1. INTRODUCTION

Acute myocardial infarction (AMI) is a major cause of morbidity and mortality around the world. AMI causes fatal conditions such as heart failure and sudden cardiac death, and it has become the leading contributor to disease burden.1 In Taiwan, despite the overall age- and gender-adjusted AMI incidence maintaining at approximately 50 per 100 000 people in the past decade, a noteworthy rise of 30.3% in young males and 24.4% in young females under 55 has been observed.2 Survivors of AMI confront an elevated risk of major cardiovascular events (MACE) following discharge. MACEs typically includes heart failure, nonfatal reinfarction, rehospitalization for cardiovascular-related concerns, repeat percutaneous coronary intervention (PCI), coronary artery bypass grafting, and all-cause mortality.3 The incidence of MACE varies from 4.2% to 51%, depending on the definition and duration of follow-up.3 Therefore, implementing lifestyle changes, ensuring quality care, and administering guideline-directed medical therapy are pivotal for secondary prevention of MACE.

The European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery advocate a postrevascularization imaging examination 6 months later for high-risk patients.4 However, echocardiography and coronary computed tomography offer limited long-term prognostic value. Stress echocardiography requires extensive training and adherence to high interoperator consistency standards.5 In contrast, cardiopulmonary exercise testing (CPET) excels in providing comprehensive insights into cardiovascular, pulmonary, and musculoskeletal integration, offering valuable assessments of functional aerobic capacity in various populations, both healthy and diseased, making it a gold standard for prognostic stratification.6,7

Among all CPET indices, maximal oxygen consumption (max VO2) is the most consistent with predicted values in patients with various diseases.8 However, challenges such as peripheral muscle fatigue, dyspnea, and significant cardiac alterations impede cardiac patients from achieving maximal exercise effort.9 Consequently, submaximal CPET parameters prove more suitable for assessing patients with cardiovascular disorders. Indices such as anaerobic threshold (AT), the slope of the relationship between minute ventilation and carbon dioxide production (VE/VCO2 slope), work efficiency, and oxygen uptake efficiency slope (OUES) exhibit strong correlations with cardiac function.10,11 Furthermore, the OUES, which estimates ventilator efficiency in relation to oxygen consumption,12 is now a well-established substitute for max VO2 in submaximal exercise effort in adults,13 older children,12 and AMI survivors.14 The OUES has a high correlation with peak VO2, good test-retest reliability, and relatively stable during the incremental exercise test.15

Per the American College of Cardiology/American Heart Association guidelines,16 AMI patients should undergo predischarge CPET for risk stratification and subsequent cardiac rehabilitation prescription. However, due to safety concerns, limited resources, and inadequate equipment, few hospitals, including tertiary ones, conduct predischarge CPET for AMI patients.17 Additionally, the prognostic significance of OUES in the predischarge status of AMI patients remains unknown. This study aims to investigate the short and long-term prognostic value of predischarge CPET variables, particularly OUES, providing valuable insights for physicians in managing post-AMI patients.

2. METHODS 2.1. Study design and participants

This follow-up study enrolled patients admitted for their first AMI at a tertiary center in southern Taiwan from December 2012 to November 2017. Patients were eligible if they were (1) aged 20 or older, (2) first diagnosed with AMI, (3) received primary PCI (PPCI), and (4) underwent CPET before discharge, (5) had completed record of transthoracic echocardiographic examination and standard 12-lead electrocardiogram (ECG). Patients were excluded from the study if they had a history of acute coronary syndrome, or if they were deemed too frail for CPET or cardiac rehabilitation. This frailty criterion included those with cognitive impairment, neuromuscular disorders, ventilator dependence, severe pulmonary disorders requiring oxygen, and those who had been bedridden for more than 3 months. Additionally, patients with missing data or incomplete CPET records, as well as those who did not have regular medical follow-ups for at least 5 years following the onset of their AMI, were also excluded from the study. Cardiologists provided the PPCI and medications for the patients. Cardiologists would refer patients to physiatrists for phase I cardiac rehabilitation if there were no immediate complications following treatments. The phase I cardiac rehabilitation protocol was modified from the American College of Sports Medicine (ACSM) guidelines18 and has been used as a standard operating procedure in this medical center with evidence to improve exercise capacity in patients after AMI.19 Physiotherapists with at least 3 years of cardiopulmonary rehabilitation experience performed phase I cardiac rehabilitation for AMI patients. The study was approved by the Kaohsiung Veterans General Hospital’s Institutional Review Board (VGHKS17-CT11-11) and we obtained informed consent from participants after explaining the CPET purpose and study objectives.

We retrospectively retrieved demographic, clinical, and angiographic data from patients’ medical records. Clinical data encompassed medical history, medications, smoking history, and body mass index (BMI). Follow-up extended from PPCI to the first occurrence of a MACE or administrative censoring. Continuous medical care was provided by the medical center’s Department of Cardiology outpatient clinic. MACEs, defined as repeat coronary revascularization, recurrent myocardial infarction, cerebral vascular accident, or cardiovascular death, were assessed at 3 months, 2 years, and 5 years post-AMI intervention. Cardiologists confirmed MACEs through medical records. Patients with MACE constituted the adverse event group, while those without MACE at each follow-up (within 3 months, 2 years, and 5 years) comprised the adverse event-free group.

2.2. Cardiopulmonary exercise testing

A symptom-limited, progressive exercise test was performed on each patient, which included leg ergometer, a flow module, a gas analyzer, and an ECG monitor (Metamax 3B, Cortex Biophysik GmbH Co., Germany). The exercise testing was carried out with an incremental workload of 10 W/min.18 We stopped the test when the patients experienced subjectively unbearable symptoms (severe shortness of breath, chest pain, severe dizziness, excessive fatigue, physical instability, and excessive pallor as indicated by ACSM18), were unable to continue, or reached the submaximal endpoint, which was defined as work ≥75 W/min, peak oxygen consumption ≥5 metabolic equivalents (MET), peak heart rate ≥70% of the age-predicted value, or respiratory exchange rate ≥1.1. All patients underwent CPET under the supervision of a physiatrist with more than 10 years of experience (K.-L. L.).

During the CPET, we measured oxygen consumption (VO2) and carbon dioxide production (VCO2) on a breath-by-breath basis. Furthermore, the AT, respiratory rate, and several derived variables such as respiratory exchange ratio (RER) and VE/VCO2 slope were determined. AT determination is commonly used when the VCO2-VO2 slope abruptly increases.20 Peak VO2 was the absolute value of peak oxygen uptake measured throughout the test, whereas peak MET was the relative value of peak VO2 divided by a constant 3.5 Ml/kg/min. The percent of peak VO2 to predicted value (predicted peak VO2%) was calculated by comparing the measured peak MET to the predicted peak MET using Taiwan’s normal standards.21 The slope of the VE/VCO2 ratio was measured from the start to just after the AT.22 The OUES was calculated using the graphic slope (a) of the equation VO2 = a log(VE) + b. The OUES was calculated using the total exercise time (OUES 100).13 Due to the anthropometric variation, the OUES was normalized by body surface area (BSA).23 Haycock formula was used to calculate the BSA.24

2.3. Statistical analysis

Before each analysis, normality and homoscedasticity were checked. To compare the outcomes of the adverse event group and adverse-free group, we used the Chi-square test or Fisher Exact test for categorical variables, independent t test for normally distributed variables and the Mann-Whitney U-test for nonnormally distributed variables. We plotted the receiver operating characteristic (ROC) curves and determined the optimal threshold values for each CPET variables for predicting 3-month, 2-year, and 5-year MACEs by selecting the point with the highest summation value of sensitivity and specificity. We used Kaplan-Meier survival analysis and the log-rank test to compare MACEs between the adverse event group and the adverse event-free group. To estimate the hazard ratio (HR) of each potential prognostic factors of CPET, we used univariate and multivariate Cox regression analysis. Because age and gender might be associated to poorer prognosis in AMI,25,26 we used a multivariable Cox regression model adjusted for age and gender to mitigate potential biases associated with these variables. To indicate statistical significance, a two-tailed p < 0.05 was used. For all analyses, we used Statistical Package for the Social Sciences for Windows, version 21.0 (IBM Corp., Armonk, NY).

For the calculation of the minimum sample size required for our study, we used the online calculator designed by the University of California, San Francisco, based on the HR regression model (URL: https://sample-size.net/sample-size-survival-analysis/). The calculator considers the type I error set at 0.05 and a test power set at 0.8. The ratio of the exposed group (with MACEs) to the non-exposed group (without MACEs) is set at 1:4, based on past literature. The relative hazard is estimated to be 2.0, referring to previous studies.3,27 The calculated minimum sample size is 102.

3. RESULTS 3.1. Study population

A total of 122 patients underwent predischarge CPET following the first AMI. Six patients were excluded (three with missing ECG data, three with incomplete echocardiography data). After thorough review, 116 patients remained, with 3 lost to follow-up within 5 years (1 at 6 months, 2 at 1 year after PPCI). Table 1 displays the baseline characteristics of participants categorized by outcome group. The analysis included the remaining 113 patients, with a mean Killip class of 1.82 ± 1.02. The interval between PPCI and CPET was 5.70 ± 3.23 days. Most patients were on dual antiplatelet therapy after PPCI. Demographics and clinical data were compared between AMI survivors with and without MACEs at 3 months, 2 years, and 5 years. No significant differences were observed in age, BMI, smoking status, AMI type, gender, comorbidities, basic biochemistry profile (including serum creatinine, total cholesterol, low density lipoprotein cholesterol, and glycated hemoglobin), or medications. Throughout follow-up, the group without MACEs exhibited higher left ventricular ejection fraction (p values: 0.011, 0.009, 0.006, respectively). Additionally, this group had fewer stenotic coronary arteries at 3-month and 2-year follow-up (p values: 0.002, 0.038).

Table 1 - Participant characteristics according to outcome group Variables Within 3 mo Within 2 y Within 5 y Adverse event-free group (n = 93) Adverse event group (n = 20) pa Adverse event-free group (n = 53) Adverse event group (n = 60) pa Adverse event-free group (n = 42) Adverse event group (n = 71) pa Sex-female, % 9 (9.7%) 1 (5.0%) 0.688 7 (13.2%) 3 (5.0%) 0.185 5 (11.9%) 5 (7.0%) 0.496 Age, y 58.3 ± 12.3 57.8 ± 13.5 0.869 59.2 ± 12.9 57.4 ± 12.2 0.433 58.5 ± 12.8 58.1 ± 12.4 0.881 BMI, kg/m2 24.9 ± 2.8 25.0 ± 2.3 0.887 24.8 ± 2.5 25.0 ± 2.8 0.708 24.8 ± 2.8 25.0 ± 2.7 0.758 Smoker, % 59 (63.4%) 14 (70.0%) 0.578 32 (60.4%) 41 (68.3%) 0.377 24 (57.1%) 49 (69.0%) 0.202 ACS type  STEMI 78 18 0.733 43 53 0.285 35 61 0.711  Non-STEMI 15 2 10 7 7 10 Comorbidities  Hypertension 60 (64.5%) 10 (50.0%) 0.132 37 (69.8%) 36 (60.0%) 0.276 28 (66.7%) 45 (63.4%) 0.724  Diabetes 29 (31.2%) 8 (40.0%) 0.446 16 (30.2%) 21 (35.0%) 0.587 12 (28.6%) 25 (35.2%) 0.467  Dyslipidemia 55 (59.1%) 15 (75.0%) 0.185 30 (56.5%) 40 (66.7%) 0.272 24 (57.1%) 46 (64.8%) 0.419 Biochemistry data  Cr, µmol/L 75.4 ± 16.1 77.1 ± 11.7 0.655 73.2 ± 14.3 77.9 ± 13.6 0.079 72.2 ± 12.9 77.8 ± 18.6 0.088  TC, mmol/L 5.0 ± 0.6 5.2 ± 0.5 0.130 4.9 ± 0.8 5.1 ± 0.6 0.161 5.0 ± 0.6 5.1 ± 0.7 0.407  LDL, mmol/L 2.7 ± 0.6 2.9 ± 0.5 0.198 2.7 ± 0.7 2.8 ± 0.6 0.258 2.7 ± 0.7 2.8 ± 0.7 0.444  HbA1C, % 6.2 ± 1.4 6.4 ± 1.6 0.532 6.1 ± 1.3 6.3 ± 1.4 0.399 6.1 ± 1.4 6.3 ± 1.6 0.415 Echocardiography  LVEF, % 48.3 ± 7.4 43.3 ± 10.1 0.011* 49.6 ± 7.1 45.6 ± 8.6 0.009* 50.2 ± 7.3 45.8 ± 8.3 0.006* Numbers of stenotic coronary arteries confirmed by initial angiography  1 vessel 44 (47.3%) 2 (10.0%) 0.002* 27 (50.9%) 19 (31.7%) 0.038* 22 (52.4%) 24 (33.8%) 0.079  2 vessels 29 (31.2%) 7 (35.0%) 17 (32.1%) 19 (31.7%) 13 (31.0%) 23 (32.4%)  3 vessels 20 (21.5%) 11 (55.0%) 9 (17.0%) 22 (36.6%) 7 (16.6%) 24 (33.8%) Medications  Ticagrelor 20 (21.5%) 6 (30.0%) 0.396 14 (26.4%) 12 (20.0%) 0.419 11 (26.2%) 15 (21.1%) 0.537  Aspirin 89 (95.7%) 19 (95.0%) 1.00 50 (94.3%) 58 (96.7%) 0.664 39 (92.9%) 69 (97.2%) 0.359  Beta-blocker 75 (80.7%) 17 (85.0%) 0.762 43 (81.1%) 49 (81.7%) 0.942 34 (81.0%) 58 (81.7%) 0.922  Clopidogrel 71 (76.3%) 14 (70.0%) 0.574 37 (69.8%) 48 (80.0%) 0.211 29 (69.1%) 56 (78.9%) 0.242  ACEI/ARB 78 (83.9%) 15 (75.0%) 0.345 47 (88.7%) 46 (76.7%) 0.095 38 (90.5%) 55 (77.5%) 0.080  Statins 61 (65.6%) 16 (80.0%) 0.210 34 (64.2%) 43 (71.7%) 0.392 28 (66.7%) 49 (69.0%) 0.796

Data are the mean ± SD or No. (percentage).

ACEI = angiotensin converting enzyme inhibitor; ACS = acute coronary syndrome; ARB = angiotensin receptor blocker; BMI = body mass index; Cr = creatinine; HbA1C = glycated hemoglobin; LDL = low-density lipoprotein cholesterol; LVEF = left ventricular ejection fraction; STEMI = ST-elevation myocardial infarction; TC = total cholesterol.

aAll the comparisons between two groups were done by independent t test except that comparisons of categorical variables between the two groups were done by Fisher exact test for gender, ACS type, medications (Ticagrelor, Aspirin, Beta-blocker, Clopidogrel, ACEI/ARB) or Chi-square test for smoker, comorbidities, numbers of stenotic coronary arteries confirmed by initial angiography and medications (Statin).

*p < 0.05.

3.2. Parameters of CPET

Among the 123 participants, 41 (36.3%) of them terminated CPET earlier. The reasons included fatigue (n = 20), dyspnea (n = 13), pallor (n = 5), chest tightness (n = 2), and dizziness (n = 1). Table 2 shows CPET parameter comparisons between the adverse event group and the adverse event-free group at each follow-up period. Patients in the adverse event-free group had higher OUES 100 (p = 0.004) and OUES 100/BSA (p = 0.002) within 3 months of the onset of AMI than those in the adverse event group. Patients without MACEs had higher peak systolic blood pressure (SBP) (p = 0.006), predicted peak VO2% (p = 0.016), OUES 100 (p = 0.002) and OUES 100/BSA (p < 0.001), and lower VE/VCO2 slope (p = 0.044) within 2 years of the onset of AMI than those with MACEs. Patients in the adverse event-free group had higher peak SBP (p = 0.031), OUES 100 (p = 0.003), and OUES 100/BSA (p = 0.001) than those in the adverse event group after 5 years.

Table 2 - Comparisons of variables of cardiopulmonary exercise testing according to outcome group Variables Within 3 mo Within 2 y Within 5 y Adverse event-free group
(n = 93) Adverse event group
(n = 20) pa Adverse event-free group
(n = 53) Adverse event group
(n = 60) pa Adverse event-free group
(n = 42) Adverse event group
(n = 71) pa Resting heart rate, bpm 75.6 ± 11.7 78.7 ± 14.7 0.314 76.9 ± 11.4 75.5 ± 13.1 0.542 78.4 ± 11.5 74.8 ± 12. 0.127 Peak heart rate—AT heart rate, bpm 10.0 ± 6.7 10.6 ± 6.1 0.722 10.0 ± 6.3 10.3 ± 6.8 0.829 9.5 ± 6.0 10.5 ± 6.9 0.404 AT heart rate—resting heart rate, bpm 17.7 ± 8.2 17.2 ± 10.2 0.822 17.6 ± 7.1 17.6 ± 9.8 0.955 17.7 ± 7.5 17.5 ± 9.2 0.926 Resting systolic BP, mmHg 116.3 ± 16.1 113.3 ± 16.0 0.437 117.0 ± 16.4 114.8 ± 15.8 0.471 114.1 ± 15.0 116.8 ± 16.7 0.383 Resting diastolic BP, mmHg 71.0 ± 11.0 67.8 ± 9.2 0.227 70.5 ± 9.9 70.3 ± 11.4 0.933 69.8 ± 9.5 70.7 ± 11.4 0.665 Peak systolic BP, mmHg 144.6 ± 24.2 136.5 ± 19.6 0.163 149.6 ± 25.3 137.6 ± 20.5 0.006* 149.4 ± 26.4 139.5 ± 21.0 0.031* Peak diastolic BP, mmHg 78.2 ± 15.5 74.1 ± 12.2 0.268 78.6 ± 15.1 76.5 ± 15.0 0.467 79.7 ± 16.1 76.2 ± 14.3 0.242 Peak respiratory exchange ratio 1.05 ± 0.12 1.04 ± 0.12 0.620 1.05 ± 0.13 1.05 ± 0.11 0.891 1.05 ± 0.14 1.05 ± 0.11 0.991 Peak MET 3.4 ± 0.8 3.1 ± 1.0 0.272 3.4 ± 0.8 3.2 ± 0.9 0.287 3.4 ± 0.8 3.3 ± 0.9 0.391 Maximal watt 53.6 ± 18.6 49.3 ± 17.5 0.349 53.4 ± 18.2 52.3 ± 18.7 0.750 54.0 ± 18.0 52.1 ± 18.7 0.597 Predicted peak VO2% 40.8 ± 10.7 37.7 ± 13.2 0.266 42.9 ± 11.2 37.9 ± 10.7 0.016* 42.3 ± 11.6 39.0 ± 10.8 0.144 Minute ventilation, L/min 25.4 ± 7.7 26.5 ± 10.8 0.592 25.1 ± 7.6 26.0 ± 8.9 0.561 25.4 ± 8.2 25.7 ± 8.4 0.877 VE/VCO2 slope 31.0 ± 8.0 32.5 ± 10.4 0.488 30.6 ± 7.3 32.8 ± 9.4 0.086 29.7 ± 5.4 32.2 ± 9.9 0.091 Peak rate pressure product 15 030 ± 3619 15 101 ± 4162 0.938 15 723 ± 3781 14 443 ± 3553 0.066 15 883 ± 3991 14 564 ± 3452 0.063 OUES 100 1.4 ± 0.4 1.1 ± 0.4 0.004* 1.5 ± 0.4 1.2 ± 0.4 0.002* 1.5 ± 0.4 1.3 ± 0.40 0.003* OUES 100/BSA 0.8 ± 0.2 0.6 ± 0.2 0.002* 0.8 ± 0.2 0.7 ± 0.2 <0.001* 0.8 ± 0.3 0.7 ± 0.2 0.001*

Data are the mean ± SD.

AT = anaerobic threshold; BP = blood pressure; BSA = body surface area; MET = metabolic equivalent; predicted peak VO2% = percentage of measured peak oxygen consumption to estimated peak oxygen consumption; OUES 100 = oxygen uptake efficiency slope calculated from data of the whole exercise duration; VE/VCO2 slope = minute ventilation to carbon dioxide production slope.

aAll the comparisons between two groups were done by independent t test.

*