Multimodality imaging: Bird’s eye view from The European Society of Cardiology Congress 2018 Munich, August 25–29, 2018
- 292 Downloads
At the European Society of Cardiology (ESC) congress of this year 2018, held in Munich from August 25th to 29th, 4594 abstracts were presented. Of those, 423 (10.8%) belonged to an imaging category. Experts in echocardiography (VD), cardiovascular magnetic resonance (CMR) (CBD), nuclear imaging (OG), and cardiac computed tomography (CT) (PMH) have selected the abstracts in their areas of expertise that were of most interest to them and are summarized in this bird’s eye view from this ESC meeting. These abstracts were integrated by one of the Editors of the Journal (JB).
KeywordsEchocardiography computed tomography nuclear imaging cardiovascular magnetic resonance
Coronary artery disease
Cardiac magnetic resonance
Global longitudinal strain
Late gadolinium enhancement
Left ventricular ejection fraction
Positron emission tomography
Cardiac amyloidosis is characterized by diffuse LV hypertrophy and on speckle tracking analysis, the LV GLS is characterized by a typical apical sparing (with more preserved values of longitudinal strain in the apex, while mid and basal segments show severe impaired strain values). This LV GLS pattern is associated with poor outcomes.2 Diffuse LV hypertrophy can be observed in other diseases (hypertension, sarcomeric mutations…) and the prevalence and prognostic implications of relative apical sparing LV longitudinal strain in those other etiologies of LV hypertrophy are unknown. In 399 patients with diffuse LV hypertrophy and no significant coronary artery disease (CAD), Saito et al3 showed that relative apical sparing of LV longitudinal strain was present in 10% of patients. During a median follow-up of 4.8 years, 50 patients presented with major adverse cardiovascular events (MACE) (cardiac death, heart failure or acute myocardial infarction). The pattern of relative apical sparing of LV longitudinal strain was more frequently observed among patients presenting with an event. A relative apical sparing LV longitudinal strain index (ratio between the value of apical longitudinal strain and the average longitudinal strain of basal and mid ventricular segments) > 0.6 had incremental prognostic value over clinical and conventional echocardiographic parameters to predict the outcome. These results extend the evidence on the prognostic relevance of relative apical sparing LV longitudinal strain to other causes of LV hypertrophy.
A novel method that integrates blood pressure measurements and LV GLS provides noninvasive LV pressure-strain loops and permits calculation of global and segmental myocardial work in various cardiac diseases. Edwards et al4 estimated global and regional myocardial work and wasted work in 10 controls, ten patients with nonischemic cardiomyopathy and 14 patients with ischemic cardiomyopathy. Patients with ischemic and nonischemic cardiomyopathies had significantly lower values of myocardial work compared with controls. In addition, patients with nonischemic cardiomyopathy showed significantly wasted work in the septal wall compared with the lateral wall, suggesting enhanced systolic LV lengthening with segmental shortening during isovolumic relaxation, which can be a target for cardiac resynchronization therapy. In contrast, ischemic patients did not show regional differences in wasted work. This technology was also applied to differentiate physiological from pathological causes of LV hypertrophy in 20 controls, 24 professional athletes and ten patients with nonischemic cardiomyopathy.5 Patients with nonischemic cardiomyopathy showed the largest LV mass (317 ± 89 g) followed by the professional athletes (203 ± 57 g) and controls. However, in terms of global myocardial work, nonischemic cardiomyopathy patients showed the most impaired values (723 ± 431 mmHg%) whereas the professional athletes showed the highest values (1931 ± 485 mmHg%). Therefore, the etiology of LV hypertrophy is an important determinant of the global LV myocardial work.
Three-dimensional (3D) transesophageal echocardiography (TEE) is key in the evaluation of patients referred for transcatheter aortic valve replacement, particularly when the anatomy of the aortic valve is bicuspid. In 104 patients with bicuspid aortic valve stenosis, who were evaluated for transcatheter aortic valve replacement but who were eventually treated with surgical aortic valve replacement, Dr Wang6 evaluated the agreement between 2-dimensional (2D) transthoracic echocardiography, 3D TEE, and computed tomography (CT) to size the aortic annulus. The reference standard was intraoperative sizing of the aortic annulus. Measurements based on CT and 3D TEE had the best agreement with intraoperative sizing of the aortic annulus. However, the agreement between 3D TEE and intraoperative sizing was negatively influenced by the presence of heavy calcifications of the aortic valve.
Finally, the use of machine learning in echocardiography provides important insights. From the national echocardiographic database of Australia (NEDA), a vendor-agnostic cloud-based database containing echocardiographic data from ten Australian laboratories, Murphy et al7 evaluated the prognostic value of left atrial (LA) volume index and LA anteroposterior diameter. Data from 352,844 individuals with a mean follow-up of 5.4 years/person were analyzed. A total of 63,142 deaths were recorded. Left atrial volume index (LAVI) showed the strongest association with all-cause mortality. Individuals with a LAVI > 42 mL/m2 (upper quintile) had the highest age- and sex-adjusted risk profile (hazard ratio [HR] 1.9 compared with patients within the lowest quintile [LA volume index < 22 mL/m2]). The analysis based on deciles of LA volume index showed that the increase in mortality risk began with the LA volume index decile of 29-31 mL/m2 (which is close to the cut-off value proposed by current recommendations to define pathologic dilation of the LA).8
Machine learning principles (e.g. deep artificial neural networks = deep learning) are raising increasing attention as innovative approaches to explore and learn complex patterns within imaging data and improve human interpretation of large data clusters. Juarez-Orozco et al10 used a pretrained very deep convolutional neural network modified through transfer learning for the identification of patients who experience MACE based on direct image processing of PET myocardial perfusion polar maps. For this purpose, polar maps of myocardial perfusion reserve from 1199 patients (625 female, mean age 68 years) undergoing 13N-ammonia PET were analyzed. Deep learning was built through transfer learning by obtaining the architecture of the open source pretrained ResNet50 convolutional neural network and replacing the last layer and associated weights with a new layer specialized for the classification of patients who experienced or not MACE using a 9:1 training to testing ratio and 5-fold cross validation. Deep learning demonstrated an overall cross-validated validation accuracy of 77% with a corresponding precision and recall of 90% and 72%, respectively, and a specificity of 87% for identifying patients who experienced a MACE on follow-up. The positive likelihood ratio was 5.51, while the negative likelihood ratio was 0.32. The authors conclude that deep learning may offer additional clinical value in the identification of patients that will present MACE on follow-up.
Hybrid 18F-fluorodeoxyglucose (FDG) PET and magnetic resonance imaging (MRI) permit differentiation between myocardial scar from fibrosis associated with inflammation. Spinelli et al11 explored the value of FDG PET/MRI to detect early signs of myocardial inflammation in patients with Anderson-Fabry disease. Cardiac FDG PET/MRI and transthoracic strain echocardiography were performed in 24 heterozygous females carrying α-galactosidase A mutation and without LV hypertrophy. Thirteen patients showed focal FDG uptake (defined by a coefficient of variation in segmental FDG uptake > 0.17), including 2 patients with late gadolinium enhancement on MRI. Patients with a coefficient of variation > 0.17 had worse LV GLS compared to those with coefficient of variation ≤ 0.17 (− 18.5 ± 2.7% vs − 22.2 ± 1.8%, P = 0.024). The authors conclude that focal FDG uptake is an early sign of disease-related myocardial damage in patients with Anderson-Fabry disease and is associated with impaired LV longitudinal function. Moreover, their findings support the notion that inflammation plays an important role in glycosphingolipids storage disorders.
Keller et al12 report feasibility and safety profiles of pharmacological stress testing with the selective A2A receptor agonist regadenoson in 5151 consecutive patients undergoing 99m-TcTetrofosmin myocardial perfusion scintigraphy. The most frequent side effects were shortness of breath (66.2%), headache (21.3%), feeling of warmth (20.7%), feeling of pressure in the chest (18.1%) and in the “stomach” (17.4%). A new first degree atrioventricular block was observed in 0.1%. In four patients (0.08%), a severe symptomatic bradycardia or even life-threatening asystole occurred which could be immediately interrupted by administration of aminophyllin and atropine. These side effects were more often observed in patients with preexistent first degree atrioventricular block. This prospective registry demonstrates that regadenoson for pharmacological stress tests is generally well tolerated with frequent but harmless and transient side effects. Although very rare, life-threatening asystole or severe bradycardia (0.08%) may occur but can be rapidly treated with aminophyllin and atropine.
Cardiovascular Magnetic Resonance Imaging
Boldrini et al14 evaluated the accuracy of noncontrast CMR for clinical diagnosis of cardiac amyloidosis in a large multicenter prospective study of 868 patients with suspected systemic amyloidosis. In addition to comprehensive clinical investigations, the imaging protocol included native (noncontrast) T1 mapping and LGE imaging. The final diagnosis was light-chain cardiac amyloidosis in 222, transthyretin cardiac amyloidosis in 214, and no cardiac amyloidosis in 427 patients. Native T1 was higher in patients with light-chain or transthyretin cardiac amyloidosis compared with patients without cardiac amyloidosis. The diagnostic accuracy of native T1 to identify patients with cardiac amyloidosis was excellent (area under the curve 0.93). A native T1 < 1036 ms was the most sensitive value (99% sensitivity) to exclude cardiac amyloidosis while a native T1 > 1164ms was the most specific (99% specificity) to diagnose cardiac amyloidosis. Based on these results, the authors proposed a diagnostic algorithm with native T1 mapping to all patients with suspected cardiac amyloidosis, whereas contrast administration and LGE imaging would be reserved for patients with native T1 mapping between 1036 and 1164ms.
Feature-tracking CMR is a novel technique that permits assessment of myocardial deformation on cine CMR images. Using this methodology, Podlesnikar et al15 evaluated the effects of early metoprolol on the LV systolic function in 214 patients with anterior ST-segment myocardial infarction treated with primary percutaneous coronary intervention. The patients were included in the METOCARD-CNIC (Effect of Metoprolol in Cardioprotection During an Acute Myocardial Infarction) trial and randomized to early intravenous metoprolol vs standard care. CMR was performed 5-7 days after reperfusion, and LV global circumferential strain (GCS) and GLS were measured using feature tracking. During a median follow-up of 2 years, 22 MACE (composite of death, heart failure readmissions, reinfarction, and malignant ventricular arrhythmias) occurred. Patients with LV GCS > − 13.2% or with LV GLS > − 11.5% (more impaired LV systolic function) who did not receive early intravenous metoprolol showed the highest cumulative event rates compared with the patients who received metoprolol or had more preserved LV systolic function. These results underscore the role of feature-tracking CMR for detailed risk stratification of patients with ST-segment elevation myocardial infarction.
Phase-contrast cine MRI of the coronary sinus is a noninvasive method to quantify CFR which has been associated with prognosis in patients with suspected CAD. In 163 patients with diabetes mellitus, Kato et al16 assessed CFR during adenosine triphosphate infusion and at rest using phase-contrast cine MRI. CFR was calculated from myocardial blood flows at stress and rest. During a mean follow-up period of 3.7 years, 20 MACEs were identified, and patients with CFR < 2.0 had worse prognosis than those with preserved CRF > 2.0. On multivariable Cox regression analysis, CFR < 2.0 was independently associated with MACE (HR 11.29, P < 0.001). The study demonstrated that a noninvasive CFR assessment by phase-contrast cine MRI could represent a novel risk stratification tool in patients with diabetes mellitus.
Cardiac Computed Tomography
The 5-year follow-up results of the Scottish Computed Tomography of the Heart (SCOT-HEART) trial were presented at this ESC congress in Munich.17 Patients with stable chest pain who were referred to a cardiology clinic were randomized to coronary computed tomography angiography (CTA) in addition to standard care (n = 2073) versus standard care alone (n = 2073). The initial results of this trial showed that coronary CTA clarified the diagnosis and altered the subsequent management of the patients at short-term follow-up.18 Specifically, coronary CTA resulted in cancelations of a significant number of functional tests, increase the number of invasive coronary angiography and led to an increase in the prescription of preventive and antianginal medications. At 5-year follow-up, coronary CTA and the associated changes in treatment resulted in significantly lower rates of death from coronary heart disease and nonfatal myocardial infarction compared with standard care (2.3% vs 3.9%; HR 0.59; P = 0.004). Importantly, the rates of invasive coronary angiography and coronary revascularization were similar at 5 years in both arms.
In the EPICardial Adipose Tissue in HEART Diseases (EPICHEART) study,19 574 patients with severe aortic stenosis referred to surgical aortic valve replacement were included and the influence of epicardial adipose tissue (EAT) volume on coronary artery calcification was evaluated. Using CT, the EAT volume was measured. In addition, quantitative proteomics of the thoracic fat (including EAT, mediastinal, and subcutaneous adipose tissue) were performed. EAT volume was not associated with the extent of coronary artery calcification on CT. However, quantitative proteomics of the EAT showed that patients with coronary artery calcification had upregulation of pro-calcifying annexins, fatty acid-binding transporters, and inflammatory signaling proteins whereas fetuin-A and antioxidant enzymes were downregulated. In EAT, pro-calcifying annexins regulation was positively correlated with coronary artery calcium. This imbalance of pro-calcifying, pro-inflammatory and lipid transporter mediators was not observed in patients without coronary artery calcification. These results indicate that the biological characteristics of the EAT rather than the volume of EAT exerts an influence on the coronary atherosclerotic process. The Cardiovascular Risk Prediction using Computed Tomography (CRISP-CT) study provided further insight into the prognostic value of perivascular fat characteristics.20 The signals released by the inflamed coronary atherosclerotic plaques exert an effect on the perivascular adipose tissue inhibiting local adipogenesis. The changes in the composition of the perivascular fat around the inflamed coronary arteries result in a change in attenuation of coronary CTA (from more negative Hounsfield unit [HU] values to less negative) which can be quantified with the perivascular fat attenuation index. On a derivation cohort of 1872 participants with a median follow-up of 72 months and a validation cohort of 2040 participants with a median follow-up of 54 months, the perivascular fat attenuation index values around the proximal right coronary artery and left anterior descending artery were measured on CTA and correlated with the occurrence of all-cause and cardiac mortality. A value of perivascular fat attenuation index of − 70.1 HU or higher (less negative) was associated with 2.5-fold increase in all-cause mortality and 9-fold increase in cardiac mortality in the derivation cohort. This cutoff value was confirmed in the validation cohort (HR 3.69, P < 0.001 for all-cause mortality and HR 5.62, P < 0.001 for cardiac mortality). Therefore, perivascular fat attenuation index reflects coronary inflammation and could guide preventive therapies in primary prevention and intensify secondary prevention in patients with CAD.
As coronary CTA is becoming a widely available imaging technique and one of the first-line tests used in patients with suspected CAD, the radiation dose exposure is highly relevant. Hausleiter et al. presented the results of the Prospective Multicenter Registry on RadiaTion Dose Estimates of Cardiac CT AngIOgraphy iN Daily Practice (PROTECTION-VI) study that evaluated the radiation dose and the utilization of dose-saving strategies for cardiac CTA.22 The study conducted in 61 hospitals from 32 countries prospectively enrolled 4502 patients undergoing cardiac CTA during one calendar month in 2017. The coronary CTA data were analyzed in a central core-lab and compared with a similar dose survey performed in 2007. The median dose-length product (DLP) of coronary CTA was 195 mGy*cm, 78% lower than the DLP reported in the 2007 survey (P < 0.001). This reduction in radiation dose was not associated with the increase in the number of nondiagnostic coronary CTA (1.7% in 2007 vs 1.9% in 2017 surveys, P = 0.55). This large international radiation dose survey demonstrates considerable reduction of radiation exposure in coronary CTA during the last decade and underlines the importance of training and adaptation of contemporary cardiac scan protocols.
The department of Cardiology receives unrestricted research grants from Biotronik, Boston Scientific, Edwards Lifesciences, GE Healthcare and Medtronic. Victoria Delgado receives speaker fees from Abbott Vascular. The remaining authors have nothing to disclose.
- 3.Saito M, Wake D, Higaki R, Inaba S, Morioka H, Sumimoto T. Prognostic implication of relative apical sparing pattern in non-ischemic patients with diffuse left ventricular hypertrophy. Eur Heart J. 2018;39:ehy566-4920.Google Scholar
- 4.Edwards NFA, Chan J, Sabapathy S, Anderson B, Chamberlain R, Scalia G, on behalf of the CATHARSIS Study. Novel non-invasive left ventricular pressure-strain loop imaging demonstrates reduced myocardial work in cardiomyopathy with significant regional variation in non-ischemic cases. Eur Heart J. 2018;39:ehy566-4922.Google Scholar
- 6.Wang Y. Optimal annulus sizing approach for patients assigned for transcatheter aortic valve replacement with bicuspid aortic valve: comparison of echocardiographic and computed tomography derived measurement. Eur Heart J. 2018;39:ehy564-P677.Google Scholar
- 8.Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging. 2015;16(3):233-70.CrossRefGoogle Scholar
- 10.Juarez-Orozco LE, Knol RJJ, Martínez-Manzanera O, van der Zant FM, Knuuti J. The prognostic value of deep learning in PET myocardial perfusion for cardiovascular events. Eur Heart J. 2018;39:ehy565-1427.Google Scholar
- 11.Spinelli L, Imbriaco M, Nappi C, Nicolai E, Giugliano G, Ponsiglione A, Diomiauiuti TC, Riccio E, Pisani A, Trimarco B, Cuocolo A. Hybrid positron emission tomography-magnetic resonance imaging and speckle tracking echocardiography to detect early cardiac involvement function in females carrying alpha-galactosidase a mutation. Eur Heart J. 2018;64:862-4.Google Scholar
- 12.Keller M, Rippel M, Silber S. Side effects of regadenoson replacing adenosine for pharmacological stress tests in real world: prospective evaluation in over 5000 patients. Eur Heart J. 2018;39:ehy564-P584.Google Scholar
- 15.Podlesnikar T, Pizarro G, Fernández-Jiménez R, Montero-Cabezas JM, Sánchez-González J, Ajmone Marsan N, Bucciarelli-Ducci C, Fras Z, Fuster V, Bax JJ, Ibáñez B, Delgado V. Prognostic value of left ventricular global circumferential and longitudinal strain with feature tracking cardiovascular magnetic resonance after ST-segment elevation myocardial infarction. Eur Heart J. 2018;39:1043.CrossRefGoogle Scholar
- 16.Kato S, Saito N, Asahima N, Iinuma N, Kamimura D, Nakachi T, Fukui K, Iwasawa T, Kosuge M, Kimura K, Tamura K. Prognostic value of magnetic resonance imaging derived coronary flow reserve for patients with diabetes mellitus. Eur Heart J. 2019;39:630.Google Scholar
- 19.Mancio J, Barros A, Conceicao G, Santa C, Ferreira W, Carvalho M, Ferreira N, Vouga L, Miranda I, Vitorino R, Falcao-Pires I, Manadas B, Gama Ribeiro V, Leite-Moreira A, Bettencourt N. Regulation of coronary calcification by epicardial adipose tissue: traits in high-risk patients. Eur Heart J. 2018;39(supplement):1278.Google Scholar
- 20.Oikonomou EK, Marwan M, Desai MY, Mancio J, Alashi A, Hutt Centeno E, Thomas S, Herdman L, Kotanidis CP, Thomas KE, et al. Non-invasive detection of coronary inflammation using computed tomography and prediction of residual cardiovascular risk (the CRISP CT study): a post-hoc analysis of prospective outcome data. Lancet. 2018;392(10151):929-39.CrossRefGoogle Scholar
- 22.Stocker TJ, Deseive S, Leipsic J, Hadamitzky M, Chen MY, Rubinshtein R, Heckner M, Bax JJ, Fang XM, Grove EL, et al. Reduction in radiation exposure in cardiovascular computed tomography imaging: results from the Prospective Multicenter Registry on RadiaTion Dose Estimates of Cardiac CT AngIOgraphy IN Daily Practice in 2017 (PROTECTION VI). Eur Heart J. 2018;39:3715-23.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.