A prospective study of patients with refractory angina: outcomes and the role of high‐sensitivity troponin T

Abstract

Background

The predictors of cardiovascular events in patients with chronic refractory angina are limited. High‐sensitivity cardiac troponin T (hs‐cTnT) assays are biomarkers that may be used to determine the prognosis of patients with stable coronary artery disease.

Hypothesis

Hs‐cTnT is a predictor of death and nonfatal myocardial infarction (MI) in patients with refractory angina.

Methods

We prospectively enrolled 117 consecutive patients in this study. A heart team ruled out myocardial revascularization feasibility after assessing recent coronary angiograms; evidence of myocardial ischemia served as an inclusion criterion. Optimal medical therapy was encouraged via outpatient visits every 6 months; plasma hs‐cTnT levels were determined at baseline. The primary endpoint was the composite incidence of death and nonfatal MI.

Results

During a median follow‐up period of 28.0 months (interquartile range, 18.0–47.5 months), an estimated 28.0‐month cumulative event rate of 13.4% was determined via the Kaplan‐Meier method. Univariate predictors of the composite endpoint were hs‐cTnT levels and LV dysfunction. Following a multivariate analysis, only hs‐cTnT was independently associated with the events in question, either as a continuous variable (hazard ratio per unit increase in the natural logarithm: 2.83, 95% confidence interval: 1.62‐4.92, P < 0.001) or as a categorical variable (hazard ratio for concentrations above the 99th percentile: 5.14, 95% confidence interval: 2.05‐12.91, P < 0.001).

Conclusions

In patients with chronic refractory angina, plasma concentration of hs‐cTnT is the strongest predictor of death and nonfatal MI. Notably, none of the outcomes in question occurred in patients with baseline plasma levels <5.0 ng/L.

1. INTRODUCTION

Refractory angina is a chronic condition affecting a growing number of individuals as the world's population ages and ischemic heart disease therapy improves. Refractory angina is currently defined as angina pectoris (AP) lasting >3 months as a result of coronary insufficiency in the setting of coronary artery disease (CAD) that cannot be controlled by a combination of medical therapy, angioplasty and coronary bypass surgery.1 Myocardial ischemia must be clinically established as the cause of the angina symptoms, although the pain experienced may arise or persist in the absence of this ischemia.2

Patients with refractory angina present with recurrent episodes of angina, some of which may be both limiting and severe, and experience very poor quality of life despite optimized medical treatment. Accurate data regarding the incidence and prevalence of refractory angina are not readily available; however, an estimated 15% of patients undergoing cardiac catheterization are not candidates for revascularization, including 6.7% of patients already receiving optimal medical therapy for symptomatic CAD,3, 4 a percentage that may increase as ischemic heart disease mortality continues to decrease.1, 5

Alternative therapies, such as enhanced external counterpulsation,6 spinal cord stimulation,7 stem cell therapy,8 and coronary‐sinus reducing device implantation,9 may be used to relieve refractory angina symptoms. However, data regarding the natural history of refractory angina and the predictors of cardiovascular events in this growing group of patients are limited. Mortality rates have ranged from approximately 4% to 10% in the largest prospective studies of patients with refractory angina.10, 11

High‐sensitivity cardiac troponin (Tn) assays have emerged as valuable biomarkers that may be used to predict major cardiovascular events in patients with stable CAD, independently of CAD risk factors, heart structure, or other cardiac biomarkers. Levels below the limit of detection of previous‐generation assays and below the 99th percentile cutoff for an apparently healthy population are strongly associated with the incidence of cardiovascular death, heart failure, or myocardial infarction (MI).12, 13, 14, 15, 16, 17 Evidence exists regarding the utility of this biomarker in aiding the risk‐stratification of patients with stable CAD; however, no evidence exists indicating that this utility would be retained in individuals with more severe and extensive disease, such as patients with refractory angina.

We sought to evaluate the predictors of death and MI in a prospective study of patients with refractory angina and the role of high‐sensitivity cardiac troponin T (hs‐cTnT) as a prognostic tool in this setting.

2. METHODS

2.1. Patients

We conducted a prospective study at the Heart Institute of the University of São Paulo (São Paulo, Brazil). The study had an “all‐comers” design involving all consecutive eligible patients who met the inclusion but not the exclusion criteria from October 2008 to September 2013. Eligible patients were those referred to a specialized clinic for refractory angina because of persistent chronic AP symptoms despite standard medical treatment. The study protocol was approved by the ethics committee of the Faculty of Medicine, University of São Paulo, and all procedures were performed in accordance with the Declaration of Helsinki. All patients were informed regarding the steps and procedures of the study and provided written informed consent to participate.

After undergoing a screening evaluation performed by the investigators, the following patients were enrolled in this study: (1) patients in whom myocardial revascularization was not feasible (recent patient coronary angiograms were reviewed by a heart team); (2) patients with grade II to IV chronic AP, as classified by the Canadian Cardiovascular Society (CCS); and (3) patients exhibiting evidence of myocardial ischemia within the previous year following any stress test. Patients unable to provide blood samples for the biomarker analysis due to venipuncture difficulty or refusal were excluded from the study. Patient eligibility and selection are summarized in Figure 1.

Figure 1.

Patient eligibility and selection. Abbreviations: CCS, Canadian Cardiovascular Society.

2.2. Follow‐up and Outcomes

Patients were followed every 6 months via outpatient visits, and optimal medical therapy was encouraged, including lifestyle interventions, risk‐factor reductions, survival‐improving medications, and maximally tolerated doses of antianginal drugs (β‐blockers, calcium‐channel blockers, short‐ and long‐acting nitrates, and trimetazidine). Patient contact information was updated at each outpatient visit or via telephone calls to reduce losses to follow‐up, and telephone calls were initiated to reschedule missed visits throughout the last month of the follow‐up period. The primary endpoint was the composite incidence of all‐cause mortality and nonfatal MI. Information regarding the primary endpoint was obtained from the medical records until September 2014. Myocardial infarction was considered for the primary outcome if an increase and/or decrease in cardiac Tn levels occurred—and ≥1 Tn value was greater than the 99th percentile upper reference limit—and symptoms of ischemia and/or electrocardiographic abnormalities (new ST‐segment/T‐wave changes, new left bundle branch block, or pathological Q waves) developed, according to the latest universal definition of MI.18 All events were adjudicated by trained cardiologists unaware of the participants’ baseline information.

2.3. Other Covariates

Information regarding each patient's baseline characteristics, including CAD risk factors, comorbidities, previous interventions, and baseline medications, was collected and subsequently entered into a prospective database during the first outpatient visit after the screening evaluation. The following data were also recorded during these visits: CCS angina class, New York Heart Association functional classification of heart failure, blood pressure, heart rate, body mass index, and biochemical and echocardiographic measurements.

All patients underwent either vasodilator stress perfusion cardiovascular magnetic resonance imaging or single‐photon emission computed tomography to detect reversible perfusion defects and to evaluate left ventricular (LV) function. Based on their left ventricular ejection fractions, the patients were categorized into the following 4 groups: preserved LV function (≥55%), mild dysfunction (≥40% but <55%), moderate dysfunction (≥30% but <40%), and severe dysfunction (<30%).

2.4. Biomarker Analysis

Blood samples were collected at baseline, immediately centrifuged, and stored in a freezer at −80°C until thawed for hs‐cTnT analysis. We used a commercially available high‐sensitivity assay for TnT (Elecsys Troponin T hs; Roche Diagnostics, Mannheim, Germany) with a limit of detection of 3 ng/L and a 99th‐percentile cutoff from an apparently healthy reference population of 14 ng/L. The technicians performing the hs‐cTnT measurements were blinded to patient outcomes.

2.5. Statistical Analysis

The categorical variables were expressed as percentages, and the continuous variables were expressed as mean ± SD. The distribution of the continuous variables was assessed via the Kolmogorov‐Smirnov test. The hs‐TnT levels were characterized by an asymmetric distribution and were expressed as medians and interquartile ranges (IQRs). Concentrations below the limit of detection were set equal to 2.99 ng/L for the statistical calculations, as previously described.19, 20

The event‐free survival curves for the composite endpoint were obtained using the Kaplan‐Meier method. Univariate analysis was performed using Cox regression to identify all the variables listed in Tables 1 and 2 that were associated with events (death and nonfatal MI), and those with P values ≤0.10 were considered for the multivariate analysis using a Cox proportional‐hazards regression model with a backward variable selection (age, chronic kidney disease, smoking status, hs‐cTnT, and LV function), enabling the calculation of both hazard ratios (HRs) and 95% confidence intervals (CIs). Hs‐cTnT levels were analyzed both as continuous variables (normalized via natural logarithmic transformation) and as categorical variables (above or below the 99th percentile of the reference population) in these models.

Table 1.

Baseline demographics, medical histories, and clinical characteristics of 117 patients

Characteristics	Value
Demographics
Age, y, mean ± SD	62.7 ± 9.4
Male sex	83 (70.9)
Medical history
HTN	108 (92.3)
DM	67 (57.3)
Dyslipidemia	105 (89.7)
Smoking status
Current	9 (7.7)
Former	64 (55.2)
Never	43 (37.1)
Family history of CAD	45 (40.5)
PAD	23 (19.8)
Valvular heart disease	10 (8.7)
CKD	16 (13.7)
COPD	4 (3.4)
Malignancy	5 (4.3)
Stroke	12 (10.3)
MI	81 (69.2)
PCI	64 (54.7)
CABG	95 (81.2)
Clinical characteristics
CAD, no. of vessels
1	8 (7.1)
2	20 (17.7)
3	85 (75.2)
Angina, CCS functional class
2	46 (39.3)
3	38 (32.5)
4	33 (28.2)
NYHA class
I	48 (41.4)
II	35 (30.2)
III	30 (25.9)
IV	3 (2.6)
LV function
Normal	72 (61.5)
Mild dysfunction	26 (22.2)
Moderate dysfunction	11 (9.4)
Severe dysfunction	8 (6.8)
Reversible perfusion defects noted via vasodilator stress tests, CMR, or SPECT	98 (83.8)

Abbreviations: CABG, coronary artery bypass grafting; CAD, coronary artery disease; CCS, Canadian Cardiovascular Society; CKD, chronic kidney disease; CMR, cardiovascular magnetic resonance; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; HTN, hypertension; LV, left ventricular; MI, myocardial infarction; NYHA, New York Heart Association; PAD, peripheral arterial disease; PCI, percutaneous coronary intervention; SD, standard deviation; SPECT, single‐photon emission computed tomography.

Data are presented as n (%) unless otherwise noted.

Table 2.

Baseline physical examination, laboratory, and medication parameters

Characteristic	Value
Parameters of physical examination, mean ± SD
SBP, mm Hg	128.3 ± 22.1
DBP, mm Hg	73.5 ± 12.6
Heart rate, bpm	65.2 ± 10.7
BMI, kg/m2	29.7 ± 5.0
Laboratory determinations
Hs‐cTnT, ng/L, median (IQR)	9.0 (4.0–14.0)
eGFR, mL/min, mean ± SD 1	76.3 ± 31.5
TC, mg/dL, mean ± SD	168.1 ± 44.3
LDL‐C, mg/dL, mean ± SD	95.6 ± 31.9
HDL‐C, mg/dL, mean ± SD	39.5 ± 11.2
FPG, mg/dL, mean ± SD	134.4 ± 54.0
Medications, n (%)
Aspirin or clopidogrel	103 (94.8)
Statins	117 (100.0)
β‐Blockers	110 (95.7)
CCBs	80 (69.6)
Long‐acting nitrates	98 (85.2)
Trimetazidine	36 (30.8)
Ivabradine	2 (1.7)
ACEIs/ARBs	97 (84.3)
Diuretics	56 (48.7)
Spironolactone	15 (13.3)
Digoxin	1 (0.9)
Anticoagulants (warfarin)	5 (4.4)
Oral antidiabetics	51 (44.3)
Insulin	17 (14.8)

Abbreviations: ACEI, angiotensin‐converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; CCB, calcium channel blocker; Cr, creatinine; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; FPG, fasting plasma glucose; HDL‐C, high‐density lipoprotein cholesterol; hs‐cTnT, high‐sensitivity cardiac troponin T; IQR, interquartile range; LDL‐C, low‐density lipoprotein cholecterol; SBP, systolic blood pressure; TC, total cholesterol.

¹eGFR calculated via the Cockcroft‐Gault formula based on Cr measurements.

We considered P values <0.05 statistically significant, and all tests were 2‐tailed for all analyses, which were performed using SPSS for Windows, version 19.0 (IBM Corp., Armonk, New York).

3. RESULTS

3.1. Baseline Characteristics of the Patients

The baseline clinical characteristics of the patients are presented in Table 1. The mean age was 62.7 ± 9.4 years; most of the patients were men (70.9%) and had preserved LV function (61.5%), with a mean ejection fraction of 52.4% ± 13.2%. Significant prevalence of 3‐vessel CAD (75.2%), chronic coronary total occlusion (95.6%), angina CCS class III or IV (60.7%), previous myocardial revascularization (91.5%), and comorbidities were identified. Reversible perfusion defects were noted in 98 patients (83.8%) via perfusion imaging and vasodilator stress testing. The patients’ baseline physical examination, laboratory, and medication parameters are provided in Table 2. At baseline, 67.9% of the patients were already taking ≥3 antianginal drugs. The median hs‐cTnT concentration was 9.0 ng/L (IQR, 4.0–14.0 ng/L), and 93 of the 117 patients (79.5%) enrolled in this study exhibited values at or above the limit of detection (3 ng/L). Thirty‐two patients (27.4%) exhibited concentrations ≥99th percentile (14 ng/L), which was coincident with the fourth quartile of hs‐cTnT levels; this group of patients exhibited ejection fractions significantly different from those of patients in the first 3 quartiles, as follows: 45.6% ± 11.8% and 54.9% ± 12.8%, respectively (P = 0.001).

3.2. Clinical Outcome

During a median follow‐up period of 28.0 months (IQR, 18.0–47.5 months), 19 combined events were noted (9 deaths and 10 nonfatal MIs), corresponding to an estimated cumulative event rate of the primary outcome of 13.4% (5.8% for all‐cause mortality). The Kaplan‐Meier analysis demonstrated a 7.7% incidence of combined events within the first year, 18.2% within 3 years, and 24.4% within 5 years of follow‐up. For death alone, the rates were 4.4% within the first year, 9.4% within 3 years, and 13.5% within 5 years. The baseline median hs‐cTnT concentrations differed significantly between patients who did and did not experience the primary outcome, as follows: 14.0 ng/L (IQR, 9.0–24.0 ng/L) and 8.0 ng/L (IQR, 3.0–12.0 ng/L), respectively (P = 0.001). No deaths occurred among patients with hs‐cTnT concentrations <9.0 ng/L, and no MIs were observed among patients with hs‐cTnT concentrations <5.0 ng/L.

The univariate predictors of death and nonfatal MI were the hs‐cTnT levels (HR per‐unit increase in the natural logarithm: 2.83, 95% CI: 1.62‐4.94, P < 0.001); an hs‐cTnT level ≥99th percentile for healthy individuals (HR: 5.18, 95% CI: 2.06‐13.00, P < 0.001), as shown in Figure 2; and either moderate or severe LV dysfunction (HR: 4.05, 95% CI: 1.62‐10.11, P = 0.003). Other clinically relevant variables exhibited a trend toward outcome prediction (P ≤ 0.10) and were included in both Model 1 (hs‐cTnT as a continuous variable) and Model 2 (hs‐cTnT as a categorical variable) in the multivariate analysis (Table 3). Only hs‐cTnT remained independently associated with combined events following multivariate analysis, whether as a continuous variable (HR: 2.83, 95% CI: 1.62‐4.92, P < 0.001) or as a categorical variable (HR: 5.14, 95% CI: 2.05‐12.91, P < 0.001).

Figure 2.

Survival free of death and nonfatal MI according to the 99th percentile cutoff of the hs‐cTnT level at baseline. The Kaplan‐Meier curves demonstrate significant increases in both mortality and nonfatal MI among patients with hs‐cTnT concentrations ≥99th percentile cutoff for healthy individuals. Abbreviations: hs‐cTnT, high‐sensitivity cardiac troponin T; MI, myocardial infarction.

Table 3.

Predictors of death and MI in patients with refractory angina ¹

	Univariate Model		Multivariate Model 1		Multivariate Model 2
	HR (95% CI)	P Value	HR (95% CI)	P Value	HR (95% CI)	P Value
Variable
Ln hs‐cTnT	2.83 (1.62‐4.94)	<0.001	2.83 (1.62‐4.92)	<0.001	—	—
Hs‐cTnT ≥ p99	5.18 (2.06‐13.00)	<0.001	—	—	5.14 (2.05‐12.91)	<0.001
Moderate/severe LVD	4.05 (1.62‐10.11)	0.003	1.73 (0.57‐5.24)	0.333	2.49 (0.93‐6.60)	0.068
Mean age	2.53 (0.91‐7.04)	0.075	2.09 (0.75‐5.81)	0.160	1.99 (0.70‐5.67)	0.195
CKD	2.30 (0.83‐6.40)	0.100	1.08 (0.32‐3.61)	0.900	1.12 (0.31‐3.99)	0.863
Angina CCS class III or IV	0.44 (0.18‐1.09)	0.077	0.49 (0.19‐1.24)	0.131	0.53 (0.21‐1.37)	0.192
Smoking status
Current	1.00	0.089	1.00	0.230	1.00	0.287
Former	0.41 (0.15‐1.14)		0.28 (0.06‐1.20)		0.32 (0.07‐1.40)
Never	1.54 (0.42‐5.62)		0.42 (0.10‐1.69)		0.55 (0.14‐2.25)

Abbreviations: CCS, Canadian Cardiovascular Society; CI, confidence interval; CKD, chronic kidney disease; HR, hazard ratio; hs‐cTnT, high‐sensitivity cardiac troponin T; Ln, natural logarithm; LVD, left ventricular dysfunction; MI, myocardial infarction; p99, 99th percentile.

¹Final multivariate Cox regression models with a backward selection of variables with a P ≤ 0.10 following univariate analysis. Hs‐cTnT entered Model 1 following logarithmic transformation and Model 2 as a categorical variable (≥99th percentile).

4. DISCUSSION

Unlike previous studies evaluating only mortality in the setting of refractory angina, this study also assessed the incidence of MI in a cohort of patients referred to a refractory angina clinic. Furthermore, a more substantial analysis of the clinically relevant variables presumably associated with these events was performed, and parameters such as myocardial ischemia and hs‐TnT, a cardiac biomarker of promising outpatient utility, were included.

The primary contribution of this study is the identification of hs‐TnT as the strongest predictor of death and nonfatal MI among patients with refractory angina, regardless of other common variables identified by other studies,7, 10, 21, 22, 23 including age, detectable myocardial ischemia, and LV systolic function.

4.1. Mortality and the Incidence of Nonfatal MI

In contrast to the high‐risk clinical profiles of patients with severe and extensive CAD, as well as the high prevalence of comorbidities, previous cardiovascular events and coronary interventions characteristic of such patients, we observed a low incidence of combined events, including an approximately 2% to 3% per year incidence of death and a 3.5% per year incidence of MI. Our study results were in accordance with historical trends demonstrating progressive decreases in mortality over time among patients with refractory angina. Previous studies reported annual mortality rates of 4% to 10%.7, 11, 22, 24 Our findings parallel those reported in the largest cohort of patients referred for refractory angina, a study that also featured the longest follow‐up period, in which a mortality rate of <4% per year was observed.10 Data regarding the incidence of MI in this population are scarce and restricted to specific studies regarding alternative therapies, such as transmyocardial revascularization and enhanced external counterpulsation; these studies reported MI rates of approximately 4% to 25% per year.21, 23, 25 The encouragingly low incidence of major cardiovascular events demonstrated in our study was most likely the result of intensive pharmacologic therapy use, as recommended in the current stable CAD guidelines,26 as well as optimal medical therapy, leading to outcomes approaching those of patients with complex CAD who are suitable for myocardial revascularization, as demonstrated in the (SYNTAX) trial.27

4.2. The Prognostic Value of hs‐cTnT in Patients With Refractory Angina

The detection of low circulating Tn levels was found to be the strongest tool for predicting cardiovascular events in patients with refractory angina. The reasons that may explain this remarkable power to predict primary cardiovascular outcomes warrant further analysis. The above‐mentioned findings may be attributable to Tn release into the bloodstream from a free cytosolic pool as a result of increased cell‐wall permeability in the setting of reversible ischemic injury,28, 29 although myocardial ischemia detection was not associated with hs‐cTnT levels in our study (perhaps because we used an indirect demonstration of perfusion defects—the vasodilator stress test—rather than a direct ischemia provocation test). Higher hs‐cTnT levels were independently associated with poorly developed coronary collateral circulation in patients with chronic total occlusion30 and coronary atherosclerotic plaques (burden, composition, and vulnerability).20, 31, 32 Therefore, a reasonable explanation for the excellent prognostic value of hs‐cTnT may be the occurrence of silent ruptures of vulnerable atherosclerotic plaques, which lead to microembolization and subsequent microinfarction in areas of myocardium not supplied by sufficient collateral circulation, and elevating Tn plasma levels before MI and cardiovascular death occurrence. A previous study demonstrated that a subsequent decrease in hs‐cTnT levels >50% was associated with a lower risk of cardiovascular death in a population of older adults.19 Future studies may entail using hs‐cTnT serial measurements in an outpatient setting to identify patients with stable CAD and chronic refractory angina who are at greater risk for cardiovascular events to develop a more intensive treatment regimen and to monitor patient responses to therapy. Whether this approach is capable of facilitating improved cardiovascular outcomes remains to be determined.

4.3. Study Limitations

The findings discussed in this manuscript should be weighed against the limitations of our study. First, this study was a single‐center prospective study limited to the number of included patients with refractory angina. However, this sample was representative of a specific patient cohort subjected to long‐term follow‐up and maximally tolerated drug therapy according to current clinical practice guidelines, irrespective of any alternative therapies utilized. Second, N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) and C‐reactive protein measurements with which to adjust the risk‐prediction model were not available. The combination of hs‐cTnT and NT‐proBNP has demonstrated improved mortality predictability in patients with stable CAD.33 Nevertheless, several studies have demonstrated only slight changes in the strength of the relationship between hs‐cTnT and cardiovascular outcomes following the inclusion of NT‐proBNP and C‐reactive protein in the model.12, 13, 19

5. CONCLUSION

In patients with refractory angina, the plasma concentration of hs‐cTnT is the strongest predictor of death and nonfatal MI. Notably, none of the outcomes in question occurred in patients with baseline plasma levels <5.0 ng/L.

Poppi NT, Gowdak LHW, Dourado LOC, Adam EL, Leite TNP, Mioto BM, Krieger JE, César LAM and Pereira AC. A prospective study of patients with refractory angina: outcomes and the role of high‐sensitivity troponin T, Clin Cardiol, 2017;40(1):11–17.