Transcatheter Aortic Valve Replacement in the Care of Older Persons with Aortic Stenosis

This past decade has certainly been one of great innovation in the treatment of heart valve disease, most importantly in the treatment of severe aortic stenosis with the arrival and dissemination of transcatheter aortic valve replacement (TAVR) as a viable alternative to surgical aortic valve replacement (SAVR) in high-risk patient populations. Due to its natural history, calcific degenerative aortic stenosis tends to be a disease of the aged, and it is only fair that it should draw the attention of the geriatric community as the care of patients with this condition is prone to be shared between cardiologists and geriatricians. In the following lines, we will review the existing body of literature on TAVR, focusing on (i) seminal randomized clinical trials that have proven the effectiveness and safety of this procedure, and (ii) observational studies that have highlighted the importance of geriatric domains, mainly frailty, on prognosis and decision-making. Two major randomized clinical trial programs, PARTNER and CoreValve, have evaluated TAVR as compared to the standard-of-care; the former using Edwards Sapien devices and the latter using Medtronic CoreValve devices. Both programs included patients with severe calcific aortic stenosis and NYHA class II or greater symptoms, and excluded patients with congenitally abnormal valves, left ventricular ejection fraction <20%, recent myocardial infarction <1 month, recent GI bleed <3 months, recent stroke <6 months, limited life expectancy <12 months, and severe dementia resulting in inability to provide informed consent, to live independently outside of a chronic care facility, or to be compliant with rehabilitation or follow-up visits. Patients deemed to be at very high “prohibitive” surgical risk, defined as >50% predicted likelihood of death or serious irreversible morbidity at 30 days, were enrolled in the PARTNER 1B and CoreValve Extreme Risk trials wherein TAVR was compared to conservative therapy (which, in many cases, involved balloon valvuloplasty).1-3 Frailty was cited as the main reason for prohibitive risk in 31% of patients.4 Patients deemed to be at high surgical risk, defined as >15% predicted likelihood of death at 30 days, were enrolled in the PARTNER 1A and CoreValve U.S. Pivotal trials wherein TAVR was compared to SAVR in order to demonstrate non-inferiority of the two procedures.5-8 More recently, patients deemed to be at intermediate surgical risk, defined as 4–8% predicted likelihood of death at 30 days, were enrolled in the PARTNER 2 and SURTAVI trials wherein TAVR was compared to SAVR using a non-inferiority design.9, 10 No age-specific cutoff was used for enrollment and participants’ mean age was 82–84 years with a standard deviation of 7–9 years.1, 2, 5, 6 Few geriatric measures aside from frailty were reported, with the available data confirming that the population undergoing TAVR was indeed frail and had on average 5 comorbid chronic conditions.1, 2, 5, 6 Disability and cognitive function were collected at baseline with the Katz Activities of Daily Living (ADL) and Mini-Mental Status Examination (MMSE) instruments, but these were not always reported in the published manuscripts. In the CoreValve trials, 80–84% of patients had slow 5-meter gait speed, 67% had weak handgrip strength, 10–22% had ADL disability, 18% had recent falls, and 28% had MMSE scores ≤24.1, 5 In the PARTNER 1A and 1B trials, the median gait speed was strikingly low at 0.38 m/s, median handgrip strength was 23.6 kg in men and 12.2 kg in women, and 29% had ADL disability.11 The primary endpoint was death from any cause at 1 year, and pre-specified secondary endpoints included: stroke, acute kidney injury, vascular complications, bleeding complications, echocardiographic valve performance, NYHA class, 6-minute walk distance, and need for repeat hospitalization. Patient-reported outcomes were collected, in particular, quality-of-life (QOL). Results for these endpoints are summarized in Table 1. In brief, for patients at prohibitive risk, TAVR was markedly superior to conservative therapy in terms of symptomatic improvement, repeat hospitalizations, and survival at 1 year; at the expense of a higher risk of peri-procedural stroke, bleeding, and vascular complications.1, 2 For patients at intermediate or high risk, TAVR was as equivalent (if not slightly superior) to SAVR in terms of survival at 1 year, superior for bleeding complications, postoperative atrial fibrillation, length of stay, and short-term symptomatic improvement, and inferior for vascular complications, residual aortic regurgitation, and permanent pacemaker implantation (mainly due to CoreValve-associated heart block).5, 6 The benefits of TAVR were similar among subgroups >85 and ≤85 years of age. Several sub-studies12-15 have shed light on QOL outcomes from the PARTNER and CoreValve trials, which, for a geriatric population, are measures of great interest. In these studies, QOL was measured using the Kansas City Cardiomyopathy Questionnaire (KCCQ) as well as with the SF-12 and EQ-5D instruments. Results showed that prohibitive risk patients who underwent TAVR enjoyed better QOL at 1, 6, and 12 months as compared to those treated conservatively. Nevertheless, 1 out of 3 patients still suffered very poor or worsening QOL or death at 6 months, collectively termed “poor outcomes”, and this increased to 1 out of 2 patients at 12 months.16, 17 High-risk patients who underwent TAVR via a trans-femoral approach enjoyed better QOL at 1 month and similar QOL at 6 and 12 months as compared to those who underwent SAVR or TAVR via a more invasive trans-apical approach. Improvement in QOL scores were generally of a large magnitude (+23 to 33 KCCQ points),12-15 and were noted across sub-scales pertaining to symptoms, physical limitations, social limitations, and mental functioning. The Society of Thoracic Surgeons (STS) risk score is the most commonly used tool to predict risk in cardiovascular surgery (http://riskcalc.sts.org), although one of its blind spots is frailty. To better identify older adults at greater risk of adverse outcomes after TAVR, Hermillier et al.18 analyzed data from the CoreValve trials and identified the following predictors of death at 1 year: falls in the past 6 months, Charlson comorbidity index ≥5, low serum albumin level at baseline, high STS risk score (≥7.0%), and use of home oxygen. The following predictors were not retained in the final model: gait speed, handgrip strength, weight loss, Katz ADL, and being wheelchair bound. Rather than focusing solely on mortality, Arnold et al.16, 17 analyzed data from the PARTNER and CoreValve trials and identified the following predictors of “poor (patient-centered) outcomes” at 6 months: low 6-minute walk distance, low mean aortic gradient, use of home oxygen, renal dysfunction, cardiac arrhythmia, cognitive dysfunction as measured by the MMSE, frailty as measured by the Fried scale (with weight loss being the most predictive domain in the scale), and disability as measured by the Katz ADL index. From early on, investigators interested in outcomes post-TAVR have studied frailty as one of its key predictors. The reasons for this interest are multiple and pertain to (i) the high burden of frailty in this complex geriatric population, (ii) the proven ability of frailty to improve risk prediction and thus guide decision making in cardiac surgery and other settings,19 (iii) the ease of use of certain measures of frailty such as gait speed and grip strength, and (iv) the potential ability of frailty to serve as a therapeutic target and improve outcomes. Our systematic review of the literature found 20 studies that focused on the implications of frailty in patients undergoing TAVR; these are reviewed in Table 2 and discussed below. 28% >10 seconds 76% >6 seconds 30-day mortality: 8.4% vs. 5.4% (>10 seconds vs. <6 seconds) In-hospital morbidity: nonsignificant except vascular complications Discharge to a facility: 47% vs. 22% Length of stay: 7 days vs. 5 days Fried scale Katz ADL disability 60% 17% 30-day mortality: OR 4.03 2.2-year mortality: OR 2.15 *Katz, Karnofsky, CHSA, Brighton scales nonsignificant 1-year mortality: 17% vs. 7%, HR 3.51 30-day mortality/morbidity: nonsignificant except major bleeding complications Length of stay: 9 days vs. 6 days 1-year mortality: 33% vs. 16%, HR 2.5 6-month “poor outcome”: 42% vs. 28%, OR 2.21 30-day mortality/morbidity: nonsignificant Albumin <3.3 g/dL Recent fall <6 months 17% 20% 3-month mortality: nonsignificant 30-day mortality/morbidity: nonsignificant Discharge to facility: 39% vs. 14%, OR 4.80 1-year mortality: 60% vs. 11%, HR 2.74 per point In-hospital mortality/morbidity: nonsignificant *ISAR scale nonsignificant 1-year mortality: HR 1.49 30-day mortality: nonsignificant Albumin <3.3 g/dL Wheelchair bound 18% 16% 2-year mortality: OR 0.56 per sex-stratified SD 30-day mortality/morbidity: OR 0.52 ICU LOS >7 days, LOS >14 days, or readmission: OR 0.56 1.5-year mortality: HR 2.67 30-day mortality: 17% vs. 6% 30-day complications: nonsignificant except acute kidney injury and need for PRBC Length of stay >14 days: 32% vs. 22% Discharge to facility/nursing: 53% vs. 9% 8-month mortality: nonsignificant 30-day mortality: nonsignificant Procedural complications: nonsignificant except need for dialysis 6-month ADL change ≥1: OR 3.34 6-month mortality: 18.6% vs. 3.3% 1-year mortality: OR 2.93 1-year cardiovascular/cerebral events: OR 4.89 The studies reviewed were published between 2011–2016 and consisted of prospective and retrospective cohort studies. Sample sizes ranged between 100–460 patients, with the exception of two studies containing 2,830–3,687 patients from the CoreValve trials16, 18 and one study containing 8,039 patients from the STS/American College of Cardiology Transcatheter Valve Therapy registry (STS/ACC TVT registry).20 Much between-study variability was observed with respect to the operating definition of frailty, which contributed to discrepancies in the proportion of frail patients encountered in each individual study, ranging from 33% to 76%. Frailty scales were generally based on variations of the frailty phenotype construct,21, 22 encompassing domains of physical performance (gait speed, handgrip strength), sarcopenia (CT-measured muscle area, self-reported weight loss), malnutrition (mini nutritional assessment, serum albumin), and often amalgamating ADL disability within the scale—even though disability is a distinct concept that most would argue should be disentangled from frailty.23 Using a traditional cutoff of >6 seconds to walk 5 meters, gait speed was found to be ubiquitously slow in >75% of patients, rendering its sensitivity high but its specificity very low to identify high-risk patients.20 A cutoff of >10 seconds to walk 5 meters, or worse yet, being wheelchair bound or unable to complete the 5-meter gait speed test were found to be more predictive in this patient population.14 Other high-risk frailty indicators were low serum albumin (<3.5 g/dL), ADL disability (≥1–2 dependencies), and unintentional weight loss. Given the inaccuracy of self-reported weight loss as a surrogate for muscle loss, investigators have used CT images to measure muscle area (CT scans are routinely acquired as part of the pre-TAVR work-up). The McGill-Munich Study was among the first to report the prognostic value of measuring psoas muscle area on a single axial CT image at the level of the L4 vertebrae using a web-based software tool (https://www.coreslicer.com).24 The recently completed Frailty-AVR study compared the prognostic value of the various frailty scales in 1,012 older adults undergoing TAVR and SAVR, and found that the short physical performance battery (SPPB) outperformed other scales to predict 1-year mortality and disability.25 The SPPB, which has been extensively validated in the geriatric literature,26, 27 consists of 5-meter gait speed, timed chair rises, and timed standing balance. Prediction was further improved by considering serum albumin, hemoglobin, and cognitive function. Despite the multitude of frailty scales used in the studies reviewed, similar observations have been made across studies. Frailty has consistently been associated with a two-to-threefold greater risk of death 1–2 years after TAVR. It also has variably been associated with short-term risk of death and complications, with some studies reporting a positive association20, 28-30 and others failing to demonstrate it in a statistically significant fashion.11, 31-35 When major complications did arise, these were more likely to be fatal in frail patients.29 Furthermore, frailty has been associated with lengthier hospital stays, a lower likelihood of being discharged home,20, 29, 32 and a greater risk of functional decline 6–12 months after TAVR.31, 36 Geriatric domains play a central role in the evaluation of older adults with severe symptomatic aortic stenosis, who may accordingly be counseled and guided towards SAVR, TAVR, or conservative medical management when an intervention is likely to be futile.37 Based on the evidence to-date, our recommendation is to adopt a tiered approach starting with an assessment of physical performance, ADL disability, and cognitive function; when deficits are encountered, deeper phenotypic characterization is indicated including comprehensive geriatric assessment. Through gains in operator expertise and valve design, the technical success of the TAVR procedure has reached 96% and the unmet challenge has become achieving QOL success which is subpar in 50% of patients. Thus, geriatric domains will likely shift from prognostic markers to therapeutic targets, earmarking frail patients that could benefit from pre- or post-procedural interventions aimed at optimizing physical recovery and preventing progressive disability. Multi-faceted interventions combining structured exercise and nutritional supplementation have shown promising results,38, 39 and are being investigated in frail TAVR patients. Optimization of patient-centered outcomes will undoubtedly require the close collaboration of cardiovascular and geriatric specialists as well as allied health professionals. The authors would like to thank Dr. Jose Morais (Director, Division of Geriatric Medicine, McGill University) for his ongoing mentorship, and the Montoni Foundation for their generous support. Financial Disclosure: Dr. Afilalo is supported by the Canadian Institutes of Health Research and Fond de recherche du Québec – Santé. Conflicts of Interest: None. Authors Contribution: Dr. Catherine Talbot-Hamon and Dr. Jonathan Afilalo contributed to the literature review, data synthesis, and manuscript production. Sponsor's Role: None.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.