Reproduced from: Zamorano JL, Pinto FJ, Solano-López J, Bucciarelli-Ducci C. The year in cardiovascular medicine 2020:
imaging. Eur Heart J. 2021 Feb 14;42(7):740-749.
© The Author(s) 2021.
All rights reserved; no part of this publication may be reproduced, stored in retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission of the Publishers.
For Permissions please email:
The opinions expressed in the Journal item reproduced as this reprint are those of the authors and contributors, and do not necessary reflect those of the European Society of Cardiology, the editors, the editorial board, Oxford University Press or the organization to which the authors are affiliated.
The mention of trade names, commercial products or organizations, and the inclusion of advertisements in this reprint do not imply endorsement by the Journal, the editors, the editorial board, Oxford University Press or the organization to which the authors are affiliated. The editors and publishers have taken all reasonable precautions to verify drug names and doses, the results of experimental work and clinical findings in the journal. The ultimate responsibility for the use and dosage of drugs mentioned in this reprint and in interpretation of published material lies with the medical practitioner, and the editors and publisher cannot accept liability for damages arising from any error or omissions in the Journal or in this reprint. Please inform the editors of any errors.
Oxford University Press, OPL, and European Society of Cardiology are not responsible or in any way liable for the accuracy of the translated reprint, for any errors, omissions, or inaccuracies, or for any consequences arising therefrom. Croatian Cardiac Society is solely responsible for the translation and this reprint.
The past year has been a unique one owing to the outbreak of COVID-19, which has affected the population worldwide, with the ensuing economic and social consequences. The field of cardiology has not escaped this reality bringing with it changes in our everyday clinical praxis. The contribution of different imaging techniques to the cardiac involvement of COVID-19 with diagnostic and prognostic implications has been published very expeditiously. It is still pending to ascertain the long-term outcome of the different degrees of cardiac injury.
The recent publication of the ISCHEMIA trial (
Furthermore, advances in the knowledge and application of artificial intelligence (AI) are consolidating the need for greater attention and interest regarding a tool that in a few years will become part of our daily clinical practice. Finally, we highlight the introduction of new recommendations in the use of imaging techniques in the new practice guidelines.
We then summarize the most outstanding studies from the last year relating to the most relevant imaging techniques in current cardiology (
Graphical Abstract - Raw 3D data were streamed from standard echocardiograph using custom connection to 3D DICOM viewer workstation (CarnaLife Holo, MedApp, Krakow, Poland) for real-time, dynamic 3D rendering and wirelessly transferred into HoloLensmixed reality display (Microsoft, Redmond, USA) to overlay non-obstructive 3D data hologram upon reality view. Data were visible as a semitransparent holographic cube positioned in a convenient sector of visual field of echocardiographist and shared by interventional cardiologist. Reproduced with permission from Kasprzak et al. (
Echocardiography continues to be one of the most used methods to better understand cardiac pathophysiology and different pathological and even normal aspects of cardiac function and also plays a central role in daily patient management. Several papers have been published in 2020, and here, we highlight just a small proportion of the large amount of literature that has been produced during this year, a very unusual one, considering the COVID-19 pandemic that affected all of us.
One area of great current interest is transthyretin amyloidosis cardiomyopathy (ATTR-CM), an increasingly recognized cause of heart failure (HF) and with the new treatment strategies underway, some already with important clinical results; its recognition is becoming a must in clinical scenarios. Echocardiography has always played a role in the diagnosis of amyloidosis and that role is further strengthened with the exponential increase in relevance of amyloidosis. Chacko et al. (
The need for normal values is very important to set the references to determine the pathological boundaries. In this regard, the NORRE study provided useful reference ranges of 2D echocardiographic measurements of left ventricular (LV) layer-specific strain from a large group of healthy volunteers of both genders over a wide range of ages. (
The importance of developing parameters that may help the clinician to better understand the severity of certain disease conditions, as well as risk stratify the patients, is of utmost clinical relevance. That is the case of patients with bicuspid aortic valve (BAV). Kong et al., (
GLS is a strong predictor of adverse cardiovascular outcome in men. However, studies have indicated that GLS may not predict cardiovascular outcomes as effectively in women. Lundorff et al. (
Some new technological developments in echocardiography have also been described in some short papers, such as the development of a method of real-time streaming of 3D-transesophageal echocardiography data into head-mounted mixed-reality holographic display allowing for touchless control and data sharing within the cath-lab. The method was tested for the first time in human during percutaneous mitral balloon commissurotomy. (
Another outstanding study exploring the use of artificial intelligence in cardiac imaging is that of Ghorbani et al. (
Finally, in the latest published guidelines, we have appreciated the inclusion of echocardiography with class I recommendation, reflecting the relevance of this technique in routine cardiology practice. (
Over the last year, cardiovascular magnetic resonance (CMR) has confirmed an established role in the diagnosis, management, and prognosis of patients with chest pain, ischaemic heart disease, and non-ischaemic cardiomyopathies, further improved by AI and machine learning (ML).
The MR-INFORM trial is an unblinded, multicentre, clinical-effectiveness trial in patients with typical angina whose management was randomly assigned to a CMR stress perfusion-based strategy or an fractional flow reserve (FFR)-based strategy. (
The Stress CMR Perfusion Imaging in the United States (SPINS) study demonstrated excellent diagnostic and prognostic value of stress CMR in single-centre study. (
Recent developments on quantitative CMR stress perfusion with automated measurements using AI (
Knott et al. assessed the prognostic significance of this new technology in 1 049 patients with known or suspected coronary artery disease reduced myocardial blood flow (MBF) and myocardial perfusion reserve (MPR) quantified automatically inline were strong independent predictors of adverse cardiovascular outcome. For each 1 mL g−
AI and ML are providing new opportunities and pushing the envelope in cardiovascular imaging on faster better image analysis. Bhuva et al. (
The increasing use of AI in CMR post-processing and image analysis is improving measurements’ precision, accuracy and reliability which become less dependent on operator’s experience. This can have the direct consequence of empowering less-experienced centres to perform CMR, thus increasing CMR availability. Moreover, the improved diagnostics is also coupled with rapid image analysis which translated in improved physician time of efficiency, an attracting feature for busy clinical schedules.
Up to 30–40% of patients undergoing cardiac resynchronization therapy (CRT) show no improvement, and there is a necessity to improve the selection of patients. In a prospective multicentre study of 200 CRT recipients, Aalen et al. demonstrated that the combination of septal and lateral wall function measured by myocardial work with pressure-strain analysis on echocardiography and myocardial scar assessed by CMR LGE can offer a precise and relative simple approach to improve selection of CRT candidates, particularly in patients with ischaemic cardiomyopathy and/or intermediate QRS complex (QRS) duration. CRT response was predicted by the work difference between septum and lateral wall with an area under the curve (AUC) of 0.77 (95% CI 0.70–0.84). The combination of septal viability by CMR combined with myocardial work difference assessment significantly increased predicted CRT response reaching an AUC of 0.88 (95% CI 0.81–0.95). (
The role of CMR in the diagnosis of cardiac amyloidosis (CA) is becoming increasingly established. One of the most impactful technical developments this year is the demonstration that a novel approach called diffusion tensor CMR (DT-CMR) can characterize the myocardial microstructural effects of amyloid infiltration in patients. Khalique et al. showed that this contrast-free and radiation-free technique can identify the location and extent of the expanded disorganized myocardium. Moreover, novel imaging biomarkers of diffusivity and fractional anisotropy can effectively discriminate CA (n = 20) from hypertrophic cardiomyopathy (HCM) (n = 11). The preliminary results of this innovative in vivo technique suggest novel pathophysiological mechanisms and improved diagnostics, proving a promising new dimension in the assessment heart muscle disorders. (
The Hypertrophic Cardiomyopathy Registry (HCMR Registry) recruited 2755 patients with HCM from 44 sites in 6 countries, and includes CMR, genetic, and biomarkers data in order to improve risk prediction. The baseline data identified two distinct subgroups of patients: a group with sarcomere positive mutation and more fibrosis by CMR and a group sarcomere mutation negative with less fibrosis. (
In a single-centre study, Raman et al. (
Since the beginning of the COVID-19 pandemic, there are an increasing number of publications on the role of CMR in detecting myocardial damage in infected individuals. Whilst CMR has a clear clinical role in identifying cardiac damage in patients with a range of cardiovascular disease, the results of the CMR studies in COVID-19 patients to date (at the time of writing this manuscript) are still preliminary. Confirmatory results are warranted from large-scale multicentre studies with robust methodology before change in clinical management can be advocated. Most notably, an observational single-centre study in Germany (
The recommendations for the use of CMR in the diagnosis and management of patients with cardiovascular disease are increasing. In the latest release of ESC guidelines in 2020, the Guidelines for the Management of Acute Coronary Syndromes in Patients Presenting without Persistent ST-segment Elevation (
Over the past year, studies concerning computed tomography (CT) in the cardiovascular scenario have strengthened its ability as a predictor of cardiovascular events, and as a therapeutic guide in primary prevention.
Recently, ROBINSCA trial assessed the effectiveness of cardiovascular disease (CVD) screening in asymptomatic participants using the SCORE model (n = 12 185) or coronary artery calcium (CAC) scoring (n = 12 950). Both arms were stratified into low, intermediate, or high 10-year risk for developing fatal and non-fatal cardiovascular disease. SCORE screening arm identified 45.1% at low risk (SCORE <10%), 26.5% at intermediate risk (10–20%), and 28.4% at high risk (≥20%). According to the CAC screening, 76.0% were at low risk (Agatston <100), 15.1% at high risk (100–399), and 8.9% at very high risk (≥400). CAC scoring significantly reduced the proportion of individuals needing preventive treatment compared to SCORE (relative reduction women: 37.2%; men: 28.8%). (
From the multicentre CAC Consortium study, 66 636 asymptomatic patients with a CT were assessed, utilizing multivariate regression models for the risk of all-cause mortality and cause-specific mortality based on their CAC score. After adjustments, individuals with CAC ≥1000 had a 5.04-, 6.79-, 1.55-, and 2.89-fold risk of CVD, CAD, cancer, and all-cause mortality, respectively, compared to those with CAC score of 0. The CAC ≥1000 group had a 1.71-, 1.84-, 1.36-, and 1.51-fold increased risk of CVD, CAD, cancer, and all-cause mortality in comparison to those with CAC scores of 400–999. These leads to consider more aggressive preventive treatment for patients with CAC score ≥1000. (
The MESA Study investigators assessed the value of CAC for guiding aspirin allocation in primary prevention. All participants (n = 6470) underwent a baseline CAC score. CVD risk was estimated using the pooled cohort equation (PCE), defining three strata: <5%, 5–20%, and >20%. Based on PCE the number needed to treat at 5 years (NNT5) was greater than or similar to the number needed to harm (NNH5) among the three estimated cardiovascular risk strata. Conversely, CAC ≥100 and CAC ≥400 identified subgroups in which NNT5 was lower than NNH5. This was true both overall (for CAC ≥100, NNT5 = 140 vs. NNH5 = 518) and within all cardiovascular risk strata. Also, CAC = 0 identified subgroups in which the NNT5 was much higher than the NNH5. (
Olesen et al. stratified 48 731 patients by diabetes status and CAD severity (no, non-obstructive, or obstructive) assessed by coronary CT angiography (CCTA). With the median follow-up of 3.6 years, they found that diabetic patients had higher death rates than non-diabetic patients, irrespective of CAD severity. Still, those diabetic patients without CAD have a low risk of MI similar to non-diabetic patients. (
Finck et al. conducted a study with 1615 patients with suspected CAD who underwent a CCTA with morphological analysis of the atheromatous plaque. After an average of 10.5 years, there were 36 cardiac deaths and 15 non-fatal MI. Among characteristics of the plaque; the spotty or gross calcification pattern and the napkin ring sign (NRS) (low-attenuating central portion with ring-like higher attenuation) were predictive for events. Yet, only spotted calcified plaques and NRS convey further prognostic value above clinical features and the severity of coronary stenosis. In a stepwise approach, the prediction of endpoint beyond clinical risk could be improved by including the severity of CAD (x2 of 27.5, P < 0.001) and further discrimination for spotty calcified plaques (x2 of 3.89, P = 0.049). (
Another study assessed whether non-calcified low-attenuation plaque burden on CCTA might have a better predictor of MI than CAC or coronary stenosis severity. They followed up 1769 patients with suspected angina for median 4.7 years finding that low-attenuation plaque burden was the strongest predictor of MI (P = 0.014), irrespective of cardiovascular risk score, CAC score, or coronary artery stenosis. Patients with low-attenuation plaque burden >4% were almost five times more likely to have subsequent MI (P < 0.001). (
From the PARADIGM Study, 2252 patients who underwent clinically indicated serial CCTA at an interscan interval of ≥2 years with non-obstructive plaques (<50%) at baseline were studied. The aim was to prove whether the plaque atheroma volume (PAV), the percentage of diameter stenosis (%DS) or high-risk plaques (HRPs) were more likely to progress to obstructive lesions (>50%). On multivariate analysis, only the baseline total PAV and %DS independently predicted the development of obstructive lesions (P < 0.05), whereas the presence of HRP did not (P > 0.05). (
The investigators of the ICONIC study performed a nested case–control study of patients who underwent a CCTA prior developing an acute coronary syndrome. Culprit lesions were confirmed by invasive coronary angiography and coregistered to baseline CCTA images. They found that HRPs on baseline CCTA were less prevalent in non-obstructive plaques (19.7%) than in obstructive plaques (46.8%). Even though non-obstructive plaque comprised 81.3% of HRP lesions overall. Among patients with identifiable culprit lesion precursors, the adjusted HR was 1.85 (95% CI 1.26–2.72) for HRP, with no interaction between %DS and HRP. Compared to non-obstructive HRP lesions, obstructive lesions without HRP exhibited a non-significant HR of 1.41 (95% CI 0.61–3.25) (
Coronary computed tomography angiograms demonstrating high-risk plaque (HRP) in culprit lesion precursors. A 61-year-old male exsmoker exhibited a high-risk plaque extending from the (A) left main to the (B) proximal left anterior descending artery with (C) 41% diameter stenosis severity, (D) positive remodelling (white arrow), and low-attenuation plaque (green arrow). There is also diffuse calcification. One month later, the patient presented with a non-ST-elevation myocardial infarction. A 55-year-old male with hypertension and hyperlipidaemia exhibited a high-risk plaque with (E) only 35% DS severity, but (F) positive remodelling, low-attenuation plaque, and napkin-ring sign. The patient presented with a non- ST-elevation myocardial infarction 2months later. Reproduced with permission from Ferraro et al. (
Recently, the ADVANCE Registry presented its 1-year results of 4288 patients with suspected CAD in whom a 30% coronary stenosis was identified by CCTA. They evaluated the relationship of fractional flow reserve derived from CCTA (FFRCT) with clinical outcomes. There were 55 events; 78% of them occurred in patients with an FFRCT ≤0.80 (P = 0.06). Time to first event (cardiovascular death or MI) occurred more in patients with an FFRCT ≤0.80 compared with FFRCT >0.80 patients (25 [0.80%] vs. 3 [0.20%]; relative risk (RR): 4.22; 95% CI: 1.28–13.95; P = 0.01). Concerning the downstream care, the majority of patients in whom medical therapy was the recommended treatment strategy following FFRCT continued on only medical therapy at 1 year (92.9%), and when the site recommendation was for revascularization, the majority (68.9%) were revascularized. (
An innovative study introduces a new parameter of dynamic CT perfusion (CTP) called stress MBF rate (SFR). This is defined as the ratio of hyperaemic (ATP infusion) MBF in an artery with stenosis to the hyperaemic MBF in a non-diseased artery. Eighty-two patients were derived to invasive angiography for suspected CAD. Stress dynamic CTP and CCTA was performed before invasive angiography. Out of 101 vessels with 30–90% stenosis on invasive angiography, FFR resulted hemodynamically significant (<0.80) in 47.5% of them. SFR was lower for invasive FFR < 0.80 lesions (0.66 vs. 0.90; P < 0.01). Compared with ≥50% stenosis by computed tomography angiography (CTA), the specificity for detecting ischaemia by SFR increased from 43% to 91%, whilst the sensitivity decreased from 95% to 62%. The combination of stenosis ≥50% by CTA and SFR resulted in an AUC of 0.91, which was significantly higher than MBF alone. (
Nowadays, the potential survival benefit of ischaemia-guided early coronary revascularization in patients with stable coronary artery disease (CAD) is still in debate.
Patel et al. performed a single-centre cohort study including 16 029 patients with suspected or known CAD (mean age 68.6 ± 11.9 years) who underwent a Rubidium-82 (Rb82) rest-stress positron emission tomography (PET) myocardial perfusion imaging (MPI), excluding those with LVEF < 40%. After a median follow-up of 3.7 years, 1277 patients underwent early revascularization (87% PCI, 13% CABG), and 2493 (15.6%) died. After a propensity score adjustment for potential confounders, a Cox model found an interaction between %ischaemia and early revascularization (P < 0.001 for both all-cause and cardiac death). They also report medical therapy survival equipoise at 5% ischaemia. This ischaemia threshold for survival benefit is lower than previously reported with single photon emission CT (SPECT) MPI. (
In a phase-III prospective multicentric clinical study, the novel PET MPI tracer Fluorine-18 flurpiridaz is evaluated for its diagnostic efficacy detecting significant CAD (>50% stenosis in quantitative ICA) vs. SPECT. 755 patients (mean age 62.3 ± 9.5 years) were included. The PET MPI with the novel tracer demonstrated to have superior sensitivity than SPECT [71.9%, 95% CI 67.0–76.3%; P < 0.001 vs. 53.7% (95% CI: 48.5–58.8%)]. It was also superior to SPECT for defect size (P < 0.001), image quality (P < 0.001), diagnostic certainty (P < 0.001), and radiation exposure (6.1 ± 0.4 vs. 13.4 ± 3.2 mSv; P < 0.001). This is a new diagnostic tool with better diagnostic performance comparing to SPECT, in particular for women, obese, and patients undergoing pharmacological stress testing. (
Kwiecinski et al. presented a post hoc analysis of 293 patients with previous CAD who underwent 18-F-NaF PET. Of those, 203 (69%) showed increased coronary activity [represented by quantitative coronary microcalcification activity (CME)]. After a median follow-up of 42 months, 20 patients (7%) experienced fatal or non-fatal MI. All of them presented previously increased coronary 18F-NaF activity. On an ROC analysis, MI prediction was better for 18F-NaF CME score than coronary calcium scoring and different clinical risk scores. This represents a powerful and safe tool for the detection of coronary atherosclerotic inflammation. (
Another proof of improvements of imaging’s ability to predict events is the international multicentre study by Miller et al. in which they sought to determine the interactions between SPECT-MPI ischaemia, high-risk non-perfusion SPECT-MPI findings and MACE. In total, 16 578 patients with known or suspected CAD were analysed. Transient ischaemic dilation (TID) and post-stress wall motion abnormalities (WMA) were non-perfusion markers of ischaemia. After a median follow-up of 4.7 years, 1842 individuals presented one event. In a univariate analysis, the authors found that patients with mild ischaemia (<10%) and TID were more likely to present MACE compared with patients without TID (adjusted HR 1.42, P = 0.023). There were similar findings in patients with post-stress WMA. However, multivariable analysis of patients with mild ischaemia, TID (adjusted HR 1.50, P = 0.037), but not WMA, was independently associated with increased MACE. (
Heart to mediastinum (H/M) ratio measured by cardiac 123I-metaiodobenzylguanidine (123I-mIBG) scintigraphy has demonstrated prognostic significance in the setting of chronic HF. The OPAR Registry investigators describe a single-centre observational cohort study with 349 patients admitted for acute decompensated HF. 123I-MIBG imaging and echocardiography were performed before discharge. Of those 127 presented reduced EF, 78 mid-range EF, and 144 preserved EF. After a median follow-up period of 2.1 (±1.4) years, 128 patients presented cardiac events (HF hospitalization or cardiac death). A multivariable Cox analysis demonstrates that late H/M (after 200 min of tracer) was significantly associated with cardiac events in overall cohort (P = 0.0038), as in each EF subgroup (P = 0.0235 in reduced, P = 0.0119 in mid-range and P = 0.0311 in preserved). The authors conclude that H/M ratio reflects cardiac sympathetic nerve dysfunction, which is associated with cardiac events in acute HF patients, irrespective of EF. (
One-third of chronic HF patients who assign to CRT therapy based on guidelines classical eligibility criteria does not present benefits. Verschure et al. presented their results in 78 stable HF individuals with guideline-based criteria for CRT who underwent a cardiac 123I-mIBG imaging before device implantation. Late H/M ratio was an independent predictor of LVEF improvement to >35% (P = 0.0014) and early H/M for LVEF improvement of at least 10% from basal. (
CA implies ominous prognosis for patients. Early diagnosis with sufficient accuracy and safety remain still challenging. Rosengren et al. published the largest study of CA patients (both AL and ATTR) examined with Pittsburgh compound (11C-PIB) PET. In this study, the diagnostic accuracy of 11C-PIB PET is remarkable with high sensitivity (94%) and specificity (93% to 100%) for distinguishing CA patients from both non-amyloid hypertrophic and healthy controls. 11C-PIB uptake was significantly higher in AL-CA patients than in ATTR-CA patients (P < 0.001). In the study from Lee et al., they also demonstrate correlation between 11C-PIB uptake and myocardial histology in CA. In addition, after a median follow-up of 423 days, the degree of myocardial 11C-PIB uptake was a significant predictor of clinical outcome (death, heart transplantation, and acute decompensated HF) on multivariate Cox regression analysis (adjusted HR: 1.185; 95% CI 1.054–1.332; P = 0.005). (
Roque et al. used serial 18F-fluorodeoxyglucose (FDG PET/CT) after 1, 6, and 12 months in 37 post-aortic or mitral valve replacement patients. They obtained the standardized uptake values (SUVs) and a new proposed value denominate valve uptake index [(SUVmax − SUVmean)/SUVmax]. Of the 111 PET/CT performed, FDG uptake was visually detectable in 79.3% of patients, presenting a diffuse, homogeneous distribution pattern in 93%. No patient presented endocarditis during follow-up (
18F-fluorodeoxyglucose uptake distribution patterns (visual assessment). 18F-fluorodeoxyglucose uptake in non-infected prostheses (left panel), compared with an example of prosthetic valve endocarditis (right panel). Positron emission tomography/CTA fusion images of the valve plane (upper row), and their corresponding attenuation-corrected positron emission tomography images (lower row). From left to right, the characteristic inflammation patterns in order of descending frequency: diffuse homogeneous (93%), diffuse heterogeneous (7%), and focal/multifocal (2%). The diffuse homogeneous pattern is characteristic of inflammation and clearly differentiable from infection, whereas more focal. 18F-fluorodeoxyglucose uptake may overlap with infective endocarditis. No anatomic lesions were detected in any patient. Reproduced with permission from Roque et al. (
Tam et al. presented a study of FDG PET/CT in suspected LV assist devices (LVAD) associating their single-centre retrospective cases between September 2015 and February 2018 with a systematic review of PubMed from database inception through March 2018 involving in total 119 scans. Pooled sensitivity was 92% (95% CI: 82%–97%) and specificity was 83% (95% CI: 24%–99%) for FDG PET/CT in diagnosing LVAD infections. The ROC curve analysis demonstrated an AUC of 0.94 (95% CI 0.91–0.95). (
Another infectious scenario in which nuclear imaging techniques play an important diagnostic role is cardiac device-related infected endocarditis (CDRIE). Holcman et al. assessed the diagnostic accuracy of the hybrid technique of SPECT CT with technetium99mhexamethylpropyleneamine oxime-labelled leucocytes (99mTc-HMPAO-SPECT/CT). In a single-centre prospective study, 103 patients with suspected CDRIE who underwent 99mTc-HMPAO-SPECT/CT were included. They found that adding this nuclear technique improves the sensitivity of the modified Duke criteria alone (87% vs. 48%, P < 0.001), whereas a negative scan excludes CDRIE with high probability. This yielded a reduction in possible CDRIE diagnoses. (
The data underlying this article will be shared on reasonable request to the corresponding author.