Author + information
- Michael J. Mack, MD∗ ( and )
- Elizabeth M. Holper, MD, MPH
- ↵∗Reprint requests and correspondence:
Dr. Michael J. Mack, The Heart Hospital Baylor Plano, 1100 Allied Drive, Plano, Texas 75093.
Accurate risk assessment for any medical procedure is a critical component of patient informed consent. Initial clinical assessment methods involved the “eyeball test” or the “end of bed-o-gram” to predict the likelihood that a patient would survive a planned procedure without major complications. Subsequently, to standardize the subjective nature of the eyeball test, The Society of Thoracic Surgeons (STS) analyzed the outcomes of thousands of cardiac operations to develop a risk algorithm. The STS Predicted Risk of Mortality (PROM) predicts 30-day mortality and major morbidity rates after the most common cardiac operations; subsequent studies also showed a correlation with 1-year mortality rates (1). The STS PROM and other risk algorithms, including the European System for Cardiac Operative Risk Evaluation (EuroSCORE), logistic EuroSCORE, and EuroSCORE II, which have been developed and validated in surgical populations, have been used to assess risk in patients considered for transcatheter aortic valve replacement (TAVR). Despite the obvious invalidity of using risk algorithms to assess candidacy for a procedure for which these algorithms were not developed or validated, no alternatives have been available until recently.
The analysis of patients’ limitations on functional capacity, or frailty, has become an increasingly important consideration for clinical decision making. None of the risk algorithms mentioned earlier includes frailty as a consideration. Frailty has been described as a “biological syndrome” marked by declines across multiple physiological systems that manifests as deterioration in homeostatic reserves and resilience (2). Although frailty is associated with age, disability, and comorbidities, there is general agreement that frailty is a distinct entity (3,4). Although no consensus operational definition has been established, frailty is nevertheless estimated to coexist in 10% to 60% of older patients with cardiovascular disease (5). The 5-m walk test of gait speed is the most commonly used metric in determining the frailty of patients with cardiovascular disease (6,7).
Green et al. (8) evaluated frailty in older patients who were undergoing TAVR for symptomatic aortic stenosis. A composite frailty score was created including gait speed, grip strength, serum albumin value, and activities of daily living, and a high score was associated with a 3-fold increase in 1-year mortality rates. One study evaluated specific risk scoring systems in TAVR, with geriatric testing included in the evaluation. However, in this Italian OBSERVANT (Observational Multicenter Registry) score, geriatric testing was not found to be significant and thus was not included in the final modeling (9). The TVT (Transcatheter Valve Therapy) Registry developed a relatively parsimonious risk algorithm constructed on the basis of patients with TAVR, but frailty is not included as a variable in the algorithm (10).
In this issue of the Journal, Hermiller et al. (11) describe a TAVR-specific risk score for 30-day and 1-year mortality rates in the extreme-risk and high-risk patients enrolled in the CoreValve U.S. Pivotal Trial and continued access registry that was sponsored by Medtronic (Minneapolis, Minnesota). This study analyzed 2,482 patients as a derivation cohort and 1,205 patients in a validation cohort. Significant multivariable predictors of 30-day and 1-year mortality rates were used to create a risk score. What is unique about this risk prediction algorithm is that in addition to the usual variables predictive of death, assessments of frailty and disability were also used. Frailty parameters assessed included gait speed, grip strength, body mass index, anemia, hypoalbuminemia, unplanned weight loss, and impaired mobility on the basis on recent falls or being wheelchair bound. Parameters of disability were the Katz Index of Independence in Activities of Daily Living and the mini-mental state examination for dementia.
In these groups of extreme-risk and high-risk patients with aortic stenosis, the mean age was 83.3 ± 7.8 years, 46.3% were women, and the mean STS PTOM was 8.9 ± 4.8%. Despite the presence of multiple comorbidities, frailty, and disability, the overall mortality rate was 5.8% at 30 days and 22.8% at 1 year. Significant predictors of 30-day death included age >85 years, home oxygen use, residence in an assisted living facility, and albumin <3.3 g/dl; with hazard ratios ranging from 1.46 to 1.74. A scoring system of risk prediction was created to stratify patients into low-, moderate-, and high-risk subsets; a 3-fold higher risk of 30-day death was observed in the high-risk subset compared with the low-risk subset (10.9% vs. 3.6%). Significant multivariable predictors of 1-year death included home oxygen use, falls in the past 6 months, a high Charlson comorbidity index score, STS PROM >7%, and albumin <3.3 g/dl, with hazard ratios ranging from 1.27 to 1.90. The validation cohort showed again a 3-fold higher mortality rate at 1 year in the high-risk subset (36.6%) as compared with the low-risk subset (12.3%).
On performing this new and detailed clinical risk assessment in the CoreValve trial, the study participants and Hermiller et al. (11) should be commended on the comprehensive assessment before the procedure to include comorbidities, traditional risk scores, and frailty and disability assessment tools. It is encouraging that the significant predictors of 30-day death included answers to 3 simple questions that can be asked on initial evaluation (age, home oxygen use, and assisted living facility residency), as well as serum albumin level. The ability to stratify such patients into high-risk and low-risk groups on the basis of the simple scoring system allows greater clinical utility. However, the predictors involved in the assessment of 1-year mortality rates are more diverse and more time consuming to obtain. Although home oxygen use and low serum albumin levels again are predictors, the more traditional risk scoring systems emerge, with an STS PROM >7% and a high Charlson comorbidity score being most predictive. This finding highlights the differences in predicting short-term versus long-term risk in these patient; short-term mortality rates are related to age and ability to function and live independently, and thus survive the insult of an invasive procedure, whereas long-term mortality rates are associated with existing comorbidities.
The decreasing 30-day and 1-year mortality rates in patients undergoing TAVR present an opportunity to focus on mitigating risk in the higher-risk groups and optimizing outcomes in the lower-risk groups. It is clear that a single unified scale or scoring system cannot adequately predict both operative risk and long-term risk in an older patient with a complex clinical situation and aortic valve stenosis. However, the paper by Hermiller et al. (11) highlights the need to incorporate current risk scores such as the STS PROM with frailty to predict short-term and long-term risk of death adequately. Given the time commitment required to complete the comorbidity calculation, frailty testing, and disability scoring in the CoreValve trial, clearly clinicians will need a unified, efficient means of evaluation.
Additionally, novel metrics of frailty associated with increased operative risk, such as sarcopenia measurable by computed tomography (12), will need to be considered in the armamentarium of risk stratification. Increased focus on this area of risk prediction before TAVR reveals that there is much more that meets the eye than just the eyeball test. By incorporating additional objective measurements of patient risk, the eyeball test can be improved, and better patient-centered decision making can result.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the author and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Mack serves as the uncompensated co-principal investigator of the PARTNER-3 trial for Edwards Lifesciences; is the principal investigator of a national trial sponsored by Abbott Vascular; and serves as the uncompensated co-principal investigator of the COAPT trial of Abbott Vascular. Dr. Holper serves on the medical advisory board for Boston Scientific.
- American College of Cardiology Foundation
- Fried L.P.,
- Tangen C.M.,
- Walston J.,
- et al.
- Afilalo J.,
- Alexander K.P.,
- Mack M.J.,
- et al.
- Rodriguez-Manas L.,
- Feart C.,
- Mann G.,
- et al.
- Afilalo J.,
- Eisenberg M.J.,
- Morin J.F.,
- et al.
- Hermiller J.B. Jr..,
- Yakubov S.J.,
- Reardon M.J.,
- et al.