Measurements of the LVVs and LVEF are important determinants of clinical decisions in patients with HFrEF and frequently used as an end-point in clinical trials (2, 9, 10, 12). The most used technique for assessing the LV size and function is 2DE, being a rapid, non-invasive and cost-effective technique (16). However, measurements of the LVVs and LVEF by 2DE are limited by the geometrical assumptions of the LV shape and by the foreshortening of the LV (13), which reduce both accuracy and reproducibility of the method (15). Limitations of the 2DE are more prominent in patients with dilated and aneurismal LVs (21), often encountered in patients with HFrEF, in which the decisions based on echo measurements are essential for patient’s therapeutic and interventional management. Conversely, 3DE, which uses newly matrix array probes, provides qualitative datasets of the LVVs that measure the true volumes of the LV with minimal post-processing (4). Yet, even if 3DE seems to offer some advantages over 2DE, it is considered as a method with a long learning curve, reliable only when used by experienced echocardiographers.
Our study shows that 3DE can be used by cardiologists with different levels of training in 2DE, just after a short period of training in 3DE, with similar feasibility, but a better reproducibility than 2DE. The main results of our study showed that, in patients with HFrEF, with a wide range of LVs shapes, LVVs, and LVEF: 1) 3DE provides feasible and more reproducible measurements of the LVVs and LVEF than 2DE, when used by novel trainees in 3DE, regardless of their basic level of training in 2DE; 2) 3DE has a substantial agreement with 2DE for measuring the LVEF, when used by trainees with different basic levels of training in 2DE, and after just one month training in 3DE; 3) 3DE re-classifies more than 10% of the patients into having a LVEF less than 35%, therefore with an indication for device therapy, when used by an experienced user in both 2D and 3D echocardiography.
Feasibility and reproducibility of 2D and 3DE measurements performed by trainees in echocardiography. Jenkins et al (22), which performed a comparative study in 110 patients, showed 99% feasibility of the 2DE, and feasibility of 97% of the 3DE offline measurements of the LVVs and LVEF, respectively. As expected, our trainees presented a feasibility of the 2DE measurements according to their level of expertise in 2DE (91%, 94%, and 98% for the Beginner, Medium, and Advanced users, respectively). However, just after one-month training in 3DE post-processing, trainees provided similar feasibility of the 3DE measurements of the LVVs and LVEF from patients with HFrEF, irrespective of their basic level of training in 2DE.
It was showed that the reproducibility is lower when measuring LVVs and LVEF with 2DE than with 3DE, and this limitation might be overcome by the use of contrast agents for the 2DE techniques (6). In our study, when the 2DE measurements provided by trainees were compared to the ones of the Expert user, they presented reproducibility according to their basic level of training in 2DE. However, when compared again to the Expert user, the trainees provided increased reproducibility for the 3DE measurements, regardless of their basic training in 2DE. Our study shows that 3DE might be a more accurate and reproducible method to follow-up patients with HFrEF, just after a short period of training in 3DE.
Mor-Avi et al (23) showed that 3DE underestimates the LVVs when compared with CMR, in patients with a wide range of LVVs and LVEF. However, the underestimations of the LVVs decreases when the trabeculae are included in the 3DE measurements and the MV plane is excluded from the CMR measurements. We acknowledge the golden standard position of the CMR for the measurement of the LVVs. However, CMR is an expensive technique, with decreased availability, and the vast majority of clinical trials were based on echocardiographic measurements. Therefore, the purpose of our study was to show that 3DE might be a more feasible and reproducible alternative for the 2DE measurements, even after a short period of training of the users in 3D echocardiography.
In the study published by Mor-Avi et al, the reproducibility of measurements of the LVVs using 3DE decreased in parallel with the decreased experience of the centers involved in the project, mainly because of the technique in measuring (the inclusion versus the exclusions of the LV trabeculae inside the LVVs measured from the 3DE dataset). Our purpose was to show that 3DE is a method that can be easily learned, and might be a more reproducible technique for the follow-up of patients with HFrEF. Therefore, we performed a uniform training of our cardiologist in 3DE, which can explain the good results offered by all of them in terms of accuracy and reproducibility of the measurements.
Thavendiranathan et al (24) showed that 3DE is the best method for sequential measurements of the LVVs and LVEF, in 56 patients with chemotherapy treatment and stable LV function (global longitudinal strain, GLS more than 16%). The mentioned study was performed in patients with non-dilated LVs (3D_LVEDV ranged from 60 to 130 ml; 3D_LVESV ranged from 30 to 50 ml), and a normal LVEF (3D_LVEF ranged from 56 to 65%), which were assessed by a single experienced observer in echocardiography. Our study especially selected patients with HFrEF, having dilated and distorted LVVs (3D_LVED ranged from 85 to 381 ml, and 3D_LVESV ranged from 30 to 293 ml), and a wide range of LVEF (3D_LVEF ranged from 13–53%), which were measured by different observers, after a short period of training in 3DE.
3DE has a substantial agreement with 2DE for measuring the LVEF when used by trainees with different basic levels of training in 2DE, and beginner level in 3DE. From the best of our knowledge, a single study mentions the benefits of 3DE when comparing its use by experienced and non-experienced observers (25). However, the study was performed with transesophageal echocardiography and aimed to compare how experts and beginners localize the prolapsing segments of the MV, in patients with organic MR. Our study showed that trainees in echocardiography might provide increased accuracy and reproducibility of 3DE measurements of LVVs and LVEF performed in patients with dilated LVs and a wide range of LVEFs, by comparison with 2DE measurements. Therefore, our study strengthens the benefit of 3DE for the long-term follow-up of patients with HFrEF, even if used by beginners in the field.
3DE reclassifies the LVEF in patients with heart failure with reduced ejection fraction. A recent meta-analysis published by Rigolli et al (21) showed that 3DE is the method with the highest accuracy to measure the LVEF, by comparison with CMR. In our study, the LVEF measured with 3DE by each of the echocardiographers included in the study was lower than the LVEF measured by 2DE. Our results agree with a recent study published by Zanella et al (26), which showed that 3DE usually provides lower LVEF than 2DE, and the 3D LVEF was an independent predictor for major arrhythmic events in patients with LV dysfunction. Moreover, in the mentioned study (26), 3DE re-assigned up to 20% of patients as having a risk for sudden cardiac death according to the LVEF, and therefore, indication for device implantation. Our study showed that more than 10% of patients are re-classified by the Expert user into having an LVEF below the threshold of 35%, and therefore, indication for device therapy. The trainees in 3DE also re-classified the LVEF when using 3DE, but with a lower percentage of the reclassification than the expert in echocardiography.
Limitations. One limitation of our study is that we did not compare our measurements against the golden standard method, CMR. However, our study aimed not to test the accuracy of 3DE versus CMR, which was already tested in previous studies (23), but to assess the reliability of 3DE for the evaluation and follow-up of patients with HFrEF, after a short period of training in 3DE.