Background
Accurate assessment and high reproducibility of left ventricular (LV) volumes, ejection fraction (EF), and myocardial wall motion are important factors for diagnosis of cardiac diseases [
1,
2]. Cardiac ultrasound (echocardiography) is currently the most widely used imaging modality in cardiac diagnostic imaging due to high cost effectiveness, wide availability in hospitals, generation of real-time images, non-ionization radiation and high feasibility. On the other hand, the image quality is suboptimal in some patients, which leads to relatively low accuracy and reproducibility compared to analysis based on magnetic resonance images in these cases [
3]. To overcome this problem, it has been demonstrated that the introduction of a contrast agent during echocardiography improves endocardial border delineation, and hence improves the accuracy of LV volume and EF measurements as well as myocardial wall motion score index (WMSI) evaluations [
4,
5]. Moreover, contrast-enhanced echocardiography has shown to reduce inter- and intraobserver variability [
4,
6].
Current guidelines from both the European and American Societies (European Association of Cardiovascular imaging (EACVI) and American Society of Echocardiography (ASE)) recommend the use of contrast-enhanced echocardiography in patients with suboptimal image quality or when there is a suspicion of structural abnormalities [
7,
8]. In spite of these recommendations, there appears to be an underuse of contrast agents during echocardiographic examinations. It has been proposed that approximately 10 to 15 % of all cardiac patients have indications for contrast-enhanced echocardiography [
5,
6]. Nevertheless, a recent study [
9] demonstrated that only 3.8 % of the echocardiographic examinations included a contrast agent during image acquisition. The underuse of contrast agent in clinical practice may alter the diagnosis and treatment strategies for the patients, as diagnostic indications can be omitted. It can be hypothesized that if a wider use of contrast-enhanced echocardiography would be adopted instead of the current selective approach, diagnoses such as myocardial ischemia and LV structural abnormalities may be detected earlier. Moreover, routine use of a contrast agent during echocardiography in all patients might also reduce the user dependency during image analysis, as the image quality increases, not only in patients with suboptimal image quality, but also in patients with good image quality without current indications for contrast-enhancement. The aim of this study was therefore to retrospectively investigate if contrast agent usage during echocardiography beyond the current recommendations affects the evaluation of cardiovascular function with respect to: 1) Assessment of regional wall motion abnormalities, 2) EF classification, and 3) Detection of LV structural abnormalities. A secondary aim was to evaluate if contrast-enhanced echocardiography reduced the inter- and intraobserver variability on LV volume and EF measurements as well as WMSI evaluations, irrespective of the reader’s experience level of contrast-enhanced echocardiography.
Discussion
The main finding from this study was that contrast-enhanced echocardiography significantly altered the assessment of cardiac function compared with greyscale echocardiography, in a patient group specifically selected due to not having an indication for contrast enhancement. This was mainly driven by an improved detection of regional wall motion abnormalities, evidenced by the fact that WMSI was significantly higher after contrast analysis compared with greyscale analysis. This was further evident when only considering patients (n = 83) with normal wall motion (WMS = 1) and sufficient greyscale image quality (no NA segments) in every segment, where as much as 55 % of these patients, previously classified without wall motion abnormalities, were re-classified having regional wall motion abnormalities after contrast-enhanced analysis. Keeping in mind that the addition of contrast agent also allowed for detection of LV structural abnormalities in 27 patients, emphasizes that a broader use of contrast-enhanced echocardiography can contribute to earlier detection of cardiovascular diseases.
Compared to these previous studies, the present study strictly selected patients without indications for contrast agent usage during standard echocardiography. Consequently, the clinical effect of contrast-enhanced echocardiography could be determined in patients without contrast indications. To our knowledge, this is the first study that investigates the potential benefits of contrast-enhanced echocardiography beyond current recommendations for resting echocardiography. Even though the image quality of the included patients would be considered sufficient for image analysis according to today’s guidelines of echocardiography (i.e. excluded if two or more contiguous segments were not visualized on greyscale images), the number of not visualized segments were remarkably reduced from 169 segments to 14 segments when using contrast. However, it should be noted that the most common WMS shift for those NA segments was a change to normal wall motion after contrast analysis (71 % of the segments), implying that improved detection of wall motion abnormalities for contrast-enhanced images is not only restricted to segments with poor image quality, and that decreased regional wall motion indeed can be more evident using contrast even when the endocardial border is considered well visualized. This is not surprising considering the extremely strong enhancement of the endocardial border and detailed visualization of trabecula due to the maximal separation of cavity and myocardium in white and black respectively in the contrast-enhanced images.
There was a significant difference in LV volumes and EF between the different imaging modes, with larger volumes and EF obtained for contrast-enhanced images. It is well known that LV volumes are underestimated when analyzing two-dimensional greyscale images, as the areas between the trabecula within the LV are excluded from the measurements, and that geometrical assumptions are made [
11]. Suboptimal image quality can also result in underestimated LV volumes as boundary detection becomes difficult. On the other hand, the difference in measured volumes between greyscale and contrast analysis had no significant impact on patient level, where most patients remained within the same EF classification (≥ 55 %, 45–54 %, 30–44 %, < 30 %). The relatively large EF intervals as well as the fact that both EDV and ESV were underestimated on greyscale images are possible explanations for these results.
The results from the study indicate improved reproducibility for WMSI and EF measurements when using contrast images instead of greyscale images, which is in line with previous findings in patients with mixed image quality [
4‐
6]. In our study, there was a sustained improved reproducibility even in this group with selected image quality. The improved reproducibility was observed both for repeated measurements by experienced readers and by in-experienced readers. Accordingly, contrast administration during echocardiography can therefore be suggested as a complement to minimize the user dependency of image analysis. This suggestion has also been purposed previously [
4].
It has been reported in literature that approximately 10 to 15 % of all cardiac patients have indications for contrast-enhanced echocardiography [
5,
6]. Following current recommendations, as much as 34 % of the examinations were actually excluded from this study due to being judged to have indication for contrast-enhanced echocardiography. One possible explanation is that this study was performed at a center with high contrast agent usage where the barrier to classify images as suboptimal is potentially lower than for a center with less experience of contrast-enhanced echocardiography, leading to increased exclusion rates.
The retrospective approach of the present study, where an objective gold standard, or alternative imaging modality to confirm the findings of regional wall motion abnormality interpretation is lacking, is a limitation. However, it should be kept in mind the contrast-enhanced imaging is known to improve diagnostic confidence and that several studies have shown that the results from the image analysis based on contrast-enhanced images correlate well with other imaging methods [
4,
12‐
14]. Additionally, a review of the results from previous stress echocardiography and/or coronary angiography examinations have been performed in a subgroup of patients (
n = 46 patients) where a shift from normal wall motion (WMS = 1 in every segment) to abnormal wall motion was observed when comparing greyscale and contrast analysis. In total, 22 of these patients were confirmed with myocardial ischemia or infarction. When combining the results from the contrast analysis at rest in the present study with previous studies, it was concluded that the wall motion abnormalities overlapped with the diseased areas in 21 patients out of these 22 patients. This confirms that the observed shift in wall motion abnormalities in the present study after contrast analysis may indeed alter the outcome for the patients and improve the diagnostic power of resting echo, even in patients with excellent image quality. Another limitation with the present study is the over representation of patients with ischemic cardiac disease. This is not a true representation of the patient cohort normally referred to echocardiography. In fact, this is a pilot study and we are about to perform a prospective randomized study in order to study the true potential benefit of routine contrast agent use.
This study is a potential first step towards a broader use of contrast-enhanced echocardiography instead of the selective approach used today. From a patient safety perspective it should be kept in mind that contrast-enhanced echocardiography is probably one of the best validated techniques, because of strict protocols and completely blinded readings required for approval of contrast agents [
15]. The contrast agents approved for clinical use are well tolerated, and serious adverse reactions are seldom observed [
16‐
18]. Adverse events are in most cases minor (e.g. headache, nausea, altered taste, sensation of heat) and self-resolving. Moreover, the risk for cavitation during ultrasound exposure is limited by applying imaging settings with a mechanical index lower than 0.5. Thus for the individual patient, there is a minor risk to broadening the clinical use of ultrasound contrast agents. The major barrier to overcome is probably more related to the clinical reality, where several obstacles have been pointed out for increased implementation of contrast use, such as absence of experienced specialists, training and accreditation for sonographers to independently perform contrast-enhanced echocardiography [
9].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MKL, RW and AB participated in the initiation and design of the study. CDS, EG, KS and RW (experienced readers) performed the ultrasound analysis including inclusion/exclusion of patients according to guidelines, volume measurements, wall motion analysis and screening for structural abnormalities. MKL and AABI (in-experienced readers) performed volume measurements and wall motion analysis in a subset of the patients. MKL, RW and AB processed all data and prepared the manuscript. All authors revised it critically and approved the final manuscript.