を予測することは難しい。今回，RFCAを施行後，洞調律が維持された基礎疾患のないLSAF患者で，1年後のMR重症度が予測可能か検討した。方法：RFCA成功1年後のMR重症度予測モデルを作成するため，孤立性LSAF患者50名のRFCAの前，1か月後，および1年後の心エコーデータを解析した。これらのデータを基礎に重回帰分析にて3つの予測モデルを作成して予測精度を検討した。結果： RFCA前のLA volume indexとMR重症度，RFCA後1か月の左室駆出率，MR重症度と僧帽弁輪縦横比を用いたモデルが，RFCA 1年後のMR重症度予測に最良であった（感度1.00，特異性0.795）。この予測モデルによるRFCA1年後のMR重症度の予測値と実測値の間（r = 0.732）とRFCA 1年後のMR重症度の予測値変化度と実測値変化度の間（r = 0.822）に非常に良好な相関関係を認めた。結論：RFCAによる洞調律達成1年後のMR重症度予測モデルを，心エコーデータを使用した重回帰分析により作成できた。

Introduction

Long standing atrial fibrillation (LSAF) and chronic AF are important causes of functional mitral regurgitation (MR) due to remodeling of the left atrium (LA) and the mitral apparatus.^1)–3) Radio-frequency catheter ablation (RFCA) is an important modality to treat LSAF, which improves patients’ prognosis^4),5) due to maintenance of sinus rhythm. Successful RFCA is also expected to reduce MR severity via reverse remodeling of LA volume and spatial relationships of the mitral apparatus. However, the extent of improvement in MR after successful RFCA can be difficult to predict. We thus conducted an analysis with the objective of rigorously specifying a model for echocardiographic prediction of MR severity one year after successful RFCA.

I Methods

1. Study population

The convenience sample of this single center cohort study was derived from consecutive patients who underwent initial RFCA for LSAF in our hospital. RFCA therapy was chosen based on shared decisions by patients and cardiologists generally following the JCS 2011 guidelines for non-pharmacotherapy of cardiac arrhythmias.⁶⁾ All patients recruited underwent RFCA between January 2014 and December 2017. The last 1-year post RFCA evaluation would have been December 2018. We identified fifty patients who met the following inclusion criteria:

1) lone LSAF having successful RFCA

2) no significant organic heart disease(s) and left ventricular ejection fraction > 55% confirmed by coronary CT angiography, two-dimensional transthoracic echocardiography (TTE) and trans-esophageal echocardiography before RFCA

3) Patients were excluded if they had an implanted pacemaker, any hormonal abnormality, or chronic end-stage renal disease requiring dialysis.

Patients’ baseline characteristics, comorbidities and AF-related medications were ascertained through medical chart review immediately prior to RFCA.

2. TTE

We performed TTE using a Vivid E9 with a M5Sc-D transducer (GE Healthcare, Chicago, IL, USA) and ECG evaluations prior to RFCA, 1-month and 1-year after successful RFCA. Routine TTE and Doppler examinations were initially performed to assess MR severity. Acquired still and video images were stored in the EchoPAC Clinical Workstation Software (GE Healthcare, Chicago, IL, USA). Measurements of the anteroposterior diameter (APD) and anterolateral-posteromedial diameter (ALPMD) of the mitral annulus were made with TTE images coincident with the ECG R wave by the EchoPAC Clinical Workstation Software. The APD was measured based on the long axis view and the ALPMD was measured based on the apical 2 chamber view, followed by calculation of the APD/ALPMD ratio. The severity of MR was classified semi-quantitatively, defined as follows: mild = 1, moderate = 2 and severe = 3 following the 2003 guidelines of the American Society of Echocardiography.⁷⁾ Furthermore, mild-to-moderate = 1.5 and moderate-to-severe = 2.5 categories were added to allow for more specific quantification of MR severity if certain parameters precluded unequivocal assignment to the standard mild-moderate-severe categories. All TTE records and findings were double-checked and reviewed for validity by a senior fellow of the Japanese Society of Ultrasonics in Medicine.

3. RFCA

RFCA was performed to create electrical isolation of all pulmonary veins from the LA by an ablation catheter (Therapy^TMCool Flex^TM, FlexAbility^TM, TactiCath^TM; Abbott, Chicago, IL, USA or Therapy^TMCool Path Duo^TM; St. Jude Medical, Saint Paul, MN, USA). Successful RFCA was defined by fulfillment of the following criteria:

1) no electrical potential in any of the pulmonary veins

2) electrical dissociation of the LA and each pulmonary vein.

According to the standards of clinical care, clinicians obtained written patient informed consent for the RFCA and data usage prior to the RFCA. This study was approved by the institutional review board of Hokkaido Cardiovascular Hospital (2022-A001).

4. Statistical analyses

We analyzed patient data such as age, gender, MR severity, existence of AF and echocardiographic variables such as LA volume index (LAVI), left ventricular ejection fraction (EF), left ventricular diastolic and systolic volumes, APD, ALPMD and APD/ALPMD ratio at 3 points in time (pre-RFCA, 1-month and 1-year post-RFCA). We then used multivariate analysis to determine the factors most predictive of MR severity 1-year after RFCA. We specified three multivariate linear regression models to predict MR severity 1-year after RFCA. Each model specified the change in MR severity (pre-RFCA–1-year post-RFCA) as the dependent variable with independent variables selected through backward elimination (variables with p > 0.1 were excluded). Age, gender and pre-RFCA MR severity were entered in as common independent variables in all three models explained below followed by backward elimination. Model A further used echocardiographic variables obtained from pre-RFCA and 1-month after RFCA. Model B used variables obtained from 1-month after RFCA and model C used only variables based on pre-RFCA data. The specificity and sensitivity of each model for worsening of MR (increasing severity of MR ≥ 0.5) were analyzed by receiver operating characteristic (ROC) curve analysis and compared based on area under the curve (AUC) statistics.

Relationships between the predicted and observed MR severity 1 year after RFCA and predicted/observed changes in MR severity 1-year after RFCA were evaluated by Pearson’s correlation coefficients. Measurements at each time point were compared to pre-RFCA using multiple comparison with Bonferroni correction following repeated measures ANOVA. All statistical analyses were performed using SPSS (SPSS, Inc., Chicago, IL, USA). Differences were defined as statistically significant if p < 0.05.

II Results

Patients’ characteristics just before RFCA are shown in Table 1. A majority of patients (74%) were male and hypertension was the most common comorbidity, present in 52% of participants. Table 2 demonstrates the TTE data (pre-RFCA, 1-month and 1-year post-RFCA). Compared to pre-RCFA, the APD, ALPMD, APD/ALPMD ratio, LAVI and EF were reduced significantly 1 year after successful RFCA (p < 0.01). Left ventricular end-diastolic and end-systolic volumes were unchanged from baseline despite successful RFCA. Relapsed AF was observed in 6 patients (12%) at 1-month and 5 patients (10%) at 1-year after RFCA.

Table 1 Baseline patient characteristics prior to RFCA

Male/female (n)		37/13
Mean age (years ± standard deviation)		65.4 ± 8.4
	≤ 64 years, n (%)	20 (40%)
	65–74 years, n (%)	23 (46%)
	75 years ≤, n (%)	7 (14%)
Comorbidities
	Hypertension, n (%)	26 (52%)
	Diabetes mellitus, n (%)	1 (2%)
	Stroke, n (%)	2 (4%)
	Vascular disease, n (%)	4 (8%)
	History of heart failure, n (%)	11 (22%)
CHA2DS2-VASc score
	0, n (%)	4 (8%)
	1, n (%)	11 (22%)
	2, n (%)	14 (28%)
	3, n (%)	14 (28%)
	4, n (%)	4 (8%)
	5, n (%)	1 (2%)
	6, n (%)	2 (4%)
Medications
	Angiotensin receptor blocker/angiotensin converting enzyme inhibitor, n (%)	19 (38%)
	β-blocker, n (%)	19 (38%)
	Calcium channel antagonist, n (%)	12 (24%)
	Diuretic, n (%)	10 (20%)
	Vitamin K antagonist, n (%)	7 (14%)
	New oral anti-coagulation drug, n (%)	37 (74%)
	Statin, n (%)	16 (32%)
	Anti-diabetic agent, n (%)	12 (24%)
	Nitrates, n (%)	3 (6%)
	Digitalis, n (%)	1 (2%)
Anti-arrhythmic drugs
	Class I, n (%)	2 (4%)
	Class III, n (%)	1 (2%)

Table 2 Echocardiographic data pre-RFCA, 1-month and 1-year after RFCA

	pre-RFCA		1-month			1-year
	mean	SD	mean	SD	p value vs. pre (Bonferoni)	mean	SD	p value vs. pre (Bonferoni)
APD of the mitral annulus, mm	30.0	3.0	28.2	3.3	< 0.001	27.2	3.6	< 0.001
ALPMD of the mitral annulus, mm	31.9	3.5	31.2	3.2	0.049	31.3	3.3	0.105
APD/ALPMD	0.95	0.10	0.91	0.10	0.01	0.87	0.09	< 0.001
LAVI, mL/m²	58.3	17.7	50.0	17.6	< 0.001	45.7	17.7	< 0.001
EF, %	54.5	7.9	56.8	6.7	0.092	59.5	6.1	< 0.001
Left ventricular end-diastolic volume, mL	87.8	51.9	86.9	24.7	1.000	89.4	20.2	1.000
Left ventricular end-systolic volume, mL	38.4	17.1	38.2	13.6	1.000	36.5	11.9	0.866

*p < 0.001 vs. the data in pre-RFCA.

In the regression models shown in Table 3, age, gender, left ventricular end-diastolic and end-systolic volumes, APD, ALPMD, existence of AF in pre and 1-month after RFCA and APD/ALPMD and EF in pre-RFCA were eliminated due to the absence of a significant relationship with severity of MR 1 year after RFCA. Variables include in the final multivariate regression model were LAVI in pre, APD/ALPMD ratio and EF at 1-month and severity of MR before and 1-month after RFCA. In accordance with the methods described above, the baseline pre-RFCA severity of MR in pre-RFCA was compulsorily included in all models. In model A (pre and 1-month), pre-RFCA LAVI and the APD/ALPMD and severity of MR at 1-month had significant positive associations with the dependent variable, while the EF at 1-month and pre-RFCA severity of MR had significant negative associations with the severity of MR 1-year after RFCA. In model B (1-month), the APD/ALPMD and LAVI in 1-month had significant positive associations and severity of MR in pre and 1-month after RFCA had significant negative association. Model C (pre) also demonstrated a statistically significant relationship between LAVI and pre-RFCA MR severity.

Table 3 Multivariate linear regression results

	β*	P	95% confidence interval
	β*	P	Lower	Upper
Model A (intercept: −0.358)
LAVI (pre-RFCA) per 100-points	0.727	0.012	0.168	1.287
APD/ALPMD of the mitral annulus (1-month) per 0.1-point	1.300	0.017	0.025	0.235
EF (1-month) per 10-points	−0.012	0.092	−0.027	0.002
Severity of MR (pre-RFCA) per 1-grade	−0.96	< 0.001	−1.182	−0.738
Severity of MR (1-month) per 1-grade	0.508	0.004	0.175	0.841
Adjusted R square = 0.638
Model B (intercept: −0.729)
APD/ALPMD of the mitral annulus (1-month) per 0.1-point	1.090	0.058	−0.004	0.221
LAVI (1-month) per 100-points	0.630	0.048	0.006	1.255
Severity of MR (pre-RFCA) per 1-grade	−0.842	< 0.001	−1.079	−0.605
Severity of MR (1-month) per 1-grade	−0.363	0.053	−0.006	0.732
Adjusted R square = 0.574
Model C (intercept: 0.374)
LAVI (pre-RFCA) per 100-points	0.739	0.031	0.070	1.408
Severity of MR (pre-RFCA) per 1-grade	−0.699	< 0.001	−0.910	−0.488
Adjusted R square = 0.478

*β corresponds to the difference between the MR pre-RFCA and 1 year after successful RFCA for each item listed.

Figure 1 demonstrates ROC curves of each model for predicting worsening of MR. Model A demonstrated the best ability to predict the severity of MR one year after successful RFCA (optimum sensitivity 1.00, specificity 0.795, AUC 0.881). Model C had the poorest predictive value (optimum sensitivity 0.500, specificity 0.591, AUC 0.548), while model B showed moderate ability to predict the severity of MR (optimum sensitivity 0.750, specificity 0.841, AUC 0.790). Figures 2A and 2B show the predicted/demonstrated MR severities and MR severity changes based on model A, respectively. The relationship between predicted and observed severity of MR 1-year after RFCA is shown in Figure 2A, and that between changes in predicted and observed severity of MR 1-year after RFCA in Figure 2B. There were strong correlations between predicted and observed severity of MR (R = 0.731) and between predicted and observed changes in severity of MR (R = 0.822) 1-year after successful RFCA.

Figure 1 ROC curves for each model

Receiver operating characteristic curves of each model.

Figure 2 Demonstrated MR severities and MR severity changes based on model A

Relationships between predicted severity of mitral regurgitation (MR) by the model A and demonstrated severity of MR 1-year after successful radio-frequency catheter ablation (RFCA) (A) and between changes in predicted severity of MR by the model A and changes in demonstrated severity of MR in 1-year after RFCA (B).

III Discussion

This observational study specifies a model which can be used for echocardiographic prediction of MR severity one year after successful RFCA among patients with LSAF.

The prevalence of AF increases with age in both in Japan and Western countries.^8)–10) In Japan, among patients in their 70s and 80s, 2–6% and 3–8% have AF, respectively.^9),10) AF is of clinical significance, including AF-induced complications, such as cerebrovascular disease,^11),12) valvular heart disease^1)–3),13) and heart failure.^14)–16) MR due to AF is one of the most important complications because of the significant risk of heart failure with which it is associated.¹⁶⁾ Recently, many reports have been published regarding AF-induced MR (i.e. atrial functional MR)^1)–3) highlighting an important mechanism by which MR develops. AF induces remodeling not only of the LA myocardium but also the mitral apparatus. Elimination of the causative pathophysiology through maintenance of sinus rhythm may induce reverse remodeling of the LA myocardium and the mitral apparatus. This reversal may, in turn, reduce MR severity. However, the sustainability of reductions in the severity of MR after successful RFCA have been difficult to predict.

The model A is highly accurate with an AUC of 0.881 based on ROC analysis, and we believe it will be a new way to predict MR severity one year after successful RFCA.

This model allows for robust prediction of MR severity even in the presence of recurrent AF. The severity of MR may be related to the extent of AF-induced remodeling of the LA and the mitral annulus and/or left ventricular systolic function. In particular, LA volume and the mitral annulus configuration may be the factors most important in determining MR severity in LSAF.

Remodeling due to AF induces fibrosis of the LA myocardium,^17)–19) the extent of which may be related to the duration of AF. If fibrosis of the LA myocardium and/or the mitral apparatus becomes irreversible, the ability of maintenance of sinus rhythm to produce reverse remodeling may be limited. This may also complicate prediction of MR severity after RFCA.

The models we have studied allows for echocardiographic prediction of MR severity after RFCA, which may be a useful tool in guiding therapeutic strategies. Also, our study offers novel, clinically-relevant information to practicing cardiologists, who can use the predictive variables to facilitate patient-specific decisions.

IV Limitations

This study is based on observational data from a small, predominantly male sample of patients at a single center. The precise duration of AF could not be known with certainty in many patients, and the extent of fibrosis of the LA myocardium and/or the mitral apparatus could not be assessed.

V Conclusions

Based on this study, we conclude that the severity of MR one year after successful RFCA can be predicted by supplementing clinical information with echocardiographic data. When clinicians consider RFCA for AF patients with MR, the effect of RFCA on reducing MR severity can be estimated by using the model we have specified and tested.

Appendix

The predicting formula of the model A is as follows:

Predicted severity of MR 1 year after RFCA = −0.358 + {0.01 × LAVI (pre-RFCA) × 0.727} + {10 × APD/ALPMD ratio of the mitral annulus (1-month) × 0.130} + {0.1 × EF (1-month) × −0.123} + (severity of MR (pre-RFCA) × −0.960) + {severity of MR (1-month) × 0.508}

Conflict of Interest

There is no potential conflict of interest to disclose.

Acknowledgements

We sincerely express our gratitude to Saki Ookubo, Takanori Kimura, Ayaka Oobayashi, Shizuka Tahara, Haruka Haraguchi, Yuki Kubota and Seiko Ohno for assisting us through their expertise in TTE.

References

1) Gertz ZM et al.: Evidence of atrial functional mitral regurgitation due to atrial fibrillation reversal with arrhythmia control. J Am Coll Cardiol, 2011; 58: 1474–1481.
2) Liang JJ, Silvestry FE: Mechanistic insights into mitral regurgitation due to atrial fibrillation: “Atrial functional mitral regurgitation”. Trends Cardiovasc Med, 2016; 26: 681–689.
3) Machino-Ohtsuka T et al.: Novel mechanistic insights into atrial functional mitral regurgitation—3-dimensional echocardiographic study. Circ J, 2016; 80: 2240–2248.
4) Marrouche NF et al.: Catheter ablation for atrial fibrillation with heart failure. N Engl J Med, 2018; 378: 417–427.
5) Friberg L et al.: Catheter ablation for atrial fibrillation is associated with lower incidence of stroke and death: Data from Swedish health registries. Eur Heart J, 2016; 37: 2478–2487.
6) JCS Joint Working Group: Guidelines for non-pharmacotherapy of cardiac arrhythmias (JCS 2011). Circ J, 2013; 77: 249–274.
7) Zoghbi WA et al.: Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography. J Am Soc Echocardiogr, 2003; 16: 777–802.
8) Heeringa J et al.: Prevalence, incidence and lifetime risk of atrial fibrillation: The Rotterdam study. Eur Heart J, 2006; 27: 949–953.
9) Inoue H et al.: Prevalence of atrial fibrillation in the general population of Japan: An analysis based on periodic health examination. Int J Cardiol, 2009; 137: 102–107.
10) Iguchi T et al.: Prevalence of atrial fibrillation in community-dwelling Japanese aged 40 years or older in Japan—Analysis of 41,436 non-employee residents in Kurashiki-city. Circ J, 2008; 72: 909–913.
11) Takabayashi K et al.: Incidence of stroke or systemic embolism in paroxysmal versus sustained atrial fibrillation: The Fushimi atrial fibrillation registry. Stroke, 2015; 46: 3354–3361.
12) Friberg L et al.: High prevalence of atrial fibrillation among patients with ischemic stroke. Stroke, 2014; 45: 2599–2605.
13) Thomas KL et al.: Prevalence, characteristics, and outcomes of valvular heart disease in patients with atrial fibrillation: Insights from the ORBIT-AF (outcomes registry for better informed treatment for atrial fibrillation). J Am Heart Assoc, 2017; 6: e006475.
14) Kotecha D et al.: Atrial fibrillation and heart failure due to reduced versus preserved ejection fraction: A systematic review and meta-analysis of death and adverse outcomes. Int J Cardiol, 2016; 203: 660–666.
15) Thihalolipavana S, Morina DP: Atrial fibrillation and heart failure: Update 2015. Prog Cardiovasc Dis, 2015: 58: 126–135.
16) Saito C et al.: Prevalence, clinical characteristics, and outcome of atrial functional mitral regurgitation in hospitalized heart failure patients with atrial fibrillation. J Cardiol, 2018; 72: 292–299.
17) Casaclang-Verzosa G et al.: Structural and functional remodeling of the left atrium: Clinical and therapeutic implications for atrial fibrillation. J Am Coll Cardiol, 2008; 51: 1–11.
18) Gal P, Marrouche NF: Magnetic resonance imaging of atrial fibrosis: Redefining atrial fibrillation to a syndrome. Eur Heart J, 2017; 38: 14–19.
19) Burstein B, Nattel S: Atrial fibrosis: Mechanisms and clinical relevance in atrial fibrillation. J Am Coll Cardiol, 2008; 51: 802–809.

Corresponding author

Register with J-STAGE for free!