Frailty and Exercise Training: How to Provide Best Care after Cardiac Surgery or Intervention for Elder Patients with Valvular Heart Disease

The aim of this literature review was to evaluate existing evidence on exercise-based cardiac rehabilitation (CR) as a treatment option for elderly frail patients with valvular heart disease (VHD). Pubmed database was searched for articles between 1980 and January 2018. From 2623 articles screened, 61 on frailty and VHD and 12 on exercise-based training for patients with VHD were included in the analysis. We studied and described frailty assessment in this patient population. Studies reporting results of exercise training in patients after surgical/interventional VHD treatment were analyzed regarding contents and outcomes. The tools for frailty assessment included fried phenotype frailty index and its modifications, multidimensional geriatric assessment, clinical frailty scale, 5-meter walking test, serum albumin levels, and Katz index of activities of daily living. Frailty assessment in CR settings should be based on functional, objective tests and should have similar components as tools for risk assessment (mobility, muscle mass and strength, independence in daily living, cognitive functions, nutrition, and anxiety and depression evaluation). Participating in comprehensive exercise-based CR could improve short- and long-term outcomes (better quality of life, physical and functional capacity) in frail VHD patients. Such CR program should be led by cardiologist, and its content should include (1) exercise training (endurance and strength training to improve muscle mass, strength, balance, and coordination), (2) nutrition counseling, (3) occupational therapy (to improve independency and cognitive function), (4) psychological counseling to ensure psychosocial health, and (5) social worker counseling (to improve independency). Comprehensive CR could help to prevent, restore, and reduce the severity of frailty as well as to improve outcomes for frail VHD patients after surgery or intervention.


Introduction
Worldwide, the population is aging, and the associated challenges for the healthcare system are significant. There is a clear association between degenerative valve disease, older age, and increasing life expectancy [1,2]. According to the Euro Heart Survey on Valvular Heart Disease (VHD), patients with diagnosed VHD are often older, with a higher prevalence of other cardiovascular risk factors and comorbidities [2].
Careful monitoring, adequate medication, and perfect surgery or intervention timing is key to a successful treatment of VHD. Due to the shift in patient population, the profile of patients who are referred to cardiac surgeon changed dramatically. The share of surgically treated elderly patients (≥80 years old) increased from 2.6-7% to 9.3-12% in a period of 10 years [3,4]. Innovation in cardiac surgery techniques [5], advance in anesthesiology, and early postsurgery care led to a lower risk of cardiac surgery in elderly patients [  evidence of improved outcomes with valve surgery or transcatheter aortic valve implantation (TAVI) [6,7]. However, older patients still have higher mortality and morbidity rates, more frequent complications related to surgery, and longer stay in hospital compared to younger patients [3,4,6]. Disease complexity in the elderly raises questions about other risk factors that exist upon well-known and established prognostic factors. Frailty is a common and relevant geriatric syndrome that could be defined as a biological syndrome with reduced reserve and resistance to stressors, resulting from cumulative deficits across multiple physiologic systems and causing vulnerability to adverse outcomes [8]. Most of the studies analyzed frailty before surgery/intervention as prognostic tool for later outcomes [9] and frailty is a reliable prognostic factor for mortality, morbidity, major complications [10][11][12][13][14][15][16], functional decline [17], quality of life [18], and risk of delirium after procedure [19,20]. Frailty status can be changed and its assessment should not only be for prognosis but also lead to "pre-rehabilitation" interventions [21] and clear comprehensive cardiac rehabilitation (CR) recommendations [9,22]. Physical rehabilitation for frail older people can positively affect physical fitness and this effect may be related to level of frailty in a setting of long-term care [23]. On the other hand, there are no clear recommendations concerning what methods and tools are sensitive enough and could be used to detect existence and dynamics of frailty syndrome in this particular patient population.
The aim of this literature review was to evaluate existing evidence on exercise-based cardiac rehabilitation (CR) as a treatment option for elderly frail VHD patients. This literature review aimed to assess the following issues: (1) Are assessment tools used for frailty screening sensitive enough to show improvement in frailty status after exercise-based CR?
(2) What exercise interventions would be the most beneficial for frail patients? (3) Would the change of frailty status improve outcomes on levels of disability, functional capacity, and quality of life after surgical/interventional VHD treatment beyond the effects of the surgery/intervention itself? frailty status strongly depended upon used assessment tools and variables measured [24]. Prevalence of frailty varied from 20-26% to 68-82%, according to the scale and the population examined [9,24]. A systematic review and meta-analysis by Anand et al. evaluated the relationship between preoperative frailty and outcomes following TAVI and demonstrated that the proportion of frail patients varied greatly across the different studies from 5 to 83% [25]. Although frailty has a good predictive value, results of the different studies are inconclusive and/or contradictory even with the same tools and similar populations. Also there is a gap of evidence concerning frailty tools sensitivity in this particular patient population. The most frequently used frailty assessment procedures for patients with VHD are discussed below.

Fried Phenotype Frailty Index Assessment (FFS). FFS
represents the definition of physical frailty and it is based on the results of the Cardiovascular Health Study and the Women's Health and Aging Studies [8]. It has been widely adopted and consists of five health domains: nutrition (unintentional weight loss), physical exhaustion (CES-D (depression) scale), low energy expenditure (or inactivity status) (Minnesota Leisure Time Activity questionnaire), mobility (5-meter walking test (5MWT)), and muscular strength (dominant hand handgrip strength) [8]. FFS covers all domains of physical frailty and is recommended by the American Geriatrics Society [26]. FFS scale also could be named as frailty index from Cardiovascular Health Study (5item CHS). However, the study that validated the FFS was performed with general population. There is no information about VHD patients, and the share of very old individuals included in the study was relatively low (3.6% of patients were 85 years or older) [8]. Moreover, used methods were subjective and based on qualitative manner and could not precisely evaluate cognitive function and level of independence. Enhanced FFS (or 7-item expanded CHS) with additional testing of cognitive function (using Mini Mental State Examination (MMSE)) and depressed mood was implemented [24,27]. Another example of FFS modification (modified fried frailty criteria) was scale for frailty evaluation that includes similar domains by using different methods to describe malnutrition (hypoalbuminemia) and clearly describe level of independence with Katz index [13,[28][29][30].
FFS [12,24,[31][32][33][34][35] and scale modifications are widely used in studies involving VHD patients for frailty screening and unfavorable intervention/surgery outcome prediction. FFS was an accurate frailty assessment tool for prediction of increased hospitalization costs after cardiac surgery [31]. Some authors did not find any relation between frailty (measured by FFS) and mortality, morbidity, or increased length of stay in the hospital [33,36,37]. However the studies concluded that frail patients are more often referred to cardiac rehabilitation facilities [33,36]. In a large study from Arnold et al. (N=2830) the authors demonstrated that frailty measured by FFS was a good predictor for negative shortterm outcome (mortality, low quality of life) [32]. Similar results were reported in other studies [34,35,38,39]. Forcillo et al. included high-and extreme-risk patients undergoing TAVI and found that serum albumin, Katz index, and 5MWT were associated with increased risk of adverse outcomes, but only albumin was predictive of 30-day all-cause mortality [28]. In contrast, in two studies from Afilalo et al. (2012 and) the FFS and its modifications were inferior compared to other assessment tools (5MWT alone and EFT) for frailty screening [12,24]. We failed to find any study that would be designed to evaluate FFS sensitivity for patients with VHD or in which FFS would be used in cardiac rehabilitation settings. According to our analysis, although FFS is a traditional method for frailty syndrome diagnosis and screening, it may be not reasonable for frailty dynamics after comprehensive CR program.   (2) Significantly more patients with delirium were considered as frail before TAVI. and were more likely to be discharged to an ICU facility (P =0.083 and P = 0.001).
(2) 4/4 frail patients had increased post-operative LOS (P = 0.014) (3) Individual components of the frailty score were also independent predictors of all-cause mortality.
(4) The HR of mortality increased with each increase in frailty score 8 BioMed Research International (2) For every 14-cm2/m2 increase in SMI, there was a 1-d reduction in LOS.
(3) None of the standard measures of frailty predicted LOS. (1) 30-d mortality and major postoperative morbidity. (1) Frailty was a better predictor of mortality than morbidity, and it was not markedly different among any of the 3 indices.
(2) Frailty was associated with an increased risk of 30 d mortality and longer LOS. (2) Total costs were independently associated with frailty and valve surgery (P < 0.001).
(3) This effect persists after adjusting for age, sex, surgery type, and surgical risk score. (1) association between baseline 6MWT and functional improvement (2) association between baseline 6MWT and mortality. (1) There were no differences in 30-d outcomes among 6MWTD groups.
(2) At 2 y, the rate of death from any cause was 42.5% in those unable to walk, 31.2% in slow walkers, and 28.8% in fast walkers (p = 0.02), driven primarily by differences in noncardiac death.
(3) Patients with poor baseline functional status exhibit the greatest improvement in 6MWTD. (2) Albumin levels <3.3 g/dl, falls in the past 6 months, predicted death at 1 y.
(2) By combining the FORECAST and the STS score, this effect was even higher (area 0.77; P = 0.021).
(3) Stratifying the patients according to the FORECAST score showed best survival in the lowest frailty group. BMI (1) all-cause mortality at 1 y of follow-up (1) All-cause mortality at 1 y was higher in the low-BMI group (log-rank p=0.003) with no significant difference among normal and above-normal BMI patients.
(2) In a multivariate model, BMI <20 kg/m2 was an independent predictor of mortality (HR=2. 45 (2) At follow-up, frail participants showed significant improvement in physical function, with physical health scores improving by 50% and 14%. Non-frail subjects did not significantly improve in these measures.
(3) Mental health scores also improved to a greater extent in frail participants (3.6 vs < (2) Frailty was an independent predictor of reduced midterm survival (HR 1.5, 95% CI 1.1 to 2.2). were significantly increased in frail patients during long-term follow-up, which did not change relevantly after adjustment (2) PMVR can be performed with equal efficacy and is associated with at least similar short-term functional improvement in frail patients.
(3) Kaplan-Meier estimated all-cause mortality to be significantly higher in the frail group (log-rank test; P = 0.042).
(4) A separate analysis by treatment group found that TPA predicted high resource utilization after SAVR (OR, 0.4; P<.001), but not after TAVI (P = .66)
(3) PMA is an independent predictor of mortality after TAVI; 5MWT combined with normalized PMA showed greater discrimination ability than alone. (2) The analysis found that the SPPB, the PARTNER frailty scale, the frailty index, CSHA and 5MWT were independently associated with unplanned readmission even after adjustment in the multivariate analysis. (2) Frailty increased the C -statistic to 0.71, 95 % CI 0.65-0.76.

Schoenenberger
(3) Frailty was strongly related to outcome and increased the discriminatory ability of the risk algorithm. (2) There was also a significant correlation between Frailty score and observed 30-day mortality ( p < 0.05).
(3) The comprehensive assessment of frailty is an additional tool to evaluate elderly patients adequately before cardiac surgical interventions. (2) 1 y mortality within each CAF subgroup shows a significantly higher mortality rate among patients in the "severely frail" compared to less frail groups.
(3) Higher frailty group had a LOS at the ICU.    [14,17,19]. It has all necessary elements for frailty assessment: mobility (TUG and subjective mobility evaluation), nutrition (MNA), cognitive functions (MMSE), and independence (ADL, IADL). In addition, Eichler et al. showed that this instrument could be used in CR settings, although there were no changes in Katz index and IADL results after 3-4 weeks of inpatient CR for patients after TAVI [40]. As a result of the comprehensive rehabilitation program, the proportion of frail patients was significantly reduced by 9% (from 36.9% to 27.9%). The overall frailty index decreased by 0.4 points, driven by the significant changes of the single parameters cognition (MMSE), nutrition (MNA), and subjective mobility disability and mobility (TUG). On the other hand, it was small (N=136), short (observation for 3-4 weeks), observational type study without control group or randomization.

Multidimensional Geriatric Assessment (MGA) or Modified Geriatric Baseline Examination (MGBE
However, other authors have not confirmed the predictive value of this instrument and found that it only partly predicted unfavorable outcome after TAVI (nutrition and mobility) [41]. More studies with VHD patients and MGA should be done in order to receive higher level evidence, but its possibility of evaluating most of complex frailty syndrome components looks promising not only in screening but also in frailty dynamics after CR.

Canadian Study of Health and Aging (CSHA) Scale or
Clinical Frailty Scale or Rockwood Scale. CSHA scale is a multidomain approach that defines frailty as a proportion of accumulated health deficits [9,26]. In the Canadian Study of Health and Aging, authors worked with 3 approaches: (1) developing a rules-based definition of frailty, (2) creating a method of counting a patient's clinical deficits, and (3) proposing the clinical frailty scale, a measure of frailty based on clinical judgment [42]. The seven-point scale is displayed in an easily visualized form including objective clinical judgment and subjective patient assessment [26]. It is a validated instrument for diagnosing physical frailty and easy to use [9,42,43]. Seiffert et al. used the Rockwood scale in the Bonn group (N=347/847 patients) undergoing TAVI and found frailty to be significantly related to 1-year mortality [26,44]. Other authors also used this scale and reported significant results related to quality of life, poor physical and mental function, physical well-being [45], and poor outcomes even in pre-frail patients [46]. Shimura et al. compared it with other well-known frailty markers and concluded that CSHA predicted late mortality in an elderly TAVR patient cohort better than 5MWT, body mass index (BMI), and handgrip strength [47]. We did not find any study that would use CSHA scale to evaluate frailty dynamics after CR for patients with VHD. What is more, CSHA did not show superiority compare to other frailty tools (such as 5MWT or handgrip test) in other reviewed studies [24,30]. On the other hand, CHSA scale is based upon patients level of disability, level of independence, that could be a link to other tools used in rehabilitation settings such as Barthel index (BI) or functional independence measure (FIM). Systematic review and meta-analysis performed by Ribeiro et al. [48] showed that frail patients after TAVI or surgical aortic valve replacement improved significantly according to BI or FIM results, compared before and after CR. What is more, recently published call of action regarding frailty in CR offered to use CSHA scale as possible frailty measure [9].

5-Meter Walking Test.
One of the requirements for frailty assessment (especially for screening) is simplicity and ability to assess frailty in a very short period of time. One of the most evaluated frailty assessments is the 5-meter (15 feet ∼ 4,5 meters) walking test which can be used as part of other scales (FFS) or as stand-alone evaluation [13,20,24,27,30,47,[49][50][51]. This test results could be evaluated in different manner: (1) measuring time taken for 5-meter walk in seconds and (2) measuring gait speed in meters per second. Different authors offer different cutoff values to describe frail patients that vary from 5 to 7 seconds according to gender and body composition [6,8]. Gait speed is an independent predictor of adverse outcomes after cardiac surgery, with each 0.1-m/s decrease conferring an 11% relative increase in mortality [52]. The slowest walkers (6-10 or more seconds) had 35% higher 30-day mortality than normal walkers (adjusted OR, 1.35; 95% CI, 1.01-1.80) [51]. Kano et al. described stepwise incremental increased risk of mortality after TAVI in patients with the slowest gait speed (<0.5 m/sec) or who were unable to walk compared to patients with a normal gait speed [50]. Studies that compared several frailty assessment tools also demonstrated 5MWT superiority among other more complex ways to evaluate frailty in VHD population [27,53]. This test is validated in large trials, has good predictive value for patients undergoing cardiac surgery, and is superior to other instruments [12,52]. Although it seems logical to use this 5MWT in cardiac rehabilitation, as it is recommended by AHA guidelines [54], none of the reviewed studies in CR setting chose this test. All of the mentioned studies regarding exercise training for patients with VHD evaluated six-minute walking test to show improvement in physical capacity, but this test failed to show predictive value in frailty screening before the procedure [55]. Future studies will be needed to evaluate 5MWT sensitivity in a setting of CR.

Serum Albumin Levels.
Pre-procedural albumin levels (as single measurement or as part of other indexes) have been demonstrated to have significant predictive value [47,[56][57][58]. Hypoalbuminemia (<3.5 g/dL) was associated with poor prognosis, highlighted by the increase in noncardiovascular mortality in elderly TAVI patients [58]. Grossman et al. showed that even higher pre-procedural albumin (< 4 g/dL) 24 BioMed Research International can significantly improve the ability of the Society of Thoracic Surgeons (STS) and EuroSCORE II scores to predict TAVRrelated mortality [57]. In addition Bogdan et al. showed that low post-procedural (4 days after procedure) serum albumin remained a strong parameter correlated with allcause mortality (HR=2.47; 95% CI: 1.284.78; p<0.01) [59]. Forcillo et al. analyzed whether MFFC predicted 30-day mortality after TAVI. The results demonstrated that although all MFFC components were associated with increased risk of adverse outcomes, only albumin (<3.5 g/dL) was predictive of 30-day all-cause mortality [28]. Eichler et al. focused on nutrition assessment but used another instrument, validated MNA scale [41]. Moreover, one of the components of the successful frailty risk score for the elderly undergoing aortic valve replacement (EFT) is low pre-procedural albumin level [24].
Albumin level seems to have higher prediction value and could be a good screening instrument for frailty. What is more, serum albumin levels can be linked to lower muscle mass in elderly [60]. Albumin as nutrition measure could be connected with low body mass index (BMI) and sarcopenia. Mok et al. showed that higher BMI in TAVI patients was associated with higher skeletal muscle mass [61]. However, 46% of these cases were sarcopenic obesity that was significantly higher than the 4% to 12% in the general population [61]. These results could give an impression that muscle mass is more important than whole body mass. Shimura et al. showed significant relation between lower BMI and higher level of frailty [47]. The results from Koifman et al. confirmed that BMI <20 kg/m2 should be considered as a frailty marker during the screening process of severe aortic stenosis TAVI patients as it is associated with higher mortality [62]. Albumin levels should be taken in to account during comprehensive CR program, which is the role of dietician counseling, helping to reach healthy body mass not only through healthy diet but also through increasing muscle mass.

Katz Index of Activities of Daily
Living. The Katz index is used for functional assessment of dependency in elderly individuals in the six functions: feeding, bathing, dressing, transferring, toileting, and urinary continence [95]. It provides objective data on a patient's independency in daily activities [26]. Frailty usually overlaps with disability; hence some authors use the same instruments for disability measurement while others use it for frailty assessment [32,69]. Earlier studies incorporated Katz index in the frailty assessment and demonstrated that it is an independent risk factor for in-hospital mortality and institutional discharge in patients after cardiac surgery [15,26,96]. Therefore, a more pragmatic approach in defining frailty, commonly called the social domain, often includes an ADL assessment [26]. Puls et al. identified a Katz index <6 as a powerful independent predictor of immediate-, short-, and long-term all-cause mortality in TAVI patients [26,77]. Most of the review studies included Katz index as part of frailty assessment and demonstrated it to produce good prediction values [17,49,68,69,72]. Recently published American College of Cardiology/American Heart Association (ACC/AHA) and European Society of Cardiology (ESC) guidelines recommend the Katz index as one of the frailty assessment instruments for risk prediction before cardiac surgery [97,98]. On the other hand, other authors did not detect Katz index as good predictor for long-and short-term outcomes; in those studies it was inferior to mobility and nutrition measures [30,41,64]. Although usage of Katz index looks promising, we failed to find any study in CR with VHD patients that would test individuals with KI before and after CR program. Eichler et al. used IADL and basic DAL [40]; other authors successfully used BI or FIM to evaluate frailty, disability, or patient independence [85,99].

What Exercise Interventions Would Be the Most Beneficial for Frail Patients?
In order to answer the second research question, we analyzed 12 studies published in English since 1980 on topic of VHD and exercise training with elderly ( Table 2). All analyzed studies demonstrated benefit of exercise training in patients after surgical/interventional VHD treatment. Although the evaluated exercise-based CR programs were differently structured, the patient population included and the primary endpoints examined were similar and focused on the results of 6-minute walking test and cardiopulmonary stress test (peak VO 2 ). 6 studies were conducted in inpatient facilities and 6 studies analyzed the results of outpatient programs only. In 8 studies the program started early (< 1 month) after surgery or intervention. Other studies included patients after ≥1 month of surgery or intervention. In all reviewed publications specified documented adverse events were not related to exercise [91,93,100].
Although recently published studies concentrate on disabled, comorbid high-risk elderly patients, all authors demonstrated that exercise training is safe and effective in improving functional and physical capacity [40,84,85,87,89,93], quality of life [40,84], and independence in daily life activities [48,84,85,87,93]. Exercise training type, method, length, and intensity differed among the studies and did not allow clear recommendations concerning how exactly this patient group should be treated.
Only two randomized controlled trials (RCT) were performed with patients aged 60 years or older [91,92]. Low sample size pilot study on exercise training after TAVI, with combined endurance and resistance training, showed significant improvements in exercise capacity, muscular strength in different muscles groups, and quality of life compared with usual care [91]. In this study, 13 TAVI patients participated in an eight-week outpatient CR, starting 2-3 months after intervention. Resistance training was added in the second week of the program and performed in 2 of the 3 weekly workouts. It included 5 different exercises starting with 1 set with 10 repetitions at 30% 1 repetition maximum (1-RM) and was gradually increased to 3 sets with 15 repetitions at 50% to 60% 1-RM. Sbilitiz et al. showed that CR group compared with control had a beneficial effect on VO2 peak at 4 months (24.8 mL/kg/min versus 22.5 mL/kg/min, p=0.045) but mental health and other measures of exercise capacity and self-reported outcomes was not affected [92]. Exercise training in this study was initiated 1 month after surgery, comprising 3 weekly exercise sessions for 12 weeks. The (3) Significant decrease in the VE/VCO2 in the medium-and high-risk patient groups (p <0.05).
(1) Exercise capacity after VHD surgery is related to the preoperative risk and to the type of surgery.
(2) Similar benefit from exercise training can be obtained, independent of the preoperative risk class or the type of surgery. (3) ET should be offered to all patients after valvular surgery, regardless of their EuroSCORE risk or type of surgery    program consisted of graduated cardiovascular training and strength exercises. None of these studies assessed level of patients frailty before and after exercise training program; on the other hand, it did show that structured training program could improve older patient physical capacity [91,92]. What is more, Sbilitiz et al. did not find improvement in patients' mental health which could be seen as a one component of complex frailty syndrome. It is interesting that latter study had special structured psychosocial education that cover relevant topics: changed body-image and self-image, recovery from major surgery, and dependency on relatives and medical issues, but did not receive positive results [92].
Only one study, conducted by Eichler et al., evaluated the effect of a multicomponent inpatient CR on frailty in patients after TAVI [40]. It is important to mention that the CR included patient education, diet counseling, psychological support, and risk factor management as well as individualized training components. Exercise training performed was mainly aerobic such as cycle-ergometer training and Nordic walking. Patients with a higher exercise capacity (> 1.0 W/kg) conducted additional strength training at 30-50% 1-RM. For frailty assessment, frailty index by Schoenberg et al. was used. As a result of the rehabilitation the proportion of frail patients was significantly reduced by 9% (from 36.9% to 27.9%). The overall frailty index decreased by 0.4 points, driven by the significant changes of the single parameters cognition (MMSE), nutrition (MNA), and subjective mobility disability and mobility (TUG). Although frailty index could be reversed during cardiac rehabilitation, it did not show a prognostic impact itself [40]. On the other hand, it was observational type study, did not have control group, and was not designed as randomized control trial.
Russo et al. showed that early cardiac rehabilitation enhances independence evaluated with Barthel index, mobility, and functional capacity for patients after surgical or interventional aortic stenosis treatment. The exercise training consisted of aerobic training and calisthenics [85]. Systematic review and meta-analysis performed by Ribeiro et al. [48] showed that patients after TAVI or surgical aortic valve replacement achieve similar benefits from cardiac rehabilitation. Frailty components analyzed in this review were independently measured by Barthel index or functional independence measure (FIM), which improved significantly in both groups. The cardiac rehabilitation programs included exercise endurance training (cycle or treadmill), resistance training, and/or inspiratory muscle training.
Summarizing of existing evidence on exercise training for patients with VHD would show that exercise program should be individualized and intensity should be measured by RPE with beginning of 40-60% of maximal VO2. All analyzed training protocols included aerobic training component implemented through various activities (Nordic walking, bicycle ergometer, gymnastics, etc.) and in 9 of them additional strength training was performed starting form the second or third week of exercise training, measuring intensity with 1 repetition maximum method. Inpatient CR program length varied from 2 to 4 weeks, while the duration of the outpatient programs was 12-24 weeks.

Would the Change of Frailty Status Improve Outcomes on Levels of Disability, Functional Capacity, and Quality of Life after Surgical/Interventional VHD Treatment beyond the Effects of the Surgery/Intervention Itself?
According to recent studies, in 23-32,8% of cases functional capacity of TAVI patients did not improve or even deteriorate in a period of 6-12 months after intervention [101][102][103][104]. If we put aside certain procedure technique, surgeon experience, risk-benefit ratio evaluation, and careful patient selection, individualized care or CR in inpatient/outpatient setting could outweigh unwelcome complications. Older VHD patients constitute a sensitive group that requires more specific comprehensive cardiac rehabilitation programs and they return to casual life could be more difficult [4,105].
After analyzing the reviewed studies, we could not directly answer to the third research question. Two studies that were designed as RCT did not evaluated frailty or its components [91,92], and study that was designed to evaluate exercise training impact on frailty did not have control group [40]. On the other hand, all reviewed studies show positive impact of CR and did not register any adverse event or clinical deterioration linked to exercise training program.
Participation in CR improves outcomes of patients with cardiovascular disease [106] and may be specifically beneficial for patients with frailty syndrome; however, cardiac rehabilitation facility remains unused.

Discussion
The results of this review highlight the importance of frailty syndrome in patients with VHD. Screening for frailty as a high-risk factor is well implemented in Heart team decision making for high-risk patients with aortic stenosis. However, there is gap of evidence considering frailty in CR settings. This literature review was designed to seek for answers of several research questions: Are assessment tools used for frailty screening sensitive enough to show improvement in frailty status after exercise-based CR? What exercise interventions would be the most beneficial for frail patients? Would the change of frailty status improve outcomes on levels of disability, functional capacity, and quality of life after surgical/interventional VHD treatment beyond the effects of the surgery/intervention itself?
First of all, our analysis revealed several different methods used for frailty assessment and evaluation. The analysis of these frailty tools was difficult because (1) a huge variety of instruments were used, (2) the same instruments were chosen to evaluate disability or cognitive functions in some studies and frailty in others, (3) the same instruments were named differently in separate studies, (4) authors were using validated scales but chose to analyze components separately, (5) authors created and offered new risk scores for frailty assessment that has not been implemented in other studies, and (6) some studies described subjective frailty evaluation manner "eye-ball test" [78,79], while others used objective, quantitative or semiquantitative tests to evaluate all possible deficits.
It is clear that the prevalence of frailty plays an important role for patients with VHD: recently updated AHA VHD and ESC guidelines emphasize the frailty issue as an important factor for Heart team decision making in patients with aortic stenosis [97,107]. In 2014, the AHA guidelines recommended the use of Katz Index of independence of daily living and the 5-meter gait speed test, and this recommendation was left in 2017 [54,107]. Frailty as worse outcome predictor is already mentioned in ESC guidelines in 2012, but still without clarification of how exactly frailty should be measured and evaluated [108]. ESC 2016 guidelines indicate that frailty assessment should not be subjective and criticize the "eyeball test" [97]. According to cited studies frailty could be measured with complex evaluation of the patient that includes Katz index of independence of daily living, cognitive function evaluation, nutrition assessment, and patient functional capacity (TUG) [97].
Existing literature confirms frailty predictive value for higher complications rate, longer length of hospital stay, and mortality and morbidity after cardiac surgery or interventional treatment [9,97]. On the other hand, most of the studies describe frailty assessment before procedure. Frailty status is dynamic and can transit over time. In most cases people become more frail while aging, but in up to 20% of the population frailty criteria also can disappear over time [109,110]. Study that analyzed community living elderly persons showed that, of the 754 participants, 434 (57.6%) had at least 1 transition between any 2 of the 3 frailty states during 54 months. Transitions to states of greater frailty were more common (rates up to 43.3%) than transitions to states of lesser frailty (rates up to 23.0%) [109,110]. A small study that was performed by Forman et al. showed significant worsening of frailty level (measured by CSHA, 5MWT, and MMMSE) and functional capacity in 10 weeks for patients that are waiting for TAVI [111]. This dynamics shows that even short-term (3-4 weeks) CR could be effective for these patients.
Vigorito et al. recently published a call for action on frailty in CR and offer mainly two instruments for frailty assessment in this specific field: Edmonton frailty scale (EFS) and clinical frailty scale from the CSHA study. The EFS could be used as a comprehensive instrument that has been validated in elderly patients after acute coronary syndromes [9]. It also includes two clinical performance items interrogating cognition (clock test) and functional performance (TUG) [9]. However, some EFS questions are developed for screening more than for effect evaluation (e.g., hospitalizations per year, functional independence, social support, and medication usage) and would not be sensitive to show results of comprehensive CR. Also we did not find any study with VHD patients where frailty would be measured with EFS. Clinical frailty scale from the CSHA is popular and looks effective for screening in this patient population [20,35,44]. It is easy-to-use standardized tool based on subjective evaluation of frailty; however, it lacks psychosocial factors [9,26]. Moreover, studies that compare different frailty tools (including CSHA frailty scale) did not show its superiority to other instruments for frailty screening and outcomes prediction [24,27,30].
Our analysis revealed the six most used tools for frailty in VHD population: 3 scales (fried frailty scale and its modifications, multidimensional geriatric assessment, and clinical frailty scale from the CSHA study) and 3 single measures (5-meter walking test, Katz index, and serum albumin levels). Together with existing guidelines and expert opinions we believe that simple and fast-performed single test would not show comprehensive CR effect on a frail patient. An instrument of frailty assessment in CR should have the various components included in different risk assessment scores (mobility, independence in daily living, cognitive functions, nutrition, muscle mass and strength, and anxiety and depression evaluation). We did not find enough studies performed in CR with frail patients to find out which of all the mentioned tools is the best and most sensitive for frailty syndrome and its dynamics after CR program. Future studies are needed to build new evidence in this field of research.
Second research question aim was to find an ideal exercise training-based CR program for frail patient with VHD. It is known that exercise-based comprehensive CR is based on strong evidence and is a well-recognized treatment for patients with different cardiovascular diseases [112,113]. Physical rehabilitation for frail older people can positively affect functional and physical capacity and this effect may be related to the level of frailty [23]. Physical activity recommendations and supervised exercise programs can be useful for frailty prevention, to improve physical and functional capacity (gait speed, TUG, and SPPB), survival, and quality of life as well as to decrease prevalence of anxiety and depression and risk of falls in community-dwelling frail older people [114][115][116][117]. According to recent published systematic review, interventions for pre-frail and frail older adults should include multicomponent exercises, including in particular resistance training, as well as aerobic, balance, and flexibility tasks [117]. A study that included older patients soon after CABG demonstrated that additional resistance training and special balance training improved functional capacity (6-MWD, TUG time) significantly more than aerobic training alone [118]. Exercise training-based cardiac rehabilitation could be a treatment option for frail VHD patients. EAPC Cardiac Rehabilitation Section raised awareness and importance of frailty topic in the field of CR [9].
However, literature data in exercise training and CR in VHD patients is scare [100]. According to a published Cochrane meta-analysis, exercise training is beneficial and provides short-term improvement; however, it was assessed by only two studies included in this meta-analysis, and both of them were performed in a young-middle aged sample [100]. Our literature review showed that different exercise training programs with VHD elderly patients after surgery/intervention were safe and effective in improving functional and physical capacity [40,84,85,87,89,93], quality of life [40,84], and independence in daily life activities [48,84,85,87,93]. Studies with frail patients demonstrated that exercise-based programs increased gait speed and improved balance and performance in activities of daily life or SPPB [9]. A systematic review on exercise-based CR for patients after aortic stenosis surgical/interventional treatment demonstrated that a short-term (2-6 weeks) program could improve Barthel index, FIM, 6MWT, and anxiety and depression [48]. Eichler et al. demonstrated that participating in comprehensive CR results in frailty index decrease by 0.4 points, driven by the significant changes of single parameters such as cognition (MMSE), nutrition (MNA), and subjective mobility disability and mobility (TUG). In addition, several studies showed improvement in frailty and physical and functional capacity in frail patients, although baseline evaluation before surgery/intervention demonstrated lower results in the frail patient group [18].
All reviewed studies had an aerobic component of exercise training that was fulfilled in different ways, and 9 of 12 reviewed studies had an additional strength-training component. All described exercise programs were individualized according to the patient capability and described careful exercise intensity dosage according to rate of perceived exertion. We believe that ideal CR for VHD patients should be supervised by a cardiologist, and its content should include (1) exercise training (endurance and strength training to improve muscle mass, strength, balance, and coordination and to avoid falls), (2) nutrition counseling (to improve healthy body mass), (3) occupational therapy (to improve independency and cognitive function), (4) psychological counseling to ensure psychosocial health, and (5) social worker counseling (to improve independency). To improve outcomes of these patients CR services need to be optimally utilized and the protocols modified to cater for frail patients and to monitor their progress over the course of the treatment [106].

Limitations
Our literature review has several limitations and weaknesses. First of all, it was not designed as a meta-analysis or systematic review, so results of our study have lower evidence level. We searched only one database and used informal and subjective methods for studies inclusion and quality check. A quality assessment of the included studies would have strengthened our study.
What is more, we failed to find concrete and practical answers for research questions that were raised in the beginning of the study. We could only provide existing situation and try to find possible goals for future studies.

Conclusions
Frailty assessment in CR settings should be based on functional, objective tests and should have similar components as tools for risk assessment (mobility, muscle mass and strength, independence in daily living, cognitive functions, nutrition, and anxiety and depression evaluation). Participating in comprehensive exercise-based CR could improve short-and long-term outcomes in frail VHD patient. Comprehensive CR not only should include exercise training, psychological interventions, and improvement in nutrition but also could prevent, restore, and reduce the severity of frailty as well as improving outcomes for frail VHD patients.