Long-term cardiac remodeling associated with heart failure following left-ventricular valve replacement surgery

Abstract To investigate long-term cardiac remodeling and prognosis of patients post-left-ventricular valve replacement, and explored related risk factors of heart failure and management strategies. Retrospective cohort of patients with left-ventricular valve replacement between 2005 and 2007. Major adverse cardiac events were recorded, including death, hospitalization, stroke, and New York Heart Association (NYHA) functional classifications. Cardiac remodeling was assessed by comparing pre-operative, post-operative, and follow-up echocardiographic images. (1) Two hundred fifty-seven patients who received left-ventricular mitral, aortic, or double-valve replacement surgery were followed up for 10.4 ± 1.5 years with an all-cause mortality rate 18.7% and an incidence of heart failure that significantly restricted daily life (NYHA III or IV) 21.3%.(2) There were no significant differences in classic cardiac-remodeling variables between baseline and long-term follow-up, such as left-ventricular diameter (47.9 ± 8.3 vs 49.9 ± 8.0 mm, P = .14) and left-ventricular ejection fraction (58.6 ± 9.6% vs 57.0 ± 10.3%, P = .34), whereas there were significant differences in terms of left-atrial anteroposterior diameter (LA) (39.7 ± 9.5 vs 49.0 ± 14.3 mm, P < .001) and tricuspid regurgitation (TR) (1.4 ± 1.0 vs 2.2 ± 1.2, P < .001). Multivariable logistic regression analysis showed that LA ≥ 50 mm (P = .011) and more than moderate tricuspid regurgitation (TR > 2) (P = .012) were associated with poor prognoses for long-term consequences of heart failure. Both LA and TR progressed with the length of time after surgery. LA enlargement and TR after left-ventricular valve replacement surgery were time-dependent events, which represented cardiac remodeling and were closely related to post-operative long-term consequences of heart failure. It is important to be cognizant of and to explore long-term preventive and treatment strategies for adverse cardiac events in patients following left-ventricular valve replacement.


Introduction
Currently, valvular heart disease remains a serious public health issue. The incidence of moderate to severe valvular heart disease with significant clinical presentations is 2.5%. [1] In patients over 65 years old, this incidence is reported to reach 6.4%. [2] Successful valvular surgery may significantly reduce mortality and is the mainstay treatment for these individuals. [3] However, it remains unknown if the benefits of cardiac valvular surgery last long-term. Studies found that post-operative survival rates after mitral-valve surgery were 97.5%, 87.5%, and 43.0% at 5, 10, and 20 years post-surgery, respectively. [4,5] The 15-year survival rate after aortic-valve replacement surgery was 60.6% for biological valves and 62.1% for mechanical vales. [6] A previous study on the long-term prognosis after aortic-valve replacement surgery found that congestive heart failure was an important cause of death (21.0%). [7] One, 5, 10, and 15 years after aortic-valve replacement surgery, freedom of heart failure was 98.6% ± 0.3%, 88.6% ± 1.0%, 73.9% ± 2.3%, and 45.2% ± 8.5%, respectively. [8] These findings suggest that patients carry a risk of poor long-term prognosis even after successful valve replacement surgery.
Although there are many studies on post-operative valvular heart disease, the long-term post-operative prognosis of this cohort has not received adequate attention. Some studies showed improvement of cardiac remodeling after left-ventricular valve replacement surgery, which suggests that this surgical treatment successfully resolves the valvular mechanical lesions and reverses cardiac remodeling. [9,10] However, there remains residual cardiac remodeling after left-ventricular valve replacement surgery, which is associated with poor prognosis. [10] These previous studies have been limited to the assessment of short-and/or medium-term recoveries. In contrast, there have been few studies on the prognosis of patients 10 years after the surgery. Additionally, few studies have systematically compared preand post-operative echocardiographic features. Due to a paucity of clinical evidence, clinical guidelines do not mention any longterm prognosis or management strategies following left-ventricular valve replacement surgery. [11] Therefore, to investigate such long-term prognoses, as well as to determine the cardiac structure and functional changes after left-ventricular valve replacement surgery, we conducted a clinical follow-up observational study and collected data from 257 consecutive patients (average followup 10

Data collection
The data collected included: patient age and gender; re-admission history (e.g., stroke, heart failure, and atrial fibrillation); etiology (e.g., rheumatic heart disease, infective endocarditis, mitral-valve prolapse, congenital heart disease, and nonrheumatic valvular heart disease); comorbidities (e.g., hypertension, hyperlipidemia, diabetes, and coronary heart disease); New York Heart Association (NYHA) functional classification during the follow-up; medications (e.g., anticoagulants, beta-blockers, reninangiotensin system inhibitors, and diuretics); and echocardiographic images at pre-operative, post-operative, and long-term follow-up periods. Pre-and post-operative baseline echocardiographic images were obtained at 2 weeks before and 2 to 4 weeks after the surgery.

Echocardiography
Patients received echocardiographic examinations during the pre-operative, post-operative, and long-term follow-up visits. Patients were placed in the supine or the left lateral position. Based on the Guidelines for echocardiographic measurements in Chinese adults, [12] the anteroposterior diameter of the left atrium, left-ventricular end-diastolic anteroposterior diameter (LVDd), end-diastolic ventricular septum thickness, and leftventricular posterior wall thickness were measured via parasternal left-ventricular long-axis and short-axis views. The leftventricular ejection fraction (LVEF) was measured by the dualplane Simpson method. The peak of the tricuspid regurgitation (TR) spectrum was measured by continuous Doppler at the apex two-chamber and four-chamber views. Pulmonary arterial hypertension can be detected non-invasively with Doppler echocardiography by pulmonary arterial systolic pressure. Right ventricular tricuspid pressure gradient (RVTG) can be considered equal to the pulmonary arterial systolic pressure after adding the estimated mean right atrial pressure of 10 mm Hg. [13]

Statistical analysis
Data were analyzed via SPSS 18.0 software (IBM). Continuous data with normal distributions are presented as means ± standard deviations and were compared by independent-sample t tests (2 groups) or analyses of variance (ANOVAs; more than 2 groups). Categorical data are presented as percentages (%) and were compared via Chi-square tests. Survival analysis was performed by the Kaplan-Meier method. Multivariable logistic regression analysis was performed to determine prognostic factors. Timetrend evaluations were based on correlational analysis using Spearman correlation. [14] Statistical significance was taken as a P value < .05.
Chen et al. Medicine (2021)  To understand long-term cardiac remodeling after left-ventricular valve replacement surgery, we assessed echocardiographic changes from post-surgery baselines (2-4 weeks after surgery) in relation to those at long-term follow-ups. A total of 104 patients had echocardiographic images at baseline and long-term followup (7-12 years). There were no statistically significant changes in LVDd (47.9 ± 8.3 vs 49.9 ± 8.0 mm, P = .14) or LVEF (58.6 ± 9.6% vs 57.0 ± 10.4%, P = .34). In contrast, there were statistically significant changes in left-atrial anteroposterior diameter (LA) (39.7 ± 9.5 vs 49.0 ± 14.3 mm, P < .001) and TR grades (1.4 ± 1.0 vs 2.2 ± 1.2, P < .001). These results indicate that cardiac remodeling occurred long-term after left-ventricular valve replacement surgery and presented mainly as LA dilatation and TR. After different valve type surgery such as the mitral valve, aortic valve, and double-valve replacement surgery, the long-term progress of LA and TR can still be seen, and the other classic cardiac remodeling indicators have not progressed significantly (Table 2B).

3.3.
Multivariable logistic regression analysis of long-term cardiac dysfunctions after left-ventricular valve replacement surgery (Table 3) To investigate correlations for the high incidence of long-term cardiac dysfunction after left-ventricular valve replacement surgery, multivariable logistic regression analysis was performed. The incidence of NYHA class III + IV was considered as a poor prognostic index. It was found that gender (P = .003), age (P = .004), long-term LA enlargement (P = .011), and greater than moderate long-term TR (P = .012) were associated with poor prognosis of heart failure, whereas classical cardiac remodeling indicators, such as LVDd and LVEF, did not show any correlations (Table 3).

Correlation analysis of long-term cardiac dysfunctions after left-ventricular valve replacement surgery (Fig. 3)
In order to further clarify the relationship between LA dilatation and TR with heart failure, long-term follow-up echocardiographic data were analyzed. Patients with LA ≥ 50 mm had a higher cardiac function score than did patients with LA < 50 mm (2.3 ± 1.0 vs 1.6 ± 0.7, P < .001; Fig. 3A). This result suggested that a larger LA was associated with more severe heart failure and worse outcomes. Patients with more than moderate tricuspid regurgitation (TR > 2) had a higher NYHA functional classification than those of patients with TR 2 group (2.2 ± 0.9 vs 1.7 ± 0.8, P = .011; Fig. 3B). This finding demonstrates that a higher TR grade is associated with more severe heart failure and worse outcomes.
3.5. Time-dependent changes in LA and TR after leftventricular valve replacement surgery (Fig. 4) Scatter plots of LA and TR overtime at baseline and at long-term follow-up visits suggested that the progressions of LA enlargement and TR grades (TR > 2) were time-dependent for both (r = 0.361, P < .001; r = 0.387, P < .001; Fig. 4A and B). This finding indicates that cardiac remodeling after left-ventricular valve replacement surgery progresses as a function of time from surgery.

Discussion
The present study found that patients who received leftventricular valve replacement surgery had unsatisfactory longterm prognosis, with the mortality rate being 18.7% and the incidence of severe heart failure (NYHA functional classifications III + IV) being 21.3%. Additionally, echocardiographic results showed that the classical left-ventricular dysfunction predictors,  LVDd and LVEF, were unable to predict long-term consequences of heart failure. Conversely, LA enlargement and TR gradually increased over time post-surgery and were associated with heart failure. Finally, following left-ventricular valve replacement, LA dilatation and TR progressed with time and displayed characteristic changes of long-term cardiac remodeling. After different valve type surgery such as the mitral valve, aortic valve, and double-valve replacement surgery, the long-term progress of LA and TR can still be seen, and the other classic cardiac remodeling indicators have not progressed significantly.
In the study cohort, the 10-year survival rate after leftventricular valve replacement was 81.3%. Among surviving patients, the proportion of patients with severe cardiac dysfunction was 21.3%. These findings suggest that the longterm prognosis was poor and that the cardiac dysfunction classification was high after left-ventricular valve replacement surgery. In a long-term follow-up study (7.1 ± 2.7 years) in 96 patients with functional tricuspid regurgitation who received mitral-valve replacement surgery, the 5-and 10-year survival Table 3 Logistic regression analysis of variables associated with NYHA classifications.

Variables
HR (95% confidence intervals) P   rates were 97.5% and 87.5%, respectively. [4] These results suggest that there is still a risk of long-term poor prognosis after successful surgical heart valve replacement surgery. The main indications for left-ventricular valve replacement surgery are mitral regurgitation and aortic regurgitation with or without aortic stenosis. In addition to ventricular hypertrophy caused by the increased left-ventricular pressure load from aortic stenosis, the main cardiac remodeling changes before valve replacement surgery are secondary to increased left-ventricular volume overload. This can lead to left-ventricular enlargement and EF deterioration from compensation to decompensation, which ultimately results in heart failure with reduced EF. After valve replacement surgery, the valvular mechanical dysfunction is corrected and cardiac enlargement can improve significantly in the short term. However, several studies, including the present study, showed that cardiac remodeling still exists and continues to progress after surgery. This in turn may contribute to poor cardiac function and decreased long-term survival. A meaningful finding from the present study is that classical cardiac remodeling indicators (such as LVDd, EF, and RVTG) do not reliably predict outcomes following valve replacement surgery. On the contrary, the present study showed that the long-term cardiac remodeling pattern after left-ventricular valve replacement was similar to that of heart failure with preserved EF. This suggests that post-  operative long-term cardiac remodeling may not be a residual or continuation of valvular heart damage, but is likely the consequence of the valve replacement procedure or from the implanted prosthetic valve. The mechanism of cardiac remodeling after left-ventricular valve replacement has not been fully clarified. Some studies suggested that the anatomical structure and position of the prosthetic valve were different from those of the original valve. These post-surgical deviations may cause hemodynamic dysfunction and/or aberrant blood flow, which have been considered to play an important role in the process of cardiac remodeling. [15] Studies showed that patients with leftventricular valve repair have a greater survival advantage than patients with replacement surgery. This survival advantage was shown to become more significant over time. [16] These findings highlight the importance of developing improved techniques for performing valve replacement surgery and in improving the selection/use of prosthetic valves. TR commonly becomes worse with pulmonary hypertension in patients with primary left-ventricular valvular disease. This is well understood since valvular heart disease can cause an increase in left-ventricular pressure and capacity load, which in turn enlarges the LA. This can increase the pulmonary arterial pressure and elevate the right-ventricular load, with the final outcome inducing progressions of TR and pulmonary hypertension. Our present study showed that, after left-ventricular valve replacement surgery, the pre-operative TR and RVTG were significantly decreased and ameliorated. However, the long-term progression of post-operative TR was not accompanied by progression of pulmonary hypertension. The incidences of moderate TR increased from 1.7% in the post-operative baseline to 31.7% in the long-term follow-up period, but the changes in RVTG were not statistically significant. Obviously, this finding cannot be explained simply by right heart failure from blood backflow due to left heart failure. A more likely explanation is that, after left-ventricular valve replacement surgery, long-term left-ventricular remodeling induces rightventricular remodeling which may promote TR. It has been shown that TR is caused by the expansion of the tricuspid annulus. [17] Tricuspid valve-annulus dilation is recommended as an indication (Class IIa and b) for simultaneous tricuspid valve surgery during left-ventricular valve surgery. [11,18] However, the improvement of TR after tricuspid valve surgery does not completely convert into a survival benefit. [19] Therefore, TR exacerbation is primarily a right-ventricular disorder. Even if surgery corrects TR, it does not fundamentally improve the prognosis of right-ventricular dysfunction. This also suggests that the mechanism and pattern of cardiac remodeling after leftventricular valve replacement surgery are different from those of primary valvular disease. [20] Nardi et al investigated a total of 108 patients who underwent left atrial ablation by means of unipolar or bipolar radiofrequency for atrial fibrillation in association with mitral or mitral and aortic valve surgery. The study showed that the left atrium size is associated with poor long-term prognosis after left-ventricular valve replacement. [21] Clinic data including present study shows that cardiac remodeling persists even after left-ventricular valve replacement surgery, which leads to poor long-term outcomes. This understanding reminds us to carefully manage patients after valve replacement surgery. The latest 2014 AHA/ACC and 2017 ESC/ EACTS guidelines for valvular heart disease do not mention any drug therapies for cardiac remodeling after cardiac valve replacement surgery. [11,18] In clinical practice, most patients do not receive long-term medication (except warfarin) after successful valve replacement surgery. In our present study, up to 96.7% of patients still took warfarin 10 years after their surgery. Traditionally, b-blockers, angiotensin converting enzyme inhibitor and/or angiotensin II receptor blocker, and aldosterone inhibitors that can reverse or delay cardiac remodeling are not used in the clinic. This might be related to the long-term cardiac dysfunction found in our present study population. However, results herein suggest that the cardiac remodeling pattern after left-ventricular valve replacement surgery is different from that of the traditional heart failure with reduced EF model. The effectiveness of above drug treatments on preventing long-term cardiac remodeling following left-ventricular valve replacement surgery remains to be determined.

Limitations
The limitations of the present study include that it was a single-center study with a small number of patients. Only 104 (40.5%) patients had complete echocardiographic images available for analysis. Also, the retrospective study design of our present study had inherent biases. Due to the small sample size, the correlation is not strong, but the long-term cardiac remodeling collected by echocardiography can reflect the trend of progress over time. In the future, we can collect short-term, mid-term, and long-term continuous echocardiography data after surgery to better prove the time trend. We did not fully consider the inclusion and exclusion criteria, especially with respect to atrial flutter or fibrillation, AV conduction defects. We did not use the right heart catheter to measure pulmonary arterial pressure and pulmonary impedance accurately. Echocardiographic estimates of pulmonary arterial pressure may be less accurate. Future multicenter studies with larger patient populations and complete data sets are required to validate the present results.

Conclusions
In the present study, it was found that left atrial enlargement and tricuspid regurgitation progressed over time and represented characteristic changes of cardiac remodeling after left-ventricular valve replacement surgery. These changes were closely related to long-term heart failure after surgery. Hence, more studies are warranted to investigate long-term prevention and treatments after left-ventricular valve replacement surgery.