What's new?

Risk prediction tools, such as the Global Registry of Acute Coronary Events (GRACE) 2.0 scale, have been developed in order to facilitate decision‑making in patients with myocardial infarction. However, regardless of the growing population of older patients, we still lack a risk assessment tool taking into account the geriatric syndromes, despite their known adverse effects on prognosis. Our study clearly shows that combining the assessment of elements of geriatric syndrome with routinely used GRACE scale increases its effectiveness in identifying the group at the highest risk of death and facilitates individualized and patient‑centered therapy of the elderly with acute coronary syndrome.

Introduction

Cardiovascular diseases (CVDs) are the leading cause of morbidity and mortality in older populations. Due to demographic trends and the atherosclerotic nature of ischemic heart disease, the proportion of seniors among patients with acute coronary syndrome (ACS) increases. According to the National Registry of Acute Coronary Syndromes PL‑ACS, the annual number of hospitalizations due to ACS is estimated at about 140 000, with 26.1% of patients aged above 75 years.¹ The average age of patients with nonpersistent ST‑segment elevation myocardial infarction (NSTEMI) in 2014 was 69 years.² According to the National Myocardial Infarction Database, in the group of patients with persistent ST‑segment elevation myocardial infarction (STEMI), the average age was 74 years for women and 63 years for men.³ Elderly patients are underrepresented in clinical trials and less frequently treated invasively, therefore data on the optimal management of older adults with ACS are limited, and, as a consequence, decisions regarding treatment of older patients in everyday clinical practice are highly individualized.⁴ Therefore, we have observed an increased interest in the problems related to older age, such as frailty syndrome (FS), cognitive impairment (CI), and multimorbidity, with regard to systematizing their impact on clinical decisions and patient risk stratification.

Frailty, a common syndrome in older patients with ACS (prevalence of 35.1% to 43.2%), is characterized by a reduction in reserves and resistance to stress factors, which results from the cumulative reduction in the efficiency of various physiological systems.^5-7 Literature indicates that frail patients with non–ST elevation ACS (NSTE‑ACS) are less frequently treated invasively, develop more complex coronary artery disease (CAD), are hospitalized longer, and are at a higher risk of death due to ACS.⁸ Multimorbidity, with the prevalence increasing with age, results in a more complex treatment process, predisposes to excessive drug consumption,⁹ and is associated with an increased risk of long‑term adverse cardiovascular events in older adults with NSTE‑ACS referred for coronary angiography.¹⁰ Multimorbidity predisposes to CI, common in older patients with ACS, which also has a proven adverse effect on the prognosis. The patients with CI are more likely to experience a major adverse cardiovascular event at 1 year,¹¹ and CI diagnosed with Short Portable Mental Status questionnaire influenced long‑term outcomes (8 years after discharge) with a continuous inverse association between CI and death or recurrent MI.¹²

Proper risk stratification is becoming an increasing challenge in clinical practice. The Global Registry of Acute Coronary Events (GRACE) 2.0 score is a widely‑recognized, guideline‑recommended standard risk stratification tool for patients with ACS (STEMI and NSTE‑ACS).⁴ According to the guidelines, in patients with GRACE score over 140, coronary angiography should be performed within 24 hours, as an early invasive intervention in this group is associated with a shorter hospital stay and a reduced risk of recurrent ischemia.⁴ In the GRACE population, age was 1 of the 2 strongest predictors of death due to ACS within 6 months.¹³ Older age is undoubtedly a factor negatively affecting prognosis and complications in patients with ACS. However, chronological age does not reflect the individual course of the aging process. The GRACE scale, despite being a very good and recommended risk stratification tool for patients with ACS, tends to overestimate the risk. On one hand, the presence of FS significantly worsens the prognosis (half of the patients with FS died in the first year after ACS), but on the other hand, in the same study older patients without FS were reclassified to a group with a lower risk of death.¹⁴ Therefore, not only the presence but also the lack of FS in an elderly person holds a prognostic value.

The aim of this study was to assess a potential advantage of adding the studied geriatric assessment elements (FS, CI, and multimorbidity) to the GRACE scale commonly used in predicting the risk of 6‑month death in patients with ACS.

Patients and methods

This was a single‑center, prospective study that included 196 patients hospitalized for ACS between 2013 and 2019 in the Cardiology Department of the Marciniak Lower Silesian Specialist Hospital in Wrocław (STEMI, n = 98; NSTE‑ACS, n = 98). The study group consisted of 119 men and 77 women, aged 65 to 97 years, at the mean (SD) age of 74.4 (8) years. In 178 patients (90.8%), an invasive strategy was undertaken, namely percutaneous coronary intervention (PCI) in 160 patients (81.6%) or coronary artery bypass grafting (CABG) in 18 patients (9.2%). The inclusion criteria comprised age of at least 65 years, clinical diagnosis of type 1 MI made according to the European Society of Cardiology criteria (third universal definition of MI),¹⁵ and written informed consent to participate in the study. The exclusion criteria were poor mental condition (deep dementia defined as the Mini‑Mental State Examination [MMSE] score below 11) and poor physical condition as assessed by a physician (vision or hearing impairment, impairment of the motor function of hands) preventing completion of the surveys. The study was approved by the Bioethics Committee of the Medical University in Wrocław (650/2018).

During hospitalization for MI (at least 48 h from admission, after a transfer from an intensive cardiac care unit to a cardiology department), the examinations and tests described below were performed by an experienced physician. CI was assessed using the Polish adaptation of the MMSE Brief Mental State Assessment Scale.^16,17 The maximal score is 30 points, the score of 27–30 is considered normal, of 24–26 indicates mild CI, of 19–23 mild dementia, of 11–18 moderate dementia, and the score below 11 reflects deep dementia (patients with deep dementia were excluded from the study). FS was assessed with the Tilburg Frailty Indicator (TFI) questionnaire, based on a broad concept of FS, taking into account the physical (8 points), mental (4 points), and social (3 points) domains.^18,19 The CAD‑specific index was used to assess multimorbidity,²⁰ and the risk of 6‑month mortality was assessed according to the GRACE scale. After 6 months, a follow‑up telephone call was performed to gain information about the late complications, such as reinfarction, coronary revascularization, stroke, or death.

Results

Discussion

Risk stratification in a growing population of elderly patients suffering from CVDs is becoming an increasing challenge in everyday clinical practice and a subject of intensive research.²² Proper risk stratification is of great importance, as it allows clinicians to identify the most vulnerable population and to improve its prognosis by planning an optimal model of care. The GRACE 2.0 score is a widely recognized, guideline‑recommended standard risk stratification tool for patients with ACS (STEMI and NSTE‑ACS).⁴ Despite the fact that age was 1 of the 2 strongest predictors of death due to ACS within 6 months in the GRACE population,¹³ there is an increasing awareness among clinicians and researchers that the process of aging is highly individual and chronological age is not the same as biological age. Therefore, increasing interest in geriatric problems, such as FS, CI, and multimorbidity, and their impact on the patient risk stratification has been recently witnessed. Literature data showed that patients with FS score higher on the GRACE scale.¹⁴ The results of 2 recent meta‑analyses indicate that FS in patients with MI is associated with a greater risk of mortality.^23,24 What is more, data from the SWEDEHEART registry²⁵ show that frailty in patients with MI is independently associated with 6‑month mortality, after adjustment for age, sex, and GRACE risk score components. In our work, the frailty group also achieved a higher mean score on the GRACE scale, and, as a consequence, presented a significantly higher estimated 6‑month risk of death than the nonfrailty group. Moreover, a positive correlation was observed between the severity of frailty and the estimated risk of death within 6 months. This raises a question whether the inclusion of geriatric assessment could improve the risk stratification in elderly patients with ACS. Some literature reports support this thesis. In the LONGEVO‑SCA registry,²⁶ a prospective, multicenter study including 532 patients aged 80 years or older hospitalized for NSTE‑ACS (with 27.3% prevalence of frailty), frailty assessed with the FRAIL scale was an independent predictor of all‑cause 6‑month mortality. Additionally, the multivariable predictive model including age, GRACE score, comorbidities, and frailty was significantly more effective in predicting death at 6 months than the GRACE score (AUC, 0.83; 95% CI, 0.77–0.89 vs AUC, 0.75; 95% CI, 0.68–0.82; P = 0.03). In the ICON 1 study²⁷ population (280 patients with NSTEMI aged ≥65 years, 27.5% assessed as frail according to the Fried criteria), the authors identified 6 prognostic factors for mortality in 12‑month follow‑up, and created a 6‑component scale named Fried‑Heart (Fried frailty, Killip–Kimball class, hypertension, age, peripheral artery disease, ability to dress independently). The Fried‑Heart score showed better predictive value than the GRACE score, which may indicate the benefits of adding the frailty status into risk stratification. Sanchis et al²⁸ showed that in 342 elderly patients (>65 years old) hospitalized for ACS, geriatric conditions such as frailty (assessed with the Green and Fried scores), CI, and comorbidities were associated with worse prognosis. The Green score of 5 or above was the strongest predictor of 30‑month mortality, and adding the Green score to the GRACE scale improved the prognostic value for mortality in 2.5‑year follow‑up (AUC GRACE, 0.726 vs AUC GRACE + Green ≥5, 0.776). Interestingly, in the study by Sanchis et al,²⁸ after adjusting for the Green score, the Fried score was not predictive of mortality. It is worth emphasizing that scales such as the Green score that include an effort testing element (walk speed, handgrip strength, stand‑up test) may be difficult to use during hospitalization for ACS.

Rittger et al²⁹ showed that in a group of 106 patients with NSTE‑ACS, the frailty index combining different parameters of functional impairment (Clinical Frailty Scale ≥7; MMSE ≤2; Instrumental Activities of Daily Living ≤7) significantly predicted conservative treatment (64% patients were treated invasively, and 36% conservatively). In this observational study, a decision to perform angiography was left to the discretion of the treating physician.²⁹

In most of the studies mentioned above, FS was defined by a phenotypic model (using the FRAIL scale, Fried scale, or Green score),^25-27 and focused only on the physical aspect of frailty. However, the TFI, used in our study, is a multidimensional questionnaire evaluating not only physical but also psychological and social aspects of frailty. It is a simple tool (assessing parameters only of the greatest prognostic significance), well‑validated, and easy to use and interpret in everyday clinical practice,^18,19 and it enables comprehensive patient assessment, including the biopsychosocial aspects.

In our study group of elderly patients with ACS, we investigated whether adding the scales used in geriatric assessment (TFI, MMSE, CAD‑specific index) to the GRACE 2.0 scale improves its prognostic power to predict death in 6‑month follow‑up. To enable comparison of the results, the parameters of the abovementioned scales were normalized, then ROC curves were determined and analyzed. We demonstrated that adding the assessment of FS (with the TFI questionnaire) to the GRACE 2.0 scale increases its prognostic value for 6‑month mortality, with the scale composed of GRACE (normalized) + TFI (normalized) having the highest predictive power (AUC = 0.737; for GRACE 2.0 score >124 points AUC was 0.713).

Our work is the first to document a benefit of adding the assessment of FS with the TFI to the GRACE 2.0 scale commonly used in ACS risk stratification in the population of elderly patients. However, as we studied a relatively small group, further research is needed to better consolidate the results.