Introduction

Due to differences in culture, healthcare systems, ethnicity [1], socioeconomic status [2] and disease prevalence [3, 4], existing guidelines for cardiac computed tomography (CT) and cardiac magnetic resonance (MR) developed by western professional societies are often not applicable in Asian countries. In March 2009, the Asian Society of Cardiovascular Imaging (ASCI), as the only society in Asia dedicated solely to cardiovascular imaging, nominated 7 representatives from different Asian countries to form a working group to provide recommendations on cardiac CT and cardiac MR. Detailed background of this project has previously been described in the ASCI cardiac CT criteria report, the first publication from the working group, which summarized the opinions of leading cardiac CT practitioners in Asia on 51 indications [5]. As the second step, we present here the ASCI cardiac MR appropriateness criteria. The purpose of this report is to serve as a reference for Asian practitioners to promote and improve their use of cardiac MR by providing appropriateness ratings for common clinical indications.

Methods

ASCI cardiac MR appropriateness criteria were developed through the same process as used for ASCI CT appropriateness criteria published earlier this year [5]. Briefly, we employed the modified Delphi method with one-round data collection to evaluate the cardiac MR appropriateness [6, 7]. A total of 25 panelists were nominated [Japan 6, Korea 5, Taiwan 4, China 3, Hong Kong (China) 3, Singapore 2, Thailand 2] by Working Group members, and approved by the Working Group with consensus.

In the development of the cardiac MR indications, the Working Group members agreed to use the 33 cardiac MR indications provided by the ACCF 2006 appropriateness criteria as the framework [8]. Indications considered for the ASCI 2010 cardiac CT appropriateness criteria were added and integrated to derive 50 indications which were approved by the Working Group. Among the 50 indications, 28 were in common with ACCF 2006 appropriateness criteria and 39 were in common with ASCI 2010 CT appropriateness criteria. Three indications [risk assessment in general populations with low, moderate and high coronary heart disease risk using coronary magnetic resonance angiography (MRA)] were original indications of ASCI cardiac MR appropriateness criteria.

A questionnaire was emailed to the 25 Technical Panel members. After completion, the questionnaires were collected by the ASCI office. The questionnaires were collected during a period between October 13 and November 11, 2009. Please refer to the online supplement for the complete questionnaire (Online Supplement 1).

Definition of cardiac MR

There are a variety of techniques used for cardiac MR [9]. Basic protocols might include cine magnetic resonance imaging (MRI) for wall motion and delayed gadolinium enhancement MRI for the assessment of scar [1019]. However, some may perform stress tests routinely using either perfusion MRI with adenosine [20, 21] or cine MRI with dobutamine [22, 23], while others may consider coronary and non-coronary MRA [24, 25] as important parts of cardiac MR examinations. Moreover, different techniques can be utilized to assess certain aspects of cardiac morphology and function [2629]. Since cardiac MR is still an intense field of research and development, it is also possible for appropriateness to be influenced by the availability of newer scanners and more sophisticated imaging techniques [30]. Thus, the Working Group decided to leave the definition of cardiac MR to the judgment of the Technical Panel members. Resulting variations in definitions might be an important reflection of the current perspectives of the leading Asian cardiac MR practitioners. In the questionnaire, the term “cardiac MR” was defined as including motion, stress and rest perfusion, delayed gadolinium enhancement, flow measurement, black blood T2-weighted imaging, and coronary MRA.

Rating system

The rating system used in this Asian survey is the same as previously used in other appropriateness criteria reports and ASCI CT appropriateness criteria. The panelists were asked to assess whether the use of cardiac MR for various indications was appropriate, uncertain or inappropriate. The Technical Panel scored each indication as follows:

  • Score 7–9: Appropriate test for the specific indication. Test is generally acceptable and a reasonable approach for the listed indication.

  • Score 4–6: Uncertain for specific indication. Test may be generally acceptable and may be a reasonable approach for the indication. Uncertainty also implies that more research or patient information or both are needed to classify the indication definitively.

  • Score 1–3: Inappropriate test for specific indication. Test is not generally acceptable and is not a reasonable approach for the indication.

In a panel with 23–25 members, ‘highly agreed’ was defined as 7 or fewer panelists rating outside the three-point region containing the median. ‘Disagreement’ was defined as at least 8 panelists rating in either extreme (1–3 and 7–9). Median values for each indication served as the final scoring if there was no disagreement among Technical Panelists [5, 7, 8]. If there was disagreement, the final appropriateness score was set as uncertain regardless of the median.

Results

The questionnaires were emailed to the Technical Panel members on October 13, 2009. Completed questionnaires were returned from 23 members [Japan 6, Korea 5, Taiwan 4, China 2, Hong Kong (China) 2, Singapore 2, Thailand 2] by November 11. Their specialties were radiology in 17 and cardiology in 6. The years of experience in the cardiovascular field ranged from 4 to 26 years while the experience of cardiac MR interpretation ranged from 300 to 3,000 examinations. For the cardiologists, the number of percutaneous coronary interventions performed range from 0 to 700 cases. The hospitals they were working in included city hospitals, medical centers, and university hospitals, with in-patient bed numbers ranging from 440 to 5,600. The complete list of Technical Panel members is provided at the beginning of this report.

Among the indications rated by Technical Panel, none showed disagreement. There were 24 appropriate, 18 uncertain and 8 inappropriate indications. Technical Panel members highly agreed in 22 indications, including 19 appropriate and 3 inappropriate indications. The ‘highly agreed’ inappropriate indications were: use of cardiac MR for evaluation of chest pain syndrome in patients with low pre-test probabilities of CAD, interpretable ECGs and ability to exercise; use of cardiac MR for detection of CAD in asymptomatic patients with low coronary heart disease risk; and use of coronary MRA for risk assessment in patients with low coronary heart disease risk. A detail appropriateness rating result is provided as an online supplement (Online Supplement 2).

Compared with the ACCF 2006 report [8], only 4/28 (14%) indications changed their category. Indication no. 38 (“evaluation of LV function following myocardial infarction or in heart failure patients”) and no. 49 (“to detect post PCI myocardial necrosis”) were shifted from uncertain to appropriate. Indication no. 30 (“evaluation of bypass grafts and coronary anatomy”) and no. 31 (“history of percutaneous revascularization with stents”) were shifted from inappropriate to uncertain.

Compared with the ASCI cardiac CT appropriateness criteria report [5], 29/39 (74%) were in the same appropriateness category. In 7 indications, cardiac CT received a more favorable category than cardiac MR: indication no. 2 (“detection of CAD: symptomatic, intermediate pre-test probability of CAD. ECG interpretable and able to exercise”), no. 27 (“use of MRI for CAD evaluation before valve surgery”), no. 29 (“evaluation of complex lesions before PCI”), no. 30 (“evaluation of bypass grafts and coronary anatomy”), no. 31 (“history of percutaneous revascularization with stents”), no. 33 (“evaluation of bypass grafts and coronary anatomy greater than or equal to 5 years after CABG”), and no. 34 (“evaluation for in-stent restenosis and coronary anatomy after PCI”). On the other hand, cardiac MR received a more favorable category than cardiac CT in 3 indications; indication no. 38 (“evaluation of LV function following myocardial infarction or in heart failure patients”), no. 48 (“to determine the location and extent of myocardial infarction including ‘no-reflow’ regions, post-acute myocardial infarction”), and no. 50 (“to determine viability prior to revascularization”).

The final ratings for cardiac MR are listed by indication sequentially (Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11) and by appropriateness category (Tables 12, 13, 14).

Table 1 Detection of CAD: symptomatic
Table 2 Detection of CAD: asymptomatic (without chest pain syndrome)
Table 3 Risk assessment: general population
Table 4 Detection of CAD with prior test results
Table 5 Risk assessment with prior test results
Table 6 CAD detection in pediatric patients with kawasaki disease
Table 7 Risk assessment: preoperative evaluation for non-cardiac surgery
Table 8 Risk assessment: preoperative evaluation for cardiac surgery or endovascular intervention
Table 9 Detection of CAD: post-revascularization (PCI or CABG)
Table 10 Structure and function
Table 11 Detection of myocardial scar and viability
Table 12 Appropriate indications (median score 7–9)
Table 13 Uncertain indications (median score 4–6)
Table 14 Inappropriate indications (median score 1–3)

Discussion

This ASCI cardiac MR appropriateness criteria report was developed in order to reflect the current status of cardiac MR in Asia and the opinions of Asian cardiac MR leaders about appropriate indications for cardiac MR. This report should prove useful in clinical practice in Asia, especially for institutes starting cardiac MR services for the first time.

Among the 50 indications evaluated in this report, 28 were in common with the ACCF 2006 appropriateness criteria report [8], 39 were also included in the ASCI 2010 cardiac CT appropriateness criteria report [5] and 3 indications were unique to this report. In contrast to the ASCI cardiac CT appropriateness criteria report in which an upward shift of appropriateness category was demonstrated in 51.3% (20/39) of the indications as compared with ACCF 2006 appropriateness criteria report, such a shift was seen in only 14.3% (4/28) of the indications in this cardiac MR appropriateness criteria report. The rapid advancement of CT technology [31] and associated accumulation of evidence of its clinical usefulness [3234] as well as reduction of its radiation levels [32] may explain the faster expansion of appropriate indications for cardiac CT compared to the expansion seen for cardiac MR, which has seen comparatively few technical advances over the past 5 years.

One of the most significant features of the ASCI cardiac CT and cardiac MR appropriateness criteria reports is the high number of indications evaluated for both CT and MR. Although cardiac CT was originally developed for visualization of coronary anatomy, recent studies have demonstrated the potential usefulness of one-stop shop cardiac examination in assessment of function, myocardial ischemia and myocardial viability [35, 36]. Meanwhile, the introduction of whole heart coronary MRA has enabled routine imaging of coronary anatomy which is completely noninvasive and without the need for radiation exposure and contrast medium [24, 37, 38]. Given the similarities in information obtainable, it is inevitable that CT and MR share many indications. In our questionnaire surveys, different panelists were selected for CT and MR. The panelists were not aware that similar surveys were being performed for the other modality, thus minimizing the extent to which their ratings were based on comparison to the other modality. Our survey demonstrated that CT received higher ratings than MR in the morphological assessment of native coronary arteries and bypass grafts before and after revascularization therapy. On the other hand, assessment of myocardial viability and fibrosis can be performed better with MR. However, most appropriateness ratings were similar for CT and MR, indicating that modality choice should be based on the technology and expertise available at each individual medical center.

“Use of coronary MRA in the risk assessment of general population” was evaluated in this survey. This indication was evaluated because coronary MRA has been gaining popularity as a screening tool in recent years, since the introduction of whole-heart coronary MRA [37, 39]. We found that experts in Asia consider this indication inappropriate in populations with low to intermediate coronary heart disease risk. Future research is needed to determine whether risk assessment of population with high coronary heart disease risk is appropriate or not.

This survey had several limitations. As was the case with the ASCI cardiac CT appropriateness criteria report, the Technical Panel in this study was dominated by experts from Eastern and Southeastern Asia reflecting the current academic contribution and participation in ASCI. We hope to see active participation in ASCI from Asian countries outside the Asia–Pacific region in the future. Secondly, many Technical Panelists proposed further clarification of the scan protocol. Although the importance of correct choice of MR scan protocol cannot be underestimated, this aspect is considered too complicated to be included in this questionnaire survey because of the diversity and rapid innovation of MR scan techniques used for cardiac examinations. Third, the comparison of CT and MR in the discussion section was done based on separate surveys. Since the panelists were not aware of the potential comparison, the comparison is not a ‘head-to-head’ comparison. Rather, the comparison is actually ‘what indications cardiac CT experts think are appropriate for cardiac CT’ vs ‘what indications cardiac MR experts think are appropriate for cardiac MR’. Although such comparison still gives us some reasonable insights on the appropriate choice of modality, ‘head-to-head’ comparison might be more desirable for appropriate use of cardiac CT and cardiac MR. However, in order to perform a ‘head-to-head’ comparison, we would need to subdivide the indications based on the patient’s age, sex, renal function, allergy to the contrast medium etc., which would run the risk of making the guidelines overly lengthy and complicated.

We expect that this ASCI 2010 cardiac MR appropriateness criteria report will serve as a timely and useful guide for the establishment of clinical cardiac MR services in Asian countries. ASCI will continue to pay close attention to this field and keep Asian practitioners updated about developments in cardiac MR and new indications as they arise.