Transcatheter aortic valve replacement (TAVR) is an effective treatment for severe symptomatic aortic stenosis (AS) in patients who are inoperable or at high risk for surgery. However, the intermediate- to long-term mortality is high, emphasizing the importance of patient selection. We, therefore, sought to evaluate the prognostic value of frailty in older recipients of TAVR, hypothesizing that frail patients would experience a higher mortality rate and a higher likelihood of poor outcome 1 year after TAVR. This substudy of the Placement of Aortic Transcatheter Valves trial was conducted at 3 high-enrolling sites where frailty was assessed systematically before TAVR. In total, 244 patients received TAVR at the participating sites. Frailty was assessed using a composite of 4 markers (serum albumin, dominant handgrip strength, gait speed, and Katz activity of daily living survey), which were combined into a frailty score. The cohort was dichotomized at median frailty score. Outcomes measures were the time to death from any cause for >1 year of follow-up and poor outcome at 1 year. Poor outcome was defined as (1) death, (2) Kansas City Cardiomyopathy Questionnaire overall summary (KCCQ-OS) score <60, or (3) decrease of ≥10 points in the KCCQ-OS score from baseline to 1 year. At 1 year, the Kaplan-Meier–estimated all-cause mortality rate was 32.7% in the frail group and 15.9% in the nonfrail group (log-rank p = 0.004). At 1 year, poor outcome occurred in 50.0% of the frail group and 31.5% of the nonfrail group (p = 0.02). In conclusion, frailty was associated with increased mortality and a higher rate of poor outcome 1 year after TAVR.
Transcatheter aortic valve replacement (TAVR) can be an effective treatment option for older adults with severe symptomatic aortic stenosis (AS) who are inoperable or at high risk for traditional surgery, However, the intermediate- to long-term mortality of patients who underwent TAVR is high, which is largely reflective of the underlying age and co-morbidities of the treated population. Accordingly, there have been numerous efforts to understand which patients are unlikely to benefit from TAVR—either from a survival or quality of life standpoint. In these analyses, several clinical and physiological factors have been associated with a higher risk of poor outcomes. In elderly populations, frailty—a syndrome of impaired physiological reserve and decreased resistance to stressors —has emerged as an important risk factor for morbidity and mortality in multiple clinical situations, including the general population, older adults with coronary artery disease, and recovery after general and cardiac surgery. In addition, small, single-center studies have shown that frailty is associated with increased morbidity and mortality after TAVR. However, the relation between frailty and quality of life after TAVR is unknown. Accordingly, we sought to evaluate the prognostic value of frailty in older adults who received TAVR in the Placement of Aortic Transcatheter Valve (PARTNER) Trial. We hypothesized that frail patients who undergo TAVR would experience increased mortality and a higher likelihood of a poor quality of life after TAVR, even after accounting for traditionally measured clinical risk factors.
Methods
The design and initial results of the PARTNER trial have been published previously. The PARTNER trial enrolled patients with severe symptomatic AS into 2 cohorts: those at high surgical risk (cohort A) and those considered inoperable because of severe coexisting conditions (cohort B). Patients in cohort B with a suitable iliofemoral vessel were randomized to transfemoral TAVR with the Edwards-Sapien heart valve system (Edwards Lifesciences, Irvine, California) or to standard medical care. Patients in cohort A were randomized to TAVR (transfemoral if iliofemoral vessels were suitable or transapical if not) or to conventional surgical aortic valve replacement. The present analyses pooled patients from both cohorts A and B (n = 244) who were randomized to and received TAVR at 1 of 3 high-enrolling PARTNER trial sites where frailty was assessed systematically before TAVR (Medical City Dallas Hospital, Dallas, Texas [n = 35]; Mayo Clinic, Rochester, Minnesota [n = 83]; Columbia University Medical Center, New York, New York [n = 126]). Of the 1,057 participants enrolled in the PARTNER trial, 519 patients were randomized to and received TAVR (344 in cohort A and 175 in cohort B). Of these, 244 patients (47%; 215 in cohort A and 29 in cohort B) also completed a baseline frailty assessment at 3 high-volume centers and were included in this analysis. The study was approved by the institutional review board at each participating site, and all patients provided written informed consent.
Frailty is defined as a syndrome of impaired physiological reserve and decreased resistance to stressors and is captured by the core domains of wasting and malnutrition, exhaustion and inactivity, weakness, and slowness. For this study, we operationalized frailty using a composite of 4 markers, which were chosen to parallel those operationalized by Fried et al. Malnutrition and wasting were assessed using serum albumin measured on the day before TAVR. Weakness was assessed by dominant handgrip strength measured using the average of 3 trials of maximal isometric grip measured in kilograms with a Jamar dynamometer (Sammons Preston, Chicago, Illinois). Slowness was assessed using gait speed on a 15-ft walk (4.57 m). Participants were instructed to “walk at your comfortable pace” until a few steps past the 15-ft line. The timer was started with the first footfall after the 0-ft line and was stopped at the first footfall after the 15-ft line. The usual assist devices of subjects (e.g., walkers, canes) were permitted. If able, each subject completed one 15-ft walk. Gait speed was calculated by dividing 4.57 m by time to walk this distance in seconds and reported in meters per second, as has been previously recommended. Those subjects unable to walk 15 ft were considered to have a gait speed of 0 m/s. Instead of self-reported physical activity, independence in activities in daily living (ADLs) was assessed by the Katz ADL survey.
The 4 assessments were then combined into a frailty score, as previously described. Briefly, gait speed and serum albumin were divided into quartiles. Grip strength was divided into quartiles stratified by gender. As nearly 75% of subjects were independent in all 6 Katz ADLs, ADL status was dichotomized into a group with dependence in any ADL versus those with no ADL dependence. With these quartiles, a frailty score was calculated in the following manner: (1) quartiles of albumin, gait speed, and grip strength were assigned values of 0 to 3 in descending order and (2) a score of 0 for ADLs was assigned for ADL independence and 3 for any ADL dependence. These components were then summed to derive a frailty score for each subject (possible range 0 to 12), with the highest score representing the most frail and the lowest score being the least frail. Frail, for the purposes of this study, was then defined as a frailty score more than the median value in this population. A patient with a frailty score of ≥6 was considered frail, whereas those <6 were considered nonfrail.
The primary clinical outcome measure was the time to death from any cause for >1 year of follow-up. Other clinical outcomes of interest included 30-day cardiac death, repeat hospitalization because of AS or complications of the valve procedure, stroke, major bleeding, major vascular complications, permanent pacemaker, and renal failure requiring dialysis. Cardiac death, stroke, and major vascular complications were defined according to a modified version of the Valve Academic Research Consortium criteria as described in the PARTNER trial protocol. All events were adjudicated by an independent clinical events committee.
In addition, because recipients of TAVR are often elderly with multiple co-morbidities, it is likely that prolonged survival alone (without improved quality of life) would not be viewed as an acceptable outcome. To account for this need, a definition of poor outcome after TAVR that considers both survival and quality of life has been operationalized, which is defined as any of the following at 6 months after TAVR (definition #1): (1) death, (2) Kansas City Cardiomyopathy Questionnaire overall summary (KCCQ-OS) score <45, or (3) decrease of ≥10 points in the KCCQ-OS score from baseline to 6 months. In addition, we used an alternative, expanded definition of poor outcome (definition #2) that included any of the following at 1 year after TAVR: (1) death, (2) KCCQ-OS score <60, or (3) decrease of ≥10 points in the KCCQ-OS score from baseline to 1 year.
All statistical analyses were based on the population of patients who actually received TAVR and underwent a baseline frailty assessment at participating sites. Continuous variables are summarized as medians (interquartile ranges [IQRs]) and were compared between frail and nonfrail patients who underwent TAVR using the Wilcoxon rank-sum test. Categorical variables are presented as proportions and were compared by the chi-square test. Thirty-day event rates were compared between frail and nonfrail patients using univariable Cox proportional hazards models. Time-to-event variables were summarized by means of Kaplan-Meier estimates and compared with the log-rank test. Cox proportional hazards models were then used to evaluate the independent association between baseline frailty status and all-cause mortality for >1 year after TAVR. Logistic regression was used to evaluate the association between frailty status and poor outcome after TAVR using definitions 1 and 2, as described earlier.
The following models were used to evaluate the relation between frailty and 1-year mortality or poor outcome after TAVR. Using separate unadjusted Cox proportional hazards models, the relation between each component of the frailty score and mortality was evaluated. Gait speed, grip strength, and albumin were modeled as continuous variables and in quartiles. Because the results did not differ, only the results of the continuous analyses are shown. ADL was modeled as a dichotomous variable (independent in all ADLs vs dependent in any ADL). The unadjusted relation between frailty score, both modeled as a continuous variable as a categorical variable (dichotomized at the median), was evaluated. Unadjusted logistic regression was used to evaluate the relation between frailty status and poor outcome after TAVR using definitions 1 and 2, as described earlier.
Multivariable modeling was used to evaluate the independent relation between frailty status (frailty score dichotomized at the median) and 1-year mortality after TAVR. Frailty status was forced into the multivariable models. Covariates for the multivariable models were selected using stepwise Cox regression with entry/stay criteria of 0.1/0.1 and a maximal ratio of 1 covariate for every 10 events to avoid overfitting. Candidate variables included age, male gender, body mass index, transfemoral TAVR, Society of Thoracic Surgery score, diabetes mellitus, hypertension, angina pectoris, heart failure, New York Heart Association class IV, coronary artery disease, previous coronary angioplasty, previous coronary bypass, cerebrovascular disease, peripheral vascular disease, previous balloon aortic valvuloplasty, permanent pacemaker, renal disease, liver disease, chronic pulmonary disease, aortic valve mean gradient, ejection fraction, and moderate or severe mitral regurgitation. Using multivariable logistic regression, a similar modeling strategy was used to evaluate the independent relation between frailty and poor outcome after TAVR. A 2-sided alpha level of 0.05 was used for all significance testing. All statistical analyses were performed using SAS software, version 9.2 (SAS Institute, Cary, North Carolina).
Results
Of the 244 patients included in this analysis, overall, median gait speed was 0.38 m/s (IQR 0.23 to 0.64 m/s), median serum albumin was 3.9 g/dl (IQR 3.6 to 4.2 g/dl), and 172 (71%) performed all ADLs independently. In men, median grip strength was 23.6 kg (IQR 17.0 to 28.3), and in women, median grip strength was 12.2 kg (IQR 10.0 to 15.7; Table 1 ) The median frailty score was 5 (IQR 3 to 7). Accordingly, for the purposes of this study, 134 (120 cohort A and 14 cohort B) participants with a frailty score <6 were considered not frail and 110 (95 cohort A and 15 cohort B) participants with a frailty score ≥6 were considered frail. Baseline demographic, clinical, and echocardiographic characteristics stratified by baseline frailty status are listed in Table 2 . Notably, frail participants walked shorter distances when performing the 6-minute walk test and were less likely to be able to perform the 6-minute walk test.
| Non-frail | Frail | |
|---|---|---|
| Independent in ADLs | 134 (100%) | 38 (35%) |
| Albumin, g/dL | ||
| median (IQR) | 4.1 [3.8, 4.4] | 3.7 [3.4, 4.0] |
| Gait speed, m/s | ||
| median (IQR) | 0.5 [0.3, 0.8] | 0.3 [0.0, 0.5] |
| Grip strength, kg (men) | ||
| median (IQR) | 27.0 [21.3, 31.7] | 18.7 [13.2, 23.2] |
| Grip strength (women) | ||
| median (IQR) | 14.0 [10.9, 16.9] | 11.0 [9.0, 14.0] |
| Variable | Non-frail (n = 134) | Frail (n = 110) | p-value |
|---|---|---|---|
| Age (yrs) | 85.4 [79.4, 89.5] | 87.1 [82.7, 90.3] | 0.11 |
| Male gender | 74 (55%) | 52 (47%) | 0.22 |
| Body mass index (kg/m2) | 25.8 [22.2, 29.6] | 24.8 [21.9, 28.3] | 0.46 |
| Transfemoral TAVR | 62 (46%) | 57 (52%) | 0.39 |
| STS Score | 10.5 [8.8, 12.4] | 11.3 [9.6, 13.8] | 0.07 |
| Diabetes mellitus | 43 (32%) | 28 (26%) | 0.26 |
| Hypertension | 122 (91%) | 95 (86%) | 0.25 |
| Angina pectoris | 33 (25%) | 18 (16%) | 0.11 |
| Heart failure | 132 (99%) | 110 (100%) | 0.50 |
| NYHA Class IV | 36 (27%) | 32 (29%) | 0.70 |
| CAD | 114 (85%) | 91 (83%) | 0.62 |
| Previous PCI | 76 (57%) | 45 (41%) | 0.01 |
| Previous coronary bypass | 70 (52%) | 47 (43%) | 0.14 |
| Cerebrovascular disease | 39 (33%) | 26 (26%) | 0.30 |
| Peripheral vascular disease | 55 (41%) | 46 (42%) | 0.89 |
| Previous BAV | 33 (25%) | 35 (32%) | 0.21 |
| Permanent pacemaker | 30 (22%) | 27 (25%) | 0.69 |
| Renal disease | 19 (14%) | 15 (14%) | 0.90 |
| Liver disease | 3 (2%) | 9 (8%) | 0.03 |
| COPD | 57 (43%) | 46 (42%) | 0.91 |
| AV mean gradient (mm Hg) | 40.9 [36.5, 53.9] | 45.2 [34.9, 59.7] | 0.26 |
| AV area (cm2) | 0.63 [0.50, 0.83] | 0.62 [0.51, 0.72] | 0.20 |
| Ejection fraction (%) | 55 [45, 60] | 55 [35, 60] | 0.11 |
| Moderate or severe MR | 9 (7%) | 21 (20%) | 0.004 |
| 6-Minute Walk Test | |||
| Could Not Perform | 23 (17%) | 38 (35%) | 0.002 |
| Total Distance Walked (m) ∗ | 192 [122, 297] | 146 [77, 238] | 0.01 |
At 30 days, there were no differences in rates of major adverse clinical events according to baseline frailty status, including death, cardiac death, stroke, or repeat hospitalization ( Table 3 ). At 1 year, the Kaplan-Meier–estimated all-cause mortality rate was 32.7% in the frail group and 15.9% in the nonfrail group (log-rank p = 0.004, Figure 1 ). In univariable analyses, none of the individual components of the frailty score was associated with mortality over the 1 year after TAVR ( Table 4 ). However, frailty—both as a continuous score and as a categorical variable—was associated with increased mortality after TAVR ( Table 4 ). Frail patients had greater rates of both cardiac death (frail vs nonfrail: 15.0% vs 8.4%) and noncardiovascular death (frail vs nonfrail: 9.5% vs 4.9%) in patients for whom the cause of death could be classified. After adjusting for important clinical and demographic characteristics, frailty remained independently associated with a 2.5-fold increased hazard of 1-year mortality after TAVR (95% confidence interval [CI] 1.40 to 4.35, p = 0.002).
