Predictors of health-related quality of life decline after transcatheter aortic valve replacement in older patients with severe aortic stenosis

Abstract

Background

Transcatheter aortic-valve implantation (TAVI) has been shown to improve survival and quality of life in patients with severe aortic stenosis. However, one-third of patients have poor outcome as death, functional decline or quality of life (QoL) decline. The aim of this study was to determine cardiac and geriatric predictors of physical and mental QoL decline 6 months after a TAVI procedure in patients aged 75 and older.

Methods

Between January 2013 and June 2014, we did a prospective and multicenter study including patients ≥ 75 years old referred for TAVI. The primary outcome was the measure of QoL, assessed by the Short Form 36 survey (SF-36), before and 6 months after the intervention. Association between QoL decline and baseline characteristics including cardiac and geriatric factors was analysed by logistic regression models.

Results

Mean age of the 150 patients studied was 83.7 years old and 56% were men. The primary end point, mean SF-36 physical summary score, significantly improved between baseline and 6-month (33.6 vs. 36.4, p=0.003) whereas mental component score significantly decreased (48.2 vs. 36.4, p-value<0.001). However, patients with presence of depressive symptoms before the intervention had mental QoL improvement at six months (OR 0.04 [0.01-0.19], p-value<0.001) and no significant geriatric predictors were associated with physical QoL decline.

Conclusion

The mental QoL significantly decreased and patients with preoperative depressive symptoms had mental QoL improvement at six months. Researches are needed to confirm that mental QoL of patients with depressive symptoms can be improved by TAVI.

Introduction

Aortic stenosis is the most common heart valve disease in patients aged 75 years and older (1). Although surgical aortic valve replacement improves survival and quality of life, at least one-third of older patients do not undergo surgery because of comorbidities or a high risk of preoperative mortality (2). For these patients, a less invasive treatment as transcatheter aortic valve implantation (TAVI) may be chosen. Indeed, numerous studies have proven that TAVI in older patients improves quality of life (QoL) and survival compared to medical treatment (3, 4, 5). However, it is established that some patients do not improve heart failure symptoms and functional limitation (5, 6, 7). Moreover, 15 to 25% of patients who underwent TAVI do not improve quality of life at 6 months (8, 9). In order to describe these patients, some studies tried to identify predictors of mortality, functional decline and worsened QoL (4, 5, 6, 7, 10). In theses studies, the predictors of poor outcomes were NYHA III or IV, oxygen-dependent lung disease, renal dysfunction, lower mean aortic valve gradient, a frailty index and the 6-minute walk test (4, 5, 6, 7, 10). Specific geriatric factors assessed as routinely during a comprehensive geriatric assessment (CGA), have not been independently analysed as predictors of mortality or functional decline or worsened QoL after TAVI (8). CGA developed by geriatricians is a multidisciplinary evaluation of comorbid medical conditions, cognitive and psychological status, nutritional status, functional status and social support (11).

The aim of this prospective study was to determine cardiac and geriatric predictors of health-related quality of life decline 6 months after a TAVI procedure in patients aged 75 and older. The secondary objectives were to analyse the predictors of mortality, the predictors of functional decline and post operative complications.

Methods

Study population

Between January 2013 and June 2014, all patients aged ≥ 75 years referred for a severe symptomatic aortic stenosis pre-operative evaluation in Nantes University Hospital or in Angers University Hospital were eligible for this prospective observational study. The inclusion criteria were 1) age of 75 years old and over, 2) presence of a symptomatic severe aortic stenosis considered severe if aortic valve area is < 1 cm² or a mean aortic-valve gradient is ≥ 40 mmHg, and 3) presence of a comprehensive pre-operative geriatric assessment. The following patients were excluded from the present study: patients under legal protection, patients for whom TAVI was done as an emergency procedure, patients with valve-in-valve procedure.

Patients underwent TAVI with the self-expandable Medtronic CoreValve bioprosthesis or the balloon-expandable Edwards Sapien transcatheter heart valve. The transcatheter aortic valve procedure was done under general anaesthesia, using the transfemoral approach whenever feasible.

This study was conducted in accordance with the ethical standards set forth in the declaration of Helsinki (1983). The Committee for the local Ethical Committee of Nantes (France) approved the project and the study is in compliance with the STROBE statement guidelines. All patients provided written informed consent for baseline and follow up assessment.

This study was registered at ClinicalTrials.gov. (Number: NCT01761357)

Baseline Data

An extensive cardiac and geriatric assessment was performed for all patients during the in-hospital pre-operative evaluation. This evaluation included the patient's history, his comorbidities, symptoms and a systematic clinical examination. The echocardiographic assessment provided information on aortic valve area (cm²), mean aortic-valve gradient (mmHg), left ventricular ejection fraction (LVEF), and pulmonary hypertension (PH). The coronary angiography and aortoiliofemoral computed tomography angiography completed the cardiologic assessment. The perioperative mortality risk was estimated by Logistic EuroSCORE, EuroSCORE II and the Society of Thoracic Surgeons score (STS). The geriatric assessment (CGA) performed by a geriatrician, included the following items: patients comorbidities evaluated by the Cumulative Illness Rating Scale for Geriatrics (CIRS-G) (12), Body Mass Index (BMI) for nutritional status, Mini-Mental State Examination (MMSE) and Frontal Assessment Battery (FAB) for cognitive assessment (13, 14), Instrumental activities of daily living of Lawton (IADL) for functional assessment (15), Timed Get Up and Go Test (TUG) for mobility impairment (16), mini-Geriatric Depression Scale (mini-GDS) (17) for presence of depressive symptoms and the presence of antidepressant treatment.

For the purpose of this analysis, geriatric test's scores were dichotomized at standard cut points which were defined according to current literature: TUG ≥ 20s (mobility impairment) vs. < 20s (normal gait function) (16), BMI ≤ 21 (malnutrition) vs. > 21 (normal nutrition), and mini-GDS ≥ 1 (presence of depressive symptoms) vs. mini-GDS=0 (no depression) (17). MMSE, FAB, CIRS-G, IADL were coded as continuous variables. CIRS-G score varies from very low (0) to high (56) burden of disease. IADL score ranges from 0 to 8. An IADL score of 8 indicates full function, and 2 or less indicates severe functional impairment (15).

Study end points

The main outcome was the decline of health-related quality of life decline 6 months after the TAVI procedure. The health-related quality of life (QoL) was evaluated with a standardized written questionnaire, the Short Form 36 (SF-36) tool. Each patient completed a questionnaire the day before the procedure. At 6 months, the questionnaire was completed by postal survey.

The SF-36 tool consists of 36 items self-administered questionnaire grouped in 8 dimensions of health parameters: physical functioning, role–physical, bodily pain, general health, vitality, social functioning, role–emotional, and mental health. These 8 scales can be summarized in 2 scores: the Physical Component Score (PCS) and the Mental Component Score (MCS). They are scaled to overall population norms of 50 and SDs of 10; higher scores meaning better QoL (18). The SF-36 has undergone extensive reliability and validity testing in various countries, in older patients and in patients with aortic stenosis (19). The SF-36 scale is currently used and validated in French (19). The individual QoL change was calculated by the difference in the physical component score and mental component scores between baseline and 6 months. Thresholds for minimal important difference have been suggested to define meaningful QoL improvement (20). But there are no thresholds for a difference meaning a “no QoL improvement” or a QoL decline with clinical relevance. In our study, a “no QoL improvement” or a QoL decline was defined by a difference in the PCS or the MCS ≤ zero at 6 months (18).

The secondary outcomes were death at 6 months, functional decline at 6 months assessed by IADL and complications by phone survey and medical records. Functional decline was defined as a decrease of ≥ 1 point in the ability to perform IADL between baseline and 6 months follow-up (15).

Clinical adverse events were defined according to the VARC criteria (21). Complications of interest were stroke, vascular complications requiring unplanned endovascular or surgical intervention, implantation of pace maker or new permanent atrial fibrillation, acute renal failure defined as an absolute increase in serum creatinine ≥ 200%, and other TAVI-related complication as ventricular perforation or cardiac tamponade. Geriatric complications were also gathered as delirium, falls or acute failure of chronic diseases requiring hospitalisation.

Statistical analysis

The participants' baseline characteristics were summarized using means ± standard deviations for continuous variables or counts and percentages for categorical variables. Patient's characteristics were compared between 2 groups according the presence of physical or mental QoL decline, by Fisher exact test for dichotomous variables and by t-test for means of continuous variables.

Baseline predictors of quality of life decline that were found with a p-value less than 0.20 in univariate analysis, confounding factors and variables of the main objective (MMS, FAB, BMI, CIRS-G, IADL, mini-GDS) were included in multivariate analyses. Baseline predictors of functional decline and mortality that were found with a p-value less than 0.20 in univariate analysis, confounding factors and variables of the main objective were respectively included in multivariate logistic regression or multivariate Cox model.

Odds ratio or Hazard ratio with their 95 % confidence intervals were presented. All p-value were two sided. p-values less than 0.05 were considered statistically significant. All statistics were performed using SAS(r) (version 9.3 SAS Institute Inc., Cary, NC, USA).

Results

Between January 2013 and June 2014, 193 patients ≥ 75 years were referred for a TAVI procedure. Forty three patients were excluded (2 patients under legal protection, 24 patients with no geriatric baseline examination or emergency procedure, 13 patient's refusal, 4 with valve-in-valve procedures). The study population consisted of 150 patients.

Patient characteristics

The mean age of the study population was 83.7±4.6 years (range 77-95) and 56% were men. A total of 73 patients (48.7%) were in NYHA functional class III or IV. The mean aortic-valve gradient was 51.7±14.8 mmHg. The mean Logistic EuroSCORE was 17.0±8.1%, and the mean STS score was 7.25±3.7%. The geriatric assessment revealed that patients had many comorbidities with a high score of CIRS-G (10.8±3.1), 55.3% had impaired gait function. The mean MMS was of 24.9±3.5. Patients had limited ability to performed activities of daily living (mean IADL= 5.5±2.2).

At baseline, the physical SF-36 component score and the mental SF-36 component score were respectively of 33.6±8.4 and 55.6±8.1. Baseline characteristics of patients with or without QoL decline at 6 months are shown in Table 1.

Table 1.

Baseline Characteristics of Patients with or without Physical QoL Decline at 6 months

Characteristics	All participants n= 150	Physical QoL Decline			Mental QoL Decline
		No n=72	Yes n=47	p-value	No n= 30	Yes n=89	p-value
Age, years, mean ± SD	83.7 ± 4.6	83.6 ± 4.4	84.0 ± 4.7	0.77	84.1 ± 4.6	82.6 ± 4.0	0.03
Male, n (%)	84 (56.0)	38 (52.8)	28 (59.6)	0.89	16 (53.3)	37 (41.6)	0.14
Dyspnea NHYA III or IV, n, (%)	73 (48.7)	33 (45.8)	20 (42.5)	0.21	15 (50.0)	38 (42.7)	0.84
Mean gradient aortic valve, mmHg, mean ± SD	51.7 ± 14.9	52.9 ± 14.4	51.2 ± 16.4	0.56	52.8 ± 15.4	50.6 ± 14.7	0.49
Ejection Fraction, mmHg, mean ± SD	55.9 ± 10.7	55.8 ± 11.1	54.9 ± 11.4	0.68	54.1 ± 11.5	59.4 ± 9.2	0.05
Pulmonary hypertension, n (%)	53 (34.7)	27 (37.5)	14 (29.8)	0.71	10 (33.3)	31 (34.8)	0.52
Logistic EuroSCORE, mean ± SD	17.0 ± 8.1	16 ± 7.2	16.9 ± 7.8	0.49	16.4 ± 7.4	16.3 ± 7.7	0.14
EuroSCORE 2, mean ± SD	5.0 ± 2.5	4.9 ± 2.4	5.1 ± 2.7	0.69	4.9 ±2.5	5.2 ± 2.6	0.66
STS risk score, mean ± SD	7.2 ± 3.7	6.6 ± 3.0	7.9 ± 4.1	0.07	6.9 ±3.2	7.6 ± 4.3	0.94
CIRS-G score, mean ± SD	10.8 ± 3.1	10.6 ± 3.3	10.8 ± 3.1	0.72	10.3 ± 3.1	11.6 ± 3.2	0.96
Body mass index ≤ 21, n (%)	18 (12.0)	11 (15.3)	5 (10.6)	0.66	6 (20.0)	10 (11.2)	0.43
Mini-Mental Status score, mean ± SD	24.9 ± 3.5	25.3 ± 3.3	24.7 ± 3.4	0.29	25.9 ± 3.3	24.4 ± 3.6	0.88
Frontal Assessment Battery score, mean ± SD	12.8 ± 3.2	13.1 ± 3.1	12.7 ± 3.1	0.47	12.9 ±2.9	12.9 ± 3.5	0.79
Depressive symptoms, n (%)	23 (15.3)	12 (16.7)	8 (17)	0.89	12 (36.4)	8 (8.9)	0.02
Antidepressant treatment, n (%)	13 (8.6)	6 (8.3)	5 (10.6)	0.99	5 (16.7)	5 (5.6)	0.28
IADL, mean ± SD	5.5 ± 2.2	5.6 ± 2.2	5.5 ± 2.1	0.97	5.4 ± 2.2	5.9 ± 2.0	0.16
Timed Up and Go test, n (%)	83 (55.3)	38 (52.8)	24 (51.1)	0.71	14 (46.6)	48 (53.9)	0.66

CIRS-G: Cumulative Illness Rating Scale-Geriatric (range 0 to 56) IADL: Instrumental activities of daily living score ranges from 0 (total dependence) to 8 (independent) STS: Society of Thoracic Surgeons

Follow up Data

During 6-month follow up, 12 patients died (8.0%) and 19 (12.6%) were lost to follow up. At 6-month follow-up, 47 patients (31.3%) had physical QoL decline and 89 (59.3%) had mental QoL decline.

Overall, the improvement of the physical component score was significant (33.6 vs. 36.4, p-value=0.003) whereas mental component score significantly decreased (55.6 vs. 48.2, p-value <0.001) (Figure 1). Sixty four patients (53.8%) had 6-month functional decline.

Figure 1.

Short Form 36-Item Health Survey scores

Predictors of QoL decline among survivors

There were no group differences age or gender ratios. There were no significant intergroup differences in any of the cardiac or geriatric factors (Table 1).

In the analysis, neither cardiac nor geriatric predictors were significantly associated with physical QoL decline at 6 months (Table 2).

Table 2.

Association of Preprocedure Factors with Physical QoL Decline 6 months after Transcatheter Aortic Valve Implantation, in multivariate analysis

Predictors	Physical QoL Decline
	OR (95% CI)	p-value*
Age, years, mean ± SD	1.03 (0.94-1.12)	0.55
Male, n (%)	0.99 (0.42-2.33)	0.55
Ejection fraction, mmHg, mean ± SD	0.99 (0.95-1.03)	0.52
CIRS-G score, mean ± SD	1.03 (0.89-1.19)	0.71
Body mass index, n (%)	0.53 (0.12-2.32)	0.40
Mini-Mental Status score, mean ± SD	1.71 (0.60-4.85)	0.31
Frontal Assessment Battery score, mean ± SD	0.72 (0.28-1.86)	0.5
Depressive symptoms, n (%)	1.28 (0.42-3.84)	0.67
IADL, mean ± SD	0.99 (0.81-1.19)	0.87
Timed Up and Go test, n (%)	0.71 (0.31-1.64)	0.43

CIRS-G: Cumulative Illness Rating Scale-Geriatric; IADL: Instrumental activities of daily living, score ranges from 0 (total dependence) to 8 (independent); p-value and OR significant (i.e., p < 0.05) indicated in bold

^*Between-group comparison based on simple t-test or Chi-square test, as appropriate

Table 3 shows the results for the prediction of mental QoL decline at 6 months. The presence of depressive symptoms before the intervention was strongly and significantly associated with mental QoL (OR 0.04 [0.01-0.19], p-value<0.001).

Table 3.

Association of Preprocedure Factors with Mental QoL Decline 6 months after Transcatheter Aortic Valve Implantation, in multivariate analysis

Predictors	Mental QoL Decline
	OR (95% CI)	p-value*
Age, years, mean ± SD	1.03 (0.92-1.16)	0.63
Male, n (%)	1.44 (0.44-4.69)	0.54
Ejection fraction, mmHg, mean ± SD	0.98 (0.93-1.03)	0.41
CIRS-G score, mean ± SD	0.97 (0.81-1.17)	0.78
Body mass index, n (%)	1.03 (0.09-11.11)	0.98
Mini-Mental Status score, mean ± SD	0.42 (0.1-1.75)	0.23
Frontal Assessment Battery score, mean ± SD	1.58 (0.41-6.16)	0.51
Risk of depression, n (%)	0.04 (0.01-0.19)	<0.001
IADL, mean ± SD	1.01 (0.78-1.31)	0.94
Timed Up and Go test, n (%)	0.71 (0.31-1.64)	0.04

CIRS-G: Cumulative Illness Rating Scale-Geriatric; IADL: Instrumental activities of daily living, score ranges from 0 (total dependence) to 8 (independent); p-value and OR significant (i.e., p<0.05) indicated in bold

^*Between-group comparison based on simple t-test or Chi-square test, as appropriate

Secondary analyses

Predictors of mortality

The Logistic EuroSCORE was significantly associated with death in the multivariate analysis (Table 4). No geriatric predictors were associated with mortality.

Table 4.

Association of Preprocedure Factors with 6-month Mortality after Transcatheter Aortic Valve Implantation, in multivariate analysis

Predictors	Mortality
	HR (95% CI)	p-value
Age, years, mean ± SD	0.92 (0.79-1.08)	0.35
Male, n (%)	1.48 (0.43-5.17)	0.53
NYHA III or IV, n (%)	1.64 (0.44-6.11)	0.46
Pulmonary hypertension, n (%)	0.75 (0.20-2.77)	0.67
Ejection fraction, mmHg, mean ± SD	1.02 (0.96-1.09)	0.41
Logistic EuroSCORE, mean ± SD	1.11 (1.03-1.19)	0.003
CIRS-G score, mean ± SD	1.08 (0.84-1.39)	0.52
Body mass index, n (%)	0.97 (0.86-1.08)	0.54
Mini-Mental Status score, mean ± SD	1.02 (0.82-1.26)	0.84
Frontal Assessment Battery score, mean ± SD	1.05 (0.89-1.24)	0.54
Depressive symptoms, n (%)	5.55 (0.58-50)	0.14
IADL, mean ± SD	1.31 (0.94-1.82)	0.11
Timed Up and Go test, n (%)	0.39 (0.11-1.41)	0.15

CIRS-G: Cumulative Illness Rating Scale-Geriatric; IADL: Instrumental activities of daily living, score ranges from 0 (total dependence) to 8 (independent); NYHA: New York Heart Association; p-value and OR significant (i.e., p<0.05) indicated in bol; * Between-group comparison based on multivariate Cox model.

Functional Status

Functional status for Instrumental Activities of Daily Living significantly decreased after the procedure (5.5±2.2 vs. 4.9±1.9, p<0.001). No geriatric predictors were significantly associated with functional decline (Table 5).

Table 5.

Association of Preprocedure Factors with 6-month Functional Decline after Transcatheter Aortic Valve Implantation, in multivariate analysis

Predictors	Functional Decline
	OR (95% CI)	p-value
Age, years, mean ± SD	1.04 (0.95-1.15)	0.38
Male, n (%)	0.44 (0.2-1.01)	0.05
Logistic EuroSCORE, mean ± SD	0.86 (0.71-1.03)	0.11
STS score, mean ± SD	1.03 (0.89-1.18)	0.71
CIRS-G score, mean ± SD	1.09 (0.92-1.31)	0.29
Body mass index, mean ± SD	1.08 (0.97-1.23)	0.15
Mini-Mental Status score, mean ± SD	0.93 (0.78-1.1)	0.38
Frontal Assessment Battery score, mean ± SD	1.18 (0.99-1.41)	0.06
Depressive symptoms, n (%)	1.26 (0.42-3.70)	0.68
Timed Up and Go test, n (%)	1.40 (0.66-3.22)	0.41

CIRS-G: Cumulative Illness Rating Scale-Geriatric; IADL: Instrumental activities of daily living, score ranges from 0 (total dependence) to 8 (independent); STS: Society of Thoracic Surgeons; p-value and OR significant (i.e., p<0.05) indicated in bold; * Between-group comparison based on simple t-test or Chi-square test, as appropriate

Complications

The rate of strokes after TAVI was of 2% (3 patients). Permanent pace maker implantation was the most frequent adverse effect, occurring in 25 patients (16.6%) and seven patients (4.6%) had a new permanent atrial fibrillation. Vascular complications after TAVI occurred in 13 patients (8.6%) and 8 patients (5.3%) had pericardial effusion with tamponade or ventricular perforation, only one had a conversion to surgery. The overall rate of acute renal failure was of 6%.

The incidence of delirium during hospital stay was of 12% (18 patients) and falls occurred in 11 patients (7.3%). Eleven patients had an acute failure of chronic diseases requiring hospitalisation.

Discussion

This survey is the first prospective study specially designed to describe geriatric factors associated with health-related quality of life decline after TAVI procedure. Overall, patients had significant physical QoL improvement and significant mental QoL decrease 6 months after TAVI. One-third of patients had physical QoL decline at 6 months but no predictor has been identified. Half of the population study had 6-month mental QoL decline and patients with presence of depressive symptoms before the intervention had 6 month mental QoL improvement. No geriatric factor was associated with functional decline. Finally, Logistic EuroSCORE was the only factor associated with 6 month-mortality.

In this study, patients had significant physical QoL improvement. Our findings are consistent with previous studies. Indeed, numerous studies shown that physical QoL improves at 6 months and later (5, 22) but these studies did not try to find out predictors of improvement or decline. Even though we did not found any geriatric predictors, our study is the first one to include geriatric factors gathered during a pre-operative complete geriatric assessment. Only one other study analysed predictors of poor outcomes defined by poor physical QoL and mortality after TAVI (PARTNER trial) (8). In their survey, the geriatric factor analysed were the MMS, the Body Mass Index and the gait speed. Gait speed, assessed by 6-minute walk test was significantly associated with a poor outcome at 6 months [8]. Two reasons can explain our different results. First, the methods were different especially with different outcome: physical QoL alone versus an arbitrary construct endpoint combining poor physical QoL and mortality. Second, the population-based studies were slightly different. Even though the mean age, predominance of male and cardiac factors were similar; the mean MMS score was lower and the mobility impairment at baseline was more frequent in our study. In the present study, mental QoL life significantly declined 6 months after TAVI. In previous studies, results on mental QoL evolution were contrasted. Studies using SF-36 tool found improvement or decline but did not reach statistical significance (22, 23). One study using another tool, the SF-12 tool, found a significant mental QoL improvement (5). However, our study showed that the patients with presence of depressive symptoms before the intervention had mental QoL improvement at six months. Our results differ from previous studies in cardiac surgery in which depression was significantly associated with health QoL decline (24, 25). However, our results are not surprising considering complications are more frequent following surgical aortic replacement than after TAVI in particularly in older patients (26, 27).

The rate of mortality was of 8% in our study, which is coherent with the literature (4, 10). The 6-month mortality was associated with Logistic EuroSCORE in our study. Our results are consistent with numerous studies (4, 10, 28). Indeed, cardiac predictors of mortality after TAVI are well known as NYHA class, Logistic EuroSCORE, LVEF, and access route. However, current surgical risk scores predict 30-day survival after conventional surgery and are used to identify high risk patients. Although these risk scores are associated with increased mortality after TAVI, they are not designed nor validated to assess mortality risk for TAVI. Few studies that have evaluated the value of conventional surgical risk scores to predict 1-year mortality after TAVI concluded that in practice TAVI should be guided by the interdisciplinary team decision (29, 30).

In our survey, the functional status assessed by the ability to perform the IADL declined in 53.8% of patients. In the only previous study that assessed functional decline after TAVI, Schoenenberger and al. found that 20.8% of patients had functional decline (7). These different results can be explained by the different endpoint. Indeed, in their study, the functional decline was defined as a decrease of ≥1 point in the ability to perform Activities of Daily Living (ADL) between baseline and follow-up. The ADL ranks adequacy of performance in everyday tasks as bathing or eating (31). The IADL, used in our study, has been shown to be more sensitive in detection of early stage of functional decline, because it measures higher order functioning such as managing medications and money (15, 32). Schoenenberger and al. also found that a frailty index was associated with functional decline. In our study, we did not found any independent geriatric predictor. Methodological reasons can explain our different results. In our study, the frailty index wasn't measured because there is no clear consensual definition of frailty and a large array of criteria has been proposed to define and assess this syndrome. The pre-operative geriatric assessment was based on consensual and validated scores done routinely in geriatric evaluation.

The incidences of cardiac complications were similar in our study than in previous studies (33). However this study is the first to show geriatric complications occurring in post-interventional period. About delirium, our findings are consistent with the only previous study. Tse and al. retrospectively showed that delirium occurred between 12% and 43% of patients depending on access route (34). There is no published study on occurrence of falls or acute failure of chronic disease. Acute failure of chronic disease seemed an interesting factor and may explain why some patients had quality of life decline after TAVI procedure.

The strengths of our study include the prospective cohort study design, the multicenter study and the absence of sponsor, which minimizes the potential for selection bias. However because of the predominance of non-significant results, it seemed important to look through the methods to exclude any bias. Although numerous methods exist to evaluate health-related QoL, the SF-36 health survey questionnaire was chosen because it is a validated generic scale for cardiac patients and for older patients (35, 36, 37). Furthermore, previous studies had proved improvement of quality of life after TAVI procedure using SF-36 scale (22). Therefore the study design seemed appropriate. Our consistent results as the population's study age, the significant physical QoL improvement and the incidence of cardiac complications confirm the study's strength.

This study has some limitations. The first limitation is the lack of the NYHA score evolution and echocardiography parameters between baseline and 6-month evaluation. This additional information could have helped to understand the QoL evolution especially in patients with QoL decline. However, it would not have changed results about predictors of QoL decline.

Second, the sample size was small. Further studies of larger patient populations are required to confirm our results.

In conclusion, this study showed that the mental QoL significantly declined and patients with presence of depressive symptoms before the intervention had mental QoL improvement at six months. No cardiac or geriatric predictors were associated with the physical QoL decline.

The present study has research implications. First, larger studies with complete geriatric assessment are needed to confirm geriatric characteristic of the TAVI population, and consensual geriatric scores as predictors of a health-related quality of life decline. Second, researches are needed to confirm that mental QoL of patients with depressive symptoms can be improved by TAVI.

Ethical Standards: This study was conducted in accordance with the ethical standards set forth in the declaration of Helsinki (1983). The Committee for the local Ethical Committee of Nantes (France) approved the project and the study is in compliance with the STROBE statement guidelines. All patients provided written informed consent for baseline and follow up assessment. This study was registered at ClinicalTrials.gov. (Number: NCT01761357)

Acknowledgements: We are grateful to the participants for their cooperation. Authors Contribution: - Boureau has full access to all study data and takes responsibility for the integrity of the data and the accuracy of the data analyses. - Study concept and design: Boureau and de Decker. - Acquisition of data: Boureau, Rouaud, Manigold, Jaafar, Letocart. - Analysis and interpretation of data: Boureau, de Decker, Guerin, Trochu, Berrut. - Drafting of the manuscript: Boureau, de Decker. - Critical revision of the manuscript for important intellectual content: de Decker, Trochu, Berrut. - Statistical expertise: de Decker. - Administrative, technical, or material support: de Decker. - Study supervision: de Decker.

Conflict of interest: The authors declare no conflict of interest.

Funding Source: None.

Sponsor Role: None.