Adenosine cardiac magnetic resonance imaging (AS-CMR) has emerged as an alternative to other stress tests for identifying coronary artery disease. From January 1, 2002 to January 1, 2009, 564 consecutive patients underwent AS-CMR for evaluation of chest pain. The clinical characteristics, AS-CMR findings, and outcomes were evaluated by retrospective chart review and telephone interview. The median follow-up was 51 months. Major adverse cardiac events (MACE) were defined as cardiac death, nonfatal myocardial infarction, and revascularization with percutaneous coronary intervention or bypass surgery. The AS-CMR findings were normal in 264, ischemic in 201, and scar in 240 patients. No cardiac death occurred in the normal AS-CMR group. Among the ischemic and scar groups, 7.2% and 8.3% experienced an event, respectively. On univariate analysis, ischemia (hazard ratio 5.3, 95% confidence interval 2.5 to 11.5, p <0.001) and the presence of scar (hazard ratio 5.7, 95% confidence interval 2.6 to 12.4, p <0.001) were independent predictors of all cardiac events. Multivariate Cox regression analysis for MACE identified the presence of ischemia (hazard ratio 2.8, 95% confidence interval 1.2 to 6.2, p = 0.01) and scarring (hazard ratio 2.9, 95% confidence interval 1.3 to 6.6, p = 0.01) as the strongest independent factors. The annual event rate for hard events was 0% in the normal, 1.7% in the scar, and 1.5% in the ischemia group. For the MACE end points, the rate was 0.5% in the normal, 2.4% in the scar, and 2.6% in the ischemia group. In conclusion, in the present, single-center cohort with chest pain, normal AS-CMR findings conferred very low risk (<1% annually) of MACE. However, the findings of ischemia or scar were a significant and independent predictor of hard events and MACE.
Cardiac magnetic resonance imaging (CMR) is an evolving technique with increasing indications for a role in stress imaging that is yet to be defined. Although several studies have examined the diagnostic accuracy of stress CMR, only recently have data from a sufficient number of patients been collected to assess the prognostic value of this approach. Most data concerning the prognostic value of stress CMR findings have come from using dobutamine as a stressor and relying on wall motion abnormality as evidence of inducible ischemia. Early data for adenosine stress CMR (AS-CMR) have found it to have excellent sensitivity and specificity. Despite the wealth of data about its diagnostic power, little is known about the prognostic value of AS-CMR. Thus, we undertook the present study to evaluate the prognostic value of AS-CMR on important clinical outcomes in patients presenting with chest pain.
Methods
We identified 626 consecutive patients who had undergone AS-CMR at our institution from January 2002 to January 2009 for the purpose of evaluating for coronary artery disease (CAD) or ischemia. This stress modality was selected at the physician’s discretion and was performed to evaluate chest pain or 12-lead electrocardiographic or stress test abnormalities suspicious for ischemia due to CAD. The present study was a retrospective evaluation of the long-term prognosis of a large cohort of patients who had undergone AS-CMR as a routine part of a cardiac evaluation.
We excluded patients with <1 month of follow-up, those who had died within 6 months from noncardiac causes, patients with primary cardiomyopathy with severe left ventricular dysfunction or severe valvular heart disease, and those whose AS-CMR studies had experienced technical problems related to the adenosine infusion, magnetic resonance imaging acquisition, possible interference by rhythm abnormalities that could have affected the gating and image quality or compromised the interpretation, or software malfunction. To avoid the possibility of a cardiac event resulting from an early revascularization procedure (within 60 days after the AS-CMR), the patients who had undergone percutaneous coronary intervention and coronary artery bypass grafting within the first 60 days after the AS-CMR were also excluded. Thus, 564 patients were included in the final analysis.
The AS-CMR studies were performed using the GE Signa CV/I 1.5T system (GE Healthcare, Waukesha, Wisconsin) with an 8-element cardiac phased array coil. In obese patients, we applied torso or body coils. Gadolinium administration, adenosine perfusion, and delayed enhancement were used according to the standard protocol at our institution. Cardiac function was assessed with the patient at rest in 3 long-axis (2-chamber, 3-chamber, and 4-chamber) and contiguous short-axis views using steady-state free precession sequences. After infusion of adenosine at a constant rate of 140 μg/kg/min for 3 minutes, first-pass kinetic of a gadolinium-based contrast agent (gadopentetate dimeglumine [Magnevist]; 0.1 mmol/kg) was measured during breath-hold in 4 to 5 contiguous short-axis orientations at every heart beat using a hybrid gradient echo/echo-planar pulse sequence. Ten minutes after stress perfusion, a second perfusion in the same orientation and with the same setting was performed with the patient at rest. Ten minutes after this second bolus, late enhancement images were acquired using an inversion-recovery prepared gated fast-gradient echo pulse sequence.
Two experienced readers interpreted the images in consensus with off-line analysis software (GE Healthcare Advantage Workstation). Segmental analysis of the cine scans was performed by the consensus of 2 observers who were unaware of the patients’ history and follow-up data, using a synchronized quad-screen image display. Similarly, the same 2 observers, unaware of the patients’ history and follow-up data, interpreted the perfusion scans. For visual assessment of inducible perfusion deficits, the adenosine stress and at rest perfusion scans were magnified and displayed simultaneously. A perfusion deficit was deemed relevant if it was affecting >1/2 of the myocardial wall thickness in ≥2 neighboring segments persisting for >5 heartbeats after maximum signal intensity in the cavity of the left ventricle. A reversible perfusion defect seen at stress after adenosine infusion that was not present during the imaging study at rest defined the presence of ischemia. Late enhancement images were analyzed visually, and bright segments from the subendocardial to epicardial border defined the presence of scar due to myocardial infarction (MI). The amount of ischemia and scar was assessed visually; no quantification was used because no standardized quantification is currently available.
The follow-up data were retrieved by electronic chart review and scripted telephone interview. The date of the last interview or chart review was used to calculate the follow-up period. The follow-up period was completed in 2010. All the patients were followed up for occurrence of cardiac death and nonfatal MI as the primary end point and major adverse cardiac events (MACE) as a composite secondary end point that included cardiac death, nonfatal MI, and late revascularization therapy (percutaneous coronary intervention or coronary artery bypass grafting). Cardiac death (confirmed by a review of the Social Security Death Index interactive search site, hospital chart, or physician’s records) was defined as death from any cardiac cause (e.g., lethal arrhythmia, MI, or pump failure) or sudden unexpected death occurring without another explanation. MI was defined according to the 2007 European Society of Cardiology/American College of Cardiology Foundation/American Heart Association/World Heart Federation expert consensus document for the universal definition of MI. Nonfatal MI had to have been diagnosed ≥7 days after the date the AS-CMR study was performed (documented by appropriate cardiac enzymes and electrographic changes). Also, late revascularization therapy (percutaneous coronary intervention or coronary artery bypass grafting) had to have been received no earlier than 60 days from the date of the AS-CMR test. The annual cardiac event rate was reported. In addition, the cardiac event rate in the first 3 years and from the fifth to the seventh year of follow-up was recorded.
The data are reported across 3 AS-CMR groups: (1) normal, (2) any scar, and (3) any ischemia. The study patients could be included in both of the latter 2 groups. Continuous data are expressed as the mean ± SD, and pairwise comparisons were made using t tests. Categorical data are expressed as counts and percentages, and pairwise comparisons were made using chi-square tests. The primary end points were hard events, defined as cardiac death and nonfatal MI. The secondary end points were MACE, defined as a composite of cardiac death, MI, and late revascularization (percutaneous coronary intervention and coronary artery bypass grafting). Patients who underwent early revascularization occurring within 60 days of AS-CMR were removed from the analysis. Cox proportional hazards models were developed to evaluate the incremental prognostic value of the CMR-derived variables. A pre-CMR model was first developed using a forward stepwise variable selection algorithm, with a p value threshold of 0.05 for inclusion in the model. Candidate predictors for the pre-CMR model included age, gender, known CAD, hypertension, diabetes, hyperlipidemia, smoking status, and left ventricular ejection fraction. Next, the CMR variables for scar and ischemia were subsequently forced into this model to evaluate their incremental prognostic value. Event-free survival was estimated at several points using the Kaplan-Meier method, and survival curves were developed for the normal, any scar, and any ischemia groups.
Results
A total of 564 patients who had undergone AS-CMR at our institution from January 2002 to January 2009 met the inclusion and exclusion criteria. In 40 patients, the follow-up data from the electronic chart review only was incomplete or inadequate and required substantiation by telephone interview with the patient or a family member. All patients had to have had ≥1 follow-up visit to be included in the present study. Only 16 patients had <6 months of follow-up, and none was reported as deceased when cross-referenced against the Social Security Death Index interactive search site. According to the AS-CMR results, the patients were separated into 3 groups: normal AS-CMR findings, inducible ischemia, and the presence of scar. The baseline characteristics and cardiac risk factors of the study participants according to the CMR result classification are listed in Table 1 .The mean follow-up period was 48 ± 22 months (median 51, range 33 to 64). The outcome results stratified by the AS-CMR findings are summarized in Table 2 . Univariate Cox models were developed to identify the predictors of the primary and secondary end points from all the clinical and CMR variables. The results from multivariate Cox regression analysis identified inducible ischemia and scar on AS-CMR as the strongest independent factors influencing the occurrence of MACE ( Tables 3 and 4 ).
| All (n = 564) | Normal (n = 264) | Any Scar (n = 201) | Any Ischemia (n = 240) | p Value | ||
|---|---|---|---|---|---|---|
| Any Scar Versus Normal | Any Ischemia Versus Normal | |||||
| Age (yrs) | 62 ± 13 | 59 ± 14 | 65 ± 11 | 66 ± 11 | <0.001 | <0.001 |
| Men | 325 (58%) | 120 (45%) | 145 (72%) | 162 (68%) | <0.001 | <0.001 |
| CAD history ∗ | 293 (52%) | 76 (29%) | 161 (80%) | 173 (72%) | <0.001 | <0.001 |
| Hypertension | 373 (66%) | 152 (58%) | 153 (76%) | 178 (74%) | <0.001 | <0.001 |
| Diabetes mellitus | 158 (28%) | 51 (19%) | 82 (41%) | 88 (37%) | <0.001 | <0.001 |
| Hyperlipidemia † | 350 (62%) | 133 (50%) | 158 (79%) | 172 (72%) | <0.001 | <0.001 |
| Ever smoked | 251 (44%) | 96 (36%) | 109 (54%) | 125 (52%) | <0.001 | <0.001 |
| LVEF (%) | 53 ± 14 | 60 ± 8 | 42 ± 14 | 48 ± 15 | <0.001 | <0.001 |
∗ CAD history: defined by cardiac catheterization or stress test findings or reported in the chart.
† Total cholesterol ≥200 mg/dl, low-density lipoprotein cholesterol ≥130 mg/dl, or current medication use.
| AS-CMR Group | Death (n) | MI (n) | Revascularization (n) | |
|---|---|---|---|---|
| PCI | CABG | |||
| Normal (n = 264) | 0 | 0 | 8 | 0 |
| Ischemia (n = 240) | 8 | 15 | 15 | 3 |
| Scar (n = 201) | 14 | 11 | 11 | 1 |
| HR (95% CI) | p Value | |
|---|---|---|
| Age per 10 yrs | 1.8 (1.3–2.4) | <0.001 |
| Men | 2.0 (0.9–4.6) | 0.08 |
| CAD history ∗ | 8.6 (2.6–28.4) | <0.001 |
| Hypertension | 4.7 (1.4–15.5) | 0.01 |
| Diabetes mellitus | 5.2 (2.5–11.2) | <0.001 |
| Hyperlipidemia † | 3.0 (1.1–7.8) | 0.03 |
| Ever smoked | 1.6 (0.8–3.2) | 0.22 |
| LVEF per 10% decrease | 1.5 (1.2–1.9) | <0.001 |
∗ CAD history: defined by cardiac catheterization or stress test findings or reported in the chart.
† Total cholesterol ≥200 mg/dl, low-density lipoprotein cholesterol ≥130 mg/dl, or current medication use.
| Univariate Analysis | Multivariate Analysis | |||
|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | |
| Age per 10 yrs | 1.5 (1.2–1.9) | <0.01 | 1.3 (1.0–1.7) | 0.04 |
| Men | 1.3 (0.7–2.2) | 0.42 | ||
| CAD history ∗ | 5.6 (2.6–11.9) | <0.001 | 2.6 (1.2–5.8) | 0.02 |
| Hypertension | 3.1 (1.4–6.5) | <0.01 | ||
| Diabetes mellitus | 2.8 (1.7–4.8) | <0.001 | 2.3 (1.3–3.9) | <0.01 |
| Hyperlipidemia † | 2.3 (1.2–4.5) | 0.01 | ||
| Ever smoked | 1.4 (0.8–2.4) | 0.24 | ||
| LVEF per 10% decrease | 1.3 (1.1–1.5) | <0.01 | ||
| Any ischemia (vs normal) | 5.3 (2.5–11.5) | <0.001 | 2.8 (1.2–6.2) | 0.01 |
| Any scar (vs normal) | 5.7 (2.6–12.4) | <0.001 | 2.9 (1.3–6.6) | 0.01 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
