Cardiac Damage Staging Predicts Outcomes in Aortic Valve Stenosis After Aortic Valve Replacement

Background The prognostic value of cardiac damage staging classification based on the extent of extravalvular damage has been proposed in moderate/severe aortic stenosis (AS). Objectives The purpose of this study was to assess the association of cardiac damage staging with mortality across the spectrum of patients with AS following aortic surgical or transcatheter aortic valve replacement (AVR). Methods We conducted a pooled meta-analysis of Kaplan-Meier-derived reconstructed time-to-event data from studies published through February 2023. Results In total, 16 studies (n = 14,499) met our eligibility criteria and included 12,282 patients with symptomatic severe AS and 2,217 patients with asymptomatic severe/moderate AS. For patients with symptomatic severe AS, all-cause mortality was 24.0%, 27.7%, 38.0%, 56.3%, and 57.3% at 5 years in patients with cardiac damage stage 0, 1, 2, 3, and 4, respectively (stage 0 as reference; HR in stage 1: 1.30 [95% CI: 1.03-1.64]; P = 0.029; stage 2: 1.74 [95% CI: 1.41-2.16]; P < 0.001; stage 3: 2.92 [95% CI: 2.35-3.64]; P < 0.001, and stage 4: 3.51 [95% CI: 2.79-4.41]; P < 0.001). For patients with asymptomatic moderate/severe AS, all-cause mortality was 19.3%, 36.9%, 51.7%, and 67.8% at 8 years in patients with cardiac damage stage 0, 1, 2, and 3 to 4, respectively (HR in stage 1: 1.70 [95% CI: 1.21-2.38]; P = 0.002; stage 2: 2.20 [95% CI: 1.60-3.02]; P < 0.001; and stage 3 to 4: 3.90 [95% CI: 2.79-5.47]; P < 0.001). Conclusions In patients undergoing AVR across the symptomatic and severity spectrum of AS, cardiac damage staging at baseline has important prognostic implications. This pooled meta-analysis in patients undergoing AVR suggests that staging of baseline cardiac damage could be considered for timing and selection of therapy in patients with moderate or severe AS to determine the need for earlier AVR or adjunctive pharmacotherapy to prevent irreversible cardiac damage and improve the long-term prognosis.

][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23] Moreover, this classification has been validated in other valvular heart diseases including mitral regurgitation, [24][25][26][27][28] tricuspid regurgitation, 29,30 and aortic regurgitation. 31e extravalvular cardiac damage staging proposed by Généreux et al 3 introduced a more holistic and multiparametric approach to assessing the various clinical and structural changes associated with AS.Such an approach of staging patients based on global cardiac health has shown prognostic utility in risk stratification and potentially in determining the optimal timing for AVR.Despite previous reports demonstrating the impact of cardiac damage staging in AS, some cardiac damage stages were underrepresented given the considerably low number of patients in individual studies, making it somewhat challenging to elicit more robust conclusions.This pooled meta-analysis of Kaplan-Meier-derived reconstructed time-to-event data aims to assess the impact of cardiac damage staging across the spectrum of severity and symptoms of AS on all-cause and cardiovascular (CV) mortality after AVR.

METHODS
SEARCH STRATEGY.This meta-analysis of observational studies [6][7][8][9][10][14][15][16][18][19][20][21][22]32 and randomized trials 4,13 was conducted to assess the long-term prognostic performance of cardiac damage staging classification across the spectrum of AS (ie, symptomatic severe, asymptomatic severe, and moderate AS) undergoing AVR. Data collectio and reporting were performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and the Meta-analysis of Observational Studies in Epidemiology reporting guidelines.33,34 The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42023406744). Ethl approval was not required for this study-level meta-analysis.

ENDPOINTS OF INTEREST.
The primary endpoint of this Kaplan-Meier-derived meta-analysis was allcause death.Purposely, composite endpoints were not included in this meta-analysis as the definitions of heart failure hospitalization and readmissions varied across studies.Secondary endpoints were CV death and stroke.Populations studied were: 1) symptomatic severe AS; and 2) asymptomatic severe and asymptomatic moderate AS who underwent AVR.
DATA EXTRACTION.Individual patient data based on published Kaplan-Meier graphs from all included studies were reconstructed using the "curve approach." 39We used the 2-stage approach described by Liu et al. 40 In the first stage, raw data coordinates (time, survival probability) were extracted from each subgroup in each of the respective Kaplan-Meier curves.
Two investigators assessed the accuracy of the reconstructed patient data by comparing the survival probability in the published Kaplan-Meier curve at each read-in point with the corresponding reconstructed survival probability and estimated number of patients at risk. 40TA POOLING.The reconstructed individual patient data from the individual studies were merged to construct the final study datasets.Freedom from the respective outcomes was visually assessed using Kaplan-Meier estimates and analyzed using Cox proportional hazards models stratified by the cardiac damage stages (0, 1, 2, 3, and 4).Studies that provided merged staging categories (ie, stage 0-1 and stage 3-4)   were reconstructed in a separate analysis.
RISK FOR BIAS.The Risk of Bias in Nonrandomized Studies of Exposure tool was systematically used to assess included studies for risk of bias. 41The studies and their characteristics were classified into 5 groups: A (low risk for bias), B (moderate risk for bias), C (serious risk for bias), D (critical risk for bias), or E (no information/unclear). Two independent reviewers assessed the risk of bias.Disagreements were resolved by consensus among both reviewers.
STATISTICAL ANALYSIS.In the Kaplan-Meier-based meta-analysis, the mean AE SD survival times, median (IQR) survival times, and percentage of survival at different time points with 95% CIs were calculated.
The differences in survival between the groups were assessed using the log-rank test for differences and a Cox proportional hazards regression model.Truncated survival analysis at 1 year was performed as a prespecified outcome, and the respective hazard ratios (HRs) and 95% CI were calculated.We also analyzed the data up to the longest follow-up.P values were 2-sided, and statistical significance was set at P <0.05.All analyses were completed with R version 4.2.1 (Foundation for Statistical Computing).
We further assessed the differences in CV mortality as a secondary outcome between symptomatic and asymptomatic AS cohorts.CV death was significantly different between the different groups (log-rank chi-square, 219.5;P < 0.001) (Supplemental Table 5, Supplemental Appendix, Figure 5).

DISCUSSION
The current study, analyzing 14,499 patients with AS undergoing AVR, summarizes several key aspects of the prognostic impact of cardiac damage staging uti- fraction and concomitant moderate AS, with a similar benefit to those with severe symptomatic AS. 46,47 When analyzed collectively, such observations would prompt early consideration of AVR in this population before the onset of irreversible damage.
Indeed, Généreux et al 4,5 demonstrated that the extravalvular damage resulting from AS remained Importantly, when comparing asymptomatic to symptomatic groups, our analysis showed that there was no difference in all-cause or CV mortality in corresponding cardiac damage stages.Our findings      Go to http://www.acc.org/jacc-journals-cme to take the CME/MOC/ECME quiz for this article.

T
he current 2020 American College of Cardiology/American Heart Association and the 2021 European Society of Cardiology/European Association for Cardiothoracic Surgery guidelines recommend aortic valve replacement (AVR) for patients with aortic stenosis (AS) presenting with either severe AS with symptoms or low left ventricular (LV) ejection fraction (<50%) or when moderate AS co-exists with another indication for open heart surgery (ie, coronary artery bypass, other valve disease, aortopathy).

Figure 1 .
Figure 1.STUDY SELECTION AND ELIGIBILITY CRITERIA.Two reviewers (O.M.A. and X.J.) independently and systematically reviewed the studies from the pooled search of databases, screened the abstracts, and confirmed their eligibility through full-text assessment.Disagreements were resolved by consensus with a third investigator (P.G.).Furthermore, we manually screened the references from articles of interest.We included studies meeting the following inclusion criteria: 1) the population consisting of patients with AS, regardless of severity or symptom presentation; 2) patients included in these studies underwent AVR (surgical or transcatheter following baseline cardiac damage staging); 3) outcomes of interest included all-cause and/or CV death; 4) Kaplan-Meier curves were provided for each of the individual/grouped cardiac damage stages; and 5) the study design was retrospective/prospective cohorts, post-hoc randomized trial populations, monocentric/multicentric, with matched/unmatched populations.Animal studies, reviews, congress abstracts, editorials, letters, comments, case reports, and case series were excluded.STAGING CLASSIFICATION FOR CARDIAC DAMAGE.The presence and magnitude of extra-aortic valvular cardiac damage were evaluated based on baseline transthoracic echocardiography and, accordingly, patients were classified into 5 independent stages as proposed by Généreux et al 3 prior to undergoing AVR: Stage 0-No signs of cardiac damage (ie, damage confined to aortic valve).
(N ¼ 12,282), 9 studies presented data on individual cardiac damage stages (Stage 0-4), while 3 studies merged cardiac damage stages (ie, stage 0-1 and stage 3-4) owing to small sample sizes in individual stages.In symptomatic severe AS population, the mean age was 79.5 years (95% CI: 77.8-81.2years), 50.5% were male, and 66.3% were in NYHA functional class III-IV.The pooled prevalence of the grouped and individual original cardiac damage stages are presented in Figure 1.According to the staging classification, 1,391 (13.3%) patients were in stage 0, 1,988 (19%) patients were in stage 1, 4,231 (40.4%) patients were in stage 2, 1,688 (16.1%) patients were in stage 3, and 1,168 (11.2%) patients were in stage 4. Patients in the more advanced stages of cardiac damage were older, presented with an increasing proportion of NYHA functional class III-IV, and had a higher prevalence of comorbidities, including coronary artery disease (CAD), chronic kidney disease, atrial fibrillation, previous cardiac surgery, myocardial infarction, and prior stroke.The clinical and echocardiographic baseline characteristics of the pooled population according to merged and individual cardiac damage stages are presented in Tables

an approximate 2 -
lizing reconstructed pooled time-to-event patient data on mortality post-AVR.Our main findings are as follows: 1) a significant proportion of patients had extensive cardiac damage at baseline regardless of severity or symptomatic status; 2) the extent of cardiac damage was associated with all-cause and CV mortality at 5 years for symptomatic severe AS; 3) the extent of cardiac damage was associated with allcause and CV mortality at 8 years for asymptomatic moderate and severe AS; 4) there was no difference in all-cause and CV mortality between the asymptomatic and symptomatic cohort following AVR at shortand long-term follow-up.The current study reinforces the prognostic value of the cardiac damage staging approach in all-comers AS patients, including symptomatic and asymptomatic cohorts undergoing either surgical or transcatheter AVR.Indeed, our pooled analysis revealed that approximately 60% of asymptomatic moderatesevere AS patients who underwent AVR had evidence of cardiac damage at baseline (stage $2) with Abdelfattah et al J A C C : A D V A N C E S , V O L . 3 , N O . 5 and 4-fold increase in mortality for stages 2 and 3 to 4, respectively.These findings indicate that triggers for AVR in the current guidelines are neither specific nor sensitive in accounting for the extent of cardiac damage and subsequent mortality risk. 1,2Moreover, a third of the population had LV cardiac damage (stage 1), which was associated with a w2-fold increase in mortality risk at 8year follow-up.Studies have shown that the cutoff of LV ejection fraction <50% in defining LV dysfunction as a trigger for AVR is not sensitive in the detection of subclinical LV dysfunction. 44,45Indeed, the development of cardiac damage could start long before symptom and severity manifestation of severe AS, prompting a reconsideration of more appropriate cutoffs in the timing of intervention in this population.In addition, several small-sized observational studies demonstrated a mortality benefit with AVR in patients with heart failure with reduced ejection

FIGURE 1
FIGURE 1 Distribution of Cardiac Damage Stages in Severe Symptomatic Aortic Stenosis

FIGURE 2
FIGURE 2 Reconstructed Kaplan-Meier Analyses According to Cardiac Damage for Severe Symptomatic Aortic Stenosis Following AVR

FIGURE 3 4 CardiacTAVR
FIGURE 3 Distribution of Cardiac Damage Stages in Asymptomatic Moderate/Severe Aortic Stenosis

FIGURE 4
FIGURE 4 Reconstructed Kaplan-Meier Analyses According to Cardiac Damage for All-Cause Mortality in Asymptomatic Moderate/ Severe Aortic Stenosis Undergoing AVR

J
A C C : A D V A N C E S , V O L . 3 , N O . 5 , 2 0 2 4 modified versions of the initial cardiac damage staging classification, such risk markers were not used across all the included studies, limiting a pooled analysis of such variables. 9,11,12,52,53Third, interval classification of cardiac damage staging during follow-up was not present in the majority of the included studies.Hence, we could not identify if the patients' corresponding cardiac damage stage improved or worsened over follow-up.Fourth, granular individual data on background medical therapy, comorbidities, and certain factors such as the extent of CAD were not available within all the pooled studies, hindering potential insights and riskadjusted analyses for certain factors that might impact cardiac damage extent and outcomes.Fifth, within the symptomatic cohort, the majority of studies encompassed patients who underwent AVR and were longitudinally monitored.In contrast, within the asymptomatic cohort, patient stratification based on initial AVR status was not consistently applied across all studies, with only 1 study encompassing patients who all underwent AVR immediately following cardiac damage staging.Thus, interpretation of these results in the context of an asymptomatic cohort undergoing AVR needs to be made cautiously.Nevertheless, it is noteworthy that approximately 40% of patients in the asymptomatic cohort eventually underwent AVR.Sixth, interestingly, several studies combined stage 0 to 1 and stage 3 to 4 due to the low number of patients and low power in each stage.Conceptually, it will seem more appropriate to consider stage 0 separately, pool stages 1 and 2 (LV and LA damage, representing left side damage), and stage 3 to 4 (pulmonary artery and right ventricle, representing right side damage).Indeed, patients in stage 0 have no detectable extravalvular damage yet, representing a unique population most likely at an earlier stage of the disease or more resilient to afterload mismatch compared with patients who have already established damage.Finally, the initial classification used simple criteria easily applicable by echocardiographers in clinical practice; however, the inclusion of more advanced analysis including LV strain might enhance

FIGURE 5
FIGURE 5 Reconstructed Kaplan-Meier Analyses for All-Cause Death in Symptomatic vs Asymptomatic Cohorts Stratified by Cardiac Damage Staging

J
A C C : A D V A N C E S , V O L . 3 , N O . 5 , 2 0 2 4

TABLE 1
Baseline Characteristics of Included Studies

TABLE 1
Society of ThoracicSurgeons Predicted Risk of Mortality; TAVR ¼ transcatheter aortic valve replacement.
Values are as n (%) or mean AE SD. a Included patients with chronic obstructive pulmonary disease, chronic bronchitis, or emphysema.b Population was moderate or severe asymptomatic aortic stenosis.AF ¼ atrial fibrillation; AFL ¼ atrial flutter; BMI ¼ body mass index; CABG ¼ coronary artery bypass graft; HLD ¼ dyslipidemia/hyperlipidemia; MI ¼ myocardial infarction; STS-PROM ¼

TABLE 3
Pooled Estimates of Characteristics in Symptomatic Severe AS, According to Cardiac Damage Stage