Impactful Cardiac CT and MRI Articles from 2023

Abstract

Cardiac imaging is important in diagnosing, treating, and predicting prognosis in patients with cardiovascular disease. Imaging protocols and analysis are consistently evolving, and the implementation of artificial intelligence–based applications is of increasing interest. This review presents recent advancements in noninvasive cardiac imaging, specifically focusing on cardiac CT and MRI, from notable publications across multidisciplinary journals in 2023 of interest to both radiologists and referring clinicians in the field. The discussion encompasses the latest trials of CT fractional flow reserve and the performance of the newest generation of photon-counting detector CT, particularly in coronary stenosis quantification. Additionally, it addresses coronary plaque quantification using artificial intelligence applications and their implications from large patient cohorts, alongside prognostic outcomes, and the value of coronary artery calcification scores. Various aspects of CT trials, such as anatomic planning before revascularization, high-risk plaque features, outcomes, and pericoronary fat index, are evaluated. New insights from cardiac MRI trials for cardiomyopathies, including cardiac amyloidosis, dilated cardiomyopathy, hypertrophic cardiomyopathy, myocarditis, and valvular disease, are also outlined. The review concludes by highlighting impactful societal statements and guidelines.

Keywords: CT Angiography, MR Imaging, Transcatheter Aortic Valve Implantation/Replacement (TAVI/TAVR), Cardiac, Coronary Arteries, Heart, Left Ventricle

© RSNA, 2024

Summary

This review explores 2023's most impactful cardiac CT and MRI studies for the evaluation of cardiovascular disease.

Essentials

• ■ The latest cardiac imaging studies highlight advances in CT fractional flow reserve–based analysis in patients with coronary artery disease and advances in photon-counting detector CT as applied to coronary stenosis quantification.

• ■ Studies also demonstrate the use of cardiac CT for assessing newer aspects of coronary artery disease including focal versus diffuse disease, high risk plaque features, and pericoronary fat.

• ■ The latest cardiac MRI trials and innovations suggest its feasibility in evaluating patients with dilated, hypertrophic, ischemic, and myocarditis cardiomyopathies.

Introduction

Cardiovascular imaging continues to play a central role in the diagnosis and management of cardiovascular disease. Across multiple modalities, noninvasive imaging continues to add to the diagnostic and prognostic tools available, with artificial intelligence complementing pre-existing concepts and technologies. Large multicenter trials and reviews facilitate our understanding of these techniques and concepts. This review aims to summarize the most relevant cardiovascular imaging articles, from the highest-impact journals, to assist radiologists and referring clinicians with an interest in the field. We selected articles published in 2023 based on the following inclusion criteria:

1. Prospective and retrospective cardiovascular imaging studies where imaging played a major role in determining primary or secondary outcomes. Such studies are crucial in highlighting the effectiveness and diagnostic value of imaging techniques, guiding clinical decision-making and improving patient care.

2. The most clinically relevant cardiovascular imaging studies from the highest-impact imaging and nonimaging journals, such as the Radiology suite of journals and the New England Journal of Medicine, respectively. Prioritizing studies from high-impact journals ensures that selected articles are of high scientific quality and widely recognized by the medical community and are thus most likely to shape clinical practice guidelines and imaging recommendations.

3. Cardiovascular imaging studies showing new and important findings in the field, ensuring they have been widely read by referring clinicians. Selecting studies with new and important findings ensures that the review is at the forefront of imaging advancements, offering new imaging insights and potential improvements in diagnostic decision-making.

An overview of included articles, with the key point from each study, can be found in the Table.

Overview of Selected Articles, Journals, and Key Points

Coronary Artery Disease

Recent CT Scan Platform Advancements for Coronary Imaging Photon-counting Detector CT

Simon et al (1) conducted a single-center study of 812 patients with suspected coronary artery disease (CAD) that compared photon-counting detector (PCD) CT to conventional CT. In patients who underwent PCD CT, the number of subsequent invasive coronary angiography (ICA) procedures was lower as compared with patients who were scanned with conventional CT. This difference was greater in patients with extensive coronary artery calcification. The prevalence of coronary stenosis was lower in the PCD CT group compared with the conventional CT group. Nonobstructive CAD was 11% more frequent in the PCD CT group, while obstructive CAD was 9% more frequent in the conventional CT group. These differences were more obvious in patients with extensive coronary artery calcification (coronary artery calcium score [CACS] ≥ 400), where obstructive CAD was 29% more frequently reported in the conventional CT group. In cases where 0.2-mm PCD CT reconstructions were available, the rate of obstructive disease was 48.6%. Conventional CT was independently associated with threefold higher odds of the need for ICA compared with PCD CT and almost fourfold higher odds in patients with extensive coronary artery calcification. This study highlighted that, particularly in patients with high calcium scores, use of PCD CT can reduce ICA rates relative to use of conventional CT.

Rapid advances in deep learning reconstruction technology are also leading to improved image quality for stenosis detection compared with conventional reconstruction methods. In a retrospective study of 58 patients who underwent coronary CT angiography (CCTA) using 320-row CT, super-resolution deep learning reconstructions resulted in improved vessel sharpness, stent strut thinness, stent lumen enlargement, and stenosis grading (Fig 1) (2). Such advancements in artificial intelligence combined with progress in PCD CT are substantially enhancing the accuracy of coronary artery evaluation.

Figure 1:

Images in a 66-year-old male individual after percutaneous coronary intervention of the left anterior descending artery. (A, B) On coronary CT angiogram, edge of stent and calcified and noncalcified plaque all appear sharper using (B) super-resolution deep learning reconstruction (SR-DLR) than using (A) conventional deep learning reconstruction (C-DLR). (C) Invasive coronary angiogram (ICA) demonstrates mild stenosis in the left anterior descending artery. The study highlights the improving capabilities of SR-DLRs to improve coronary image reading. (Reprinted, with permission, from reference 2.)

CT Fractional Flow Reserve

The Prospective Randomized Trial of the Optimal Evaluation of Cardiac Symptoms and Revascularization (PRECISE) study by Douglas et al (3) aimed to build on prior randomized trials investigating the clinical utility of cardiac CT in clinically stable patients with suspected CAD. The study analyzed individuals in an outpatient setting and explored whether CCTA combined with computation of CT fractional flow reserve (CT-FFR) would impact cardiac catheterization efficiency and incidence of major adverse cardiovascular events (MACE) compared with traditional testing. Participants were randomly assigned 1:1 to the precision strategy or usual testing groups. Within the precision strategy arm, participants’ risk for CAD was calculated prior to any testing by using the validated Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) minimal risk score (PMRS) aiming to reduce testing in low-risk participants (even allowing for no testing) and using CT in those at higher risk. Participants in the precision strategy group with a PMRS value of greater than 0.46 were assigned to deferred testing, accounting for approximately 20% of participants. Participants in the precision strategy group with a PMRS less than 0.46 or with known atherosclerosis underwent CCTA with CT-FFR for stenoses visually graded between 30% and 90%. Site clinicians chose the initial testing modality for participants in the usual testing group, including exercise electrocardiography, stress echocardiography, stress nuclear myocardial perfusion imaging (SPECT or PET), stress cardiac MRI, or ICA (Fig 2). This precision strategy approach resulted in significantly lower rates of ICA without obstructive disease (27 of 1047 [2.6%]) compared with the usual testing group (107 of 1046 [10.2%]) and an increase in the percutaneous coronary intervention rate. The PRECISE study showed that this pathway of risk stratification with deferred testing for low-risk patients and CT with selective CT-FFR for the remaining patients reduced rates of ICA without obstructive CAD and had no statistically significant adverse effect on safety (death or myocardial infarction at 1 year), building on the work of the Scottish Computed Tomography of the Heart (SCOT-HEART) (4) and PROMISE (5) trials. Notably, the precision strategy group had higher use of optimal medical therapy (OMT) at 1 year compared with the usual testing group, despite less frequent use of ICA. These findings closely follow SCOT-HEART and PROMISE, highlighting that detailed anatomic and physiologic data from CT (with or without CT-FFR) enhance the use of appropriate OMT.

Figure 2:

The Prospective Randomized Trial of the Optimal Evaluation of Cardiac Symptoms and Revascularization (PRECISE) study evaluated the clinical utility of CT in the stable chest pain setting evaluating whether a combined CT and CT-FFR strategy would impact catheterization laboratory efficiency and incident major adverse clinical events (MACE) compared with the usual testing pathways. Each phase of the PRECISE trial is outlined in A, a key point being that the low pretest probability group (20.2%) underwent deferred testing. Kaplan-Meier curves at a median of 11.8 months show (B) the primary composite end point (death from any cause, nonfatal myocardial infarction [MI], invasive catheterization without obstructive coronary artery disease [CAD]) and (C) death or nonfatal MI. The insets show the same data on an enlarged y-axis. The most important results were that catheterization without obstructive CAD was improved using CT-FFR and there was no statistically significant impact on safety (death, nonfatal MI) at 1 year. CT-FFR = CT fractional flow reserve, PMRS = Prospective Multicenter Imaging Study for Evaluation of Chest Pain minimal risk score. (Adapted, with permission under a CC BY-NC-ND 4.0 license, from reference 3.)

Udelson et al (6) further assessed the PRECISE trial and showed that the strategy of deferred testing among participants identified as minimal risk using the PMRS was safe and was associated with less ICA without obstructive CAD compared with the usual testing group. There were fewer ICA procedures overall and a similarly low number of revascularizations in each group (0.5%). Among those in the precision strategy group of deferred initial testing, 36% were ultimately referred for a test (median of 48 days after randomization) for new, worsening, or uncontrolled symptoms or clinical or participant concern. When subsequent tests were performed in this group, results showed no CAD or ischemia in 96% of participants.

The Effect of On-Site CT-Derived Fractional Flow Reserve on the Management of Decision Making for Patients with Stable Chest Pain, or TARGET, trial randomly assigned patients with stable angina to either the CT-FFR group (guideline-directed medical therapy or ICA guided by on-site CT-FFR) or the standard care group (guideline-directed medical therapy or ICA guided by stress tests). The study recruited individuals with intermediate to high pretest probability of obstructive CAD who were found to have CAD at CT with between 30% and 90% stenosis in at least one major coronary artery (7). Within the CT-FFR group, participants with CT-FFR of 0.80 or less in one or more major coronary artery were referred for ICA; OMT was recommended when the CT-FFR was greater than 0.80. The primary end point was the proportion of participants undergoing ICA who did not have obstructive CAD or those with obstructive disease who did not undergo intervention within 90 days. Compared with the standard care group, the proportion of participants undergoing ICA within 90 days was reduced in the CT-FFR care group (46.2% vs 28.3%, P < .001), with a smaller proportion of participants having nonobstructive CAD in the CT-FFR group (20.9% vs 38.0%). The proportion of participants with obstructive disease who did not undergo intervention within 90 days was 7.4% and 8.1% in the CT-FFR and standard care groups, respectively. Revascularization at 90 days was higher in the CT-FFR group (49.7% vs 42.8%, P = .02). There was a significant increase in the use of OMT from baseline to 1 year in both groups.

Quantitative Coronary Artery Plaque Evaluation

Tzimas et al (8) evaluated 11 808 patients from multiple sites who underwent CT for quantitative assessment of total plaque and calcified, noncalcified, and low-attenuation plaque volumes using Artificial Intelligence–Enabled Quantitative Coronary Plaque Analysis (AI-QCPA; HeartFlow). Age- and sex-stratified percentile nomograms for atherosclerotic plaque volumes were developed based on CT findings by using this new AI-QCPA tool (Fig 3). The results from this study provide context for quantitative plaque volumes and a nomographic framework to encourage future investigation of the relationship between clinical outcomes and quantitative plaque and whether quantitative plaque impacts clinical decision-making beyond visual assessment.

Figure 3:

(A) Artificial Intelligence–Enabled Quantitative Coronary Plaque Analysis (AI-QCPA; HeartFlow) at CT in a study of 11 808 patients provides age- and sex-stratified percentile nomograms for coronary plaque volumes from quantitative coronary analysis for reference. (B) Distribution of total plaque volume reported in deciles by age and total population. (C) Percentage of calcified plaque and noncalcified plaque volumes by age in the entire population. (D) Prevalence of calcified plaque and noncalcified plaque and of noncalcified plaque without coronary calcifications by age in the entire population. (Adapted, with permission under a CC BY-NC-ND 4.0 license, from reference 8.)

Nurmohamed et al (9) investigated the 10-year prognostic value of atherosclerotic plaque burden derived from artificial intelligence–guided quantitative CCTA analysis as compared with plaque burden manually assessed from CT, CACS, and clinical risk characteristics. In 536 patients, the authors used a quantitative plaque staging system to estimate 10-year risk of atherosclerotic disease. Artificial intelligence–guided quantitative CCTA analysis plaque staging showed important prognostic value for MACE and additional discriminatory value over clinical risk factors, CACS, and manual guideline–recommended CT assessment (Fig 4).

Figure 4:

Through 10-year follow-up, artificial intelligence–guided coronary CT angiography (AI-QCT)–guided plaque staging based on percentage atheroma volume (PAV) (left) showed important prognostic value for atherosclerotic cardiovascular disease (CVD) (middle) and provided important reclassification benefit compared with clinical risk characteristics and coronary artery calcium scoring (CACS) (right). AUC = area under the receiver operating characteristic curve, CAD = coronary artery disease, MACE = major adverse cardiac events, NRI = net reclassification improvement. (Adapted, with permission under a CC BY-NC-ND 4.0 license, from reference 9.)

Coronary Plaques in Focal versus Diffuse Coronary Disease

Sakai et al (10) performed a subanalysis of the Precise Percutaneous Coronary Intervention Plan (or, P3) trial exploring the relationship between CCTA-based atherosclerotic plaque phenotypes in focal and diffuse CAD and their relationship with coronary hemodynamics. A total of 117 participants in five countries were enrolled who underwent invasive FFR measurements showing functionally significant lesions. Further plaque analysis was performed with CT and optical coherence tomography. In this study, the use of motorized pullback pressure gradient, derived from hyperemic pullback pressure curves, quantified the longitudinal distribution of epicardial resistance, therefore providing another dimension to single-point FFR. The main finding was the presence of distinctive plaque features in participants with focal versus diffuse CAD. Focal disease (high pullback pressure gradient) was associated with anatomically more-severe lesions, increased overall and lipid-rich plaque burden, and a higher prevalence of thin-cap fibroatheromas. Calcifications were more common in vessels with diffuse disease and low pullback pressure gradient (Fig 5). Furthermore, translesional pressure gradients correlated inversely with fibrous cap thickness. Percentage atheroma volume was increased in diffuse disease. Percentage atheroma volume and calcifications were considered prognostic of the general burden of disease.

Figure 5:

The Precise Percutaneous Coronary Intervention Plan (P3) trial evaluated the relationship between atherosclerotic plaque phenotypes and focal and diffuse coronary artery disease (CAD) defined by coronary hemodynamics. The study included patients with hemodynamically significant CAD based on fractional flow reserve (FFR), with plaque characterization based on CT and optical coherence tomography (OCT). Using an OCT pullback pressure gradient (PPG) index, patients were divided into focal (PPG > 0.66) or diffuse (PPG ≤ 0.66) CAD (upper panel). Plaque phenotypes were subsequently analyzed at CT and OCT (middle panel). The graph shows vessels with focal CAD (red bars) had a higher plaque burden and predominantly lipid-rich plaque with a high prevalence of high-risk thin-cap fibroatheroma, whereas calcifications were the hallmark of vessels with diffuse disease (green bars). The study highlights the different plaque subtypes in focal versus diffuse disease, which may influence therapeutic pathways. (Adapted, with permission under a CC BY-NC-ND 4.0 license, from reference 10.)

Coronary Artery Calcium and Outcomes

In the Coronary CT Angiography Evaluation for Clinical Outcomes: An International Multicenter, or CONFIRM, registry, Budoff et al (11) showed that atherosclerotic cardiovascular disease (ASCVD) event rates in patients without a history of myocardial infarction or revascularization correlated with event rates in patients with established ASCVD, above a certain CACS threshold. Incidence of MACE was higher in patients with higher CACS, with the highest rates found in patients with CACS greater than 300 and those with prior ASCVD. There was no evidence of differences in all-cause mortality, MACE, MACE with late revascularization, and myocardial infarction event rates between patients with CACS greater than 300 and those with established ASCVD (all P > .05). A CACS less than 300 had substantially lower event rates. Patients with CACS greater than 300 were at an equivalent risk of MACE as those treated for established ASCVD. This interesting observation highlights the need for further study into secondary prevention treatment targets in those without prior ASCVD but with high CACS and reminds us of the utility of CACS for risk stratification.

Pericoronary Fat Attenuation Index

Kuneman et al (12) shed light on the role of pericoronary adipose tissue in suspected CAD by using CT. Patients who developed an acute coronary syndrome within 2 years after CT were identified, and patients with stable CAD (defined as any coronary plaque ≥ 30% luminal diameter stenosis) were 1:2 propensity score matched for age, sex, and cardiac risk factors. The study defined pericoronary adipose tissue at CT as tissue with attenuation values between −190 and −30 HU and within a radial distance from the vessel wall equal to the vessel diameter. The mean pericoronary adipose tissue attenuation was analyzed and found to be significantly increased across culprit lesion precursors in patients with acute coronary syndrome as compared with both nonculprit lesions in the same patients and lesions of patients with stable CAD (Fig 6). Their findings suggest a higher intensity of inflammation in patients with acute coronary syndrome and suggest that pericoronary adipose tissue attenuation at CT may represent a newly identified marker to detect high-risk plaques (HRPs).

Figure 6:

Kuneman et al evaluated the association between pericoronary adipose tissue (PCAT) and culprit plaques by using CT. Patients with acute coronary syndrome (ACS) within 2 years after CT were identified and compared with controls. The PCAT at CT was defined as tissue with attenuation between −190 and −30 HU and within a radial distance from the vessel wall equal to the vessel diameter. (A) Multiplanar reconstructions of CT scans show mixed plaques in the proximal (right) and middle (left) left anterior descending coronary artery, with surrounding PCAT (orange-yellow colored areas) across a precursor of a culprit lesion in a patient who developed an ACS (left) and across a lesion in a patient with stable coronary artery disease (CAD) (right). (B) Mean PCAT attenuation across precursors of culprit lesions versus nonculprit lesions in patients who developed ACS versus lesions in patients with stable CAD. The study highlights the evolving role of inflammatory pericoronary fat evaluation as an additional prognostic imaging biomarker of high-risk coronary plaque in predicting subsequent events. (Adapted, with permission under a CC BY-NC-ND 4.0 license, from reference 12.)

High-risk Coronary Plaques

Patients with HRP features at CT are thought to have a higher risk of future cardiovascular events (13,14). However, individual HRP features at CT imaging have not previously been systematically validated against high-resolution intravascular imaging. Kinoshita et al (15) looked at patients who underwent both CT and optical coherence tomography followed by coronary intervention with 3-year follow-up. HRP was defined as a plaque with at least two of the following features: positive remodeling, low-attenuation plaque, napkin ring sign, and spotty calcification. All four HRP features at CT were associated with features of plaque vulnerability at optical coherence tomography (Fig 7). HRP features at CT showed similar morphology, clinical significance, and association with cardiovascular events compared with features of plaque vulnerability at optical coherence tomography, further highlighting the importance of detecting and reporting HRP at CT (15).

Figure 7:

Kinoshita et al evaluated patients who underwent both CT and optical coherence tomography (OCT) to analyze high-risk plaque features at CT (positive remodeling, low-attenuation plaque, napkin ring sign, and spotty calcification) that correlated with vulnerable OCT plague features. Statistically significant associations between multiple high-risk plaque (HRP) features and each OCT feature of vulnerability are depicted by highlighted dots and lines connecting features. Positive remodeling was the most foundational HRP feature and associated with all six OCT features of plaque vulnerability. The others have more specific associations with individual OCT features of vulnerability. The prevalence of each OCT feature is represented as follows: +++ indicates ≥75%, ++ indicates 50%–74%, + indicates 25%–49%, and – indicates <25%. The prevalence pattern was similar among the four HRP CT features. In the representative images of each HRP feature, positive remodeling is indicated by green arrowheads, low-attenuation plaque by yellow dotted outlines, napkin ring sign by orange arrowheads, and spotty calcification by a white arrow. Yellow arrowheads indicate the labeled representative OCT feature of vulnerability. TCFA = thin-cap fibroatheroma. (Adapted, with permission, from reference 15.)

Updated CT-SYNTAX Score

The anatomic SYNergy between percutaneous coronary intervention with TAXus and Cardiac Surgery, or SYNTAX, score (aSS) was developed originally to quantify the anatomic extent and complexity of CAD at conventional ICA (16). The aSS was then adapted to CCTA in 2012. Given the advancements in CCTA since then, Kageyama et al (17) aimed to update the definitions of the CT-based aSS (hereafter, CT-aSS) in a study of patients with complex CAD. The modifications included an updated definition and detection of total occlusion at CT, based on (a) lesion length, (b) inclusion of scoring for serial bifurcations in one single coronary segment, (c) inclusion of score points for lesions located distal to a total occlusion, not visualized at ICA but visible at CT, and (d) removal of thrombus and bridging collateral vessels. The aim of these modifications in the CT-aSS was to reinforce the prognostic value of the score, with the hope that integration with artificial intelligence would promote its use in clinical practice. The updated CT-aSS showed comparable interobserver variability to the original CT-aSS, both with excellent interreader agreement (κ = 0.82 vs κ = 0.80) between two CT specialist readers. Serial lesions detected distally to a total occlusion and with stenosis of 50% or greater involving serial bifurcations in the same coronary segment are the major modifications in the updated CT-aSS. Increasingly, CT rather than ICA is being used to assess the appropriateness of revascularization. Patient prognosis after percutaneous coronary intervention or coronary artery bypass grafting can be predicted using the ICA-based aSS, and this prognostic capability might be extended to CT-aSS.

Studies Investigating CT Strategies for Stable Chest Pain

Noninvasive imaging investigations for CAD are increasing, and in 2016, the National Institute for Health and Care Excellence Clinical Guideline Number 95 (“Chest pain of recent onset”) (CG95) recommended CT as the first-line test for possible angina. Weir-McCall et al (18) reviewed cardiac imaging investigations performed between 2012 and 2018 that were linked to hospital admission and mortality registries to capture changes across the United Kingdom since the CG95 recommendation. Imaging investigations increased overall, with greater regional increases in CT associated with fewer hospitalizations for myocardial infarction and faster decline in CAD mortality (18). The data showed consistent and sustained increase in the use of imaging investigations to diagnose and manage CAD. There were considerable changes to specific individual imaging modalities over time; particularly, increased CT use and reduced ICA use were observed. While there was an increased cost associated with the overall increase in imaging, this remained on par with inflation.

Machado et al (19) completed a systematic review and meta-analysis comparing CT first versus direct ICA strategies in patients with stable chest pain initially referred to ICA. Five randomized controlled trials were found comparing the two strategies, with a total of 5727 participants. The main findings were that CT was associated with a 26% reduction in coronary revascularizations and, despite the lower incidence of revascularizations, the incidence of cardiovascular mortality, MACE, cardiovascular hospitalizations, and acute myocardial infarction were similar in both groups. CT was also associated with a reduced risk of stroke when compared with direct ICA.

Cardiomyopathies

Cardiac Amyloid

Several publications in 2023 evaluated cardiac amyloidosis, focusing on identification of new amyloid subtypes and emerging treatment strategies. While transthyretin cardiac amyloidosis (ATTR) typically shows characteristic imaging patterns, rarer variants demonstrate more variable and challenging findings. Ioannou et al (20) performed a retrospective analysis of patients with a confirmed diagnosis of immunoglobulin light chain amyloidosis, ATTR, apolipoprotein A-I amyloidosis (AApoAI), and apolipoprotein A-IV amyloidosis (AApoAIV) at the National Amyloidosis Center database in the United Kingdom over a 21-year period. Following extensive characterization with multimodality imaging, the authors identified 4364 patients with light chain amyloidosis, 2057 patients with ATTR, 45 patients with AApoAI, and 21 patients with AApoAIV; cardiac involvement was found in 62.6%, 96.5%, 28.9%, and 71.4% of patients, respectively. AApoAI was commonly associated with multisystem disease, with cardiac involvement manifesting with heart failure (Fig 8). AApoAI was commonly associated with right-sided involvement, right ventricular thickening, tricuspid valve dysfunction, right atrial and right ventricular radiotracer uptake, and right atrial and right ventricular late gadolinium enhancement (LGE). AApoAIV commonly manifested as classic cardiac amyloidosis (eg, subendocardial or transmural LGE). Patients with AApoAI and AApoAIV had a good prognosis and were found to have a lower risk of mortality compared with matched patients with light chain amyloidosis. Suspicion of rarer forms of cardiac amyloidosis should be prompted in patients with either characteristic imaging features of cardiac amyloidosis or right-sided cardiac disease at cardiac MRI, in combination with cardiac uptake at bone scintigraphy and the absence of biochemical evidence of a plasma cell dyscrasia. The low-grade or absent cardiac radiotracer uptake in AApoAI and AApoAIV was not proportional to the degree of cardiac infiltration; therefore, stand-alone bone scintigraphy cannot be used to diagnose these rarer forms of cardiac amyloid.

Figure 8:

Ioannou et al performed a retrospective analysis of patients with a confirmed diagnosis of immunoglobulin light chain amyloidosis (AL), transthyretin amyloidosis (ATTR), apolipoprotein A-I amyloidosis (AApoAI), and apolipoprotein A-IV amyloidosis (AApoAIV) at the National Amyloidosis Center database in the United Kingdom over a 21-year period. Diagram illustrates key clinical and imaging features that should raise the suspicion of the different forms of cardiac amyloidosis. Top left, AApoAI cardiac amyloidosis can manifest with laryngeal involvement, multiorgan involvement, and a strong family history. Echocardiograms demonstrate right-sided disease with thickening of the tricuspid valve and tricuspid regurgitation. Cardiac MR image demonstrates right atrial and right ventricular thickening and right atrial and right ventricular late gadolinium enhancement (LGE). Top right, AApoAIV cardiac amyloidosis has a male predominance and can manifest with renal involvement. Echocardiograms demonstrate biventricular wall thickening and a typical apical-sparing strain pattern. Cardiac MR image demonstrates left ventricular wall thickening, biventricular transmural LGE, and an elevated extracellular volume (ECV). Bottom left, immunoglobulin AL can manifest with macroglossia, multisystem involvement, and nephrotic syndrome. Echocardiograms demonstrate biventricular wall thickening. Cardiac MR image demonstrates diffuse biventricular transmural LGE and an elevated ECV. Bottom right, transthyretin (ATTR) cardiac amyloidosis has a male predominance and can manifest with polyneuropathy and a strong family history. Echocardiograms demonstrate biventricular wall thickening. Cardiac MR image demonstrates diffuse biventricular transmural LGE and an elevated ECV. hATTR = hereditary ATTR. (Adapted, with permission under a CC BY-NC-ND 4.0 license, from reference 20.)

An interesting method to diagnose cardiac amyloidosis at cardiac MRI was explored by Kidoh et al (21) in 136 patients with cardiac amyloidosis. Diagnostic performances of the traditional MRI extracellular volume method and the newer myocardium to lumen R1 ratio on postcontrast T1 maps (a simplified index not requiring a native T1 map and hematocrit level) were compared (Fig 9). There was no evidence of a difference in diagnostic performance for the new, simpler measure in diagnosing cardiac amyloidosis, and this more pragmatic approach has the potential to become incorporated into routine clinical use for cardiac MRI investigation of amyloidosis if corroborated in further larger trials.

Figure 9:

Kidoh et al explored the diagnostic performances of the traditional MRI ECV method and the newer method of myocardium to lumen R1 ratio on postcontrast T1 maps, which has the advantage of not requiring a native T1 map and hematocrit levels. (A) Example of region of interest placement on postcontrast T1 maps in a patient with cardiac amyloidosis. The ratio of myocardial R1 to luminal R1 on the postcontrast T1 map was defined as postcontrast myocardium to lumen R1 ratio. (B) Postcontrast T1 map in a 70-year-old male individual with suspected cardiac amyloidosis. The mean R1 values of the septal wall were higher than those of the lumen (postcontrast myocardium to lumen R1 ratio = 1.36). The ECV value was 78% (abnormally high). Wild-type transthyretin amyloidosis (ATTRwt) was confirmed using genetic testing. (C) Receiver operating characteristic curves for the detection of patients with cardiac amyloidosis. The highest area under the receiver operating characteristic curve (AUC) was attained with ECV (0.99 [95% CI: 0.97, 1.00], P < .001), followed by Λ (0.98 [95% CI: 0.96, 0.99], P < .001), and then, postcontrast myocardium-to-lumen R1 ratio (0.98 [95% CI: 0.95, 0.99], P < .001). There was no evidence of a difference in AUC between ECV and postcontrast myocardium to lumen R1 ratio (P = .10) or between Λ and postcontrast myocardium to lumen R1 ratio (P = .19). ECV = extracellular volume fraction. (Adapted, with permission, from reference 21.)

Ischemic Cardiomyopathies and Stress Perfusion Imaging

Although cardiovascular disease remains the leading cause of mortality in female individuals, recognition of CAD is often delayed or deferred because of atypical clinical presentation and lack of adherence to management guidelines. Stress perfusion cardiac MRI has become a mainstream diagnostic tool for the evaluation of suspected or known cardiovascular disease in many centers, owing to its safety, high spatial and temporal resolutions, and freedom from radiation or iodinated contrast media. In a post hoc analysis of the Stress CMR Perfusion Imaging in the United States (SPINS) trial, Heydari et al (22) sought to evaluate the sex-specific diagnostic quality, prognostic value, and cost-effectiveness of stress perfusion cardiac MRI for patients referred for suspected stable ischemic heart disease. The primary outcome measure was a composite of cardiovascular death and nonfatal myocardial infarction; secondary outcomes were hospitalization for unstable angina or heart failure and late unplanned coronary artery bypass grafting. The SPINS trial included 1104 female individuals (47% of the cohort). Female individuals had a higher prevalence of chest pain (62% vs 50%, P < .0001) but lower use of OMT compared with male individuals. At the 5.4-year median follow-up, female individuals with normal stress cardiac MRI findings had a low annualized rate of primary composite outcomes (0.54%/year), which was similar to male individuals. Female individuals with abnormal cardiac MRI findings at follow-up were at higher risk for both primary (3.74%/year vs 0.54%/year, P < .0001) and secondary (9.8%/year vs 1.6%/year, P < .0001) outcomes compared with female individuals with normal cardiac MRI findings. Multivariable analysis showed that abnormal stress cardiac MRI findings independently predicted the primary (hazard ratio: 2.64 [95% CI: 1.20, 5.90], P < .02) and secondary (hazard ratio: 2.09 [95% CI: 1.43, 3.08], P < .0001) outcome measures (Fig 10). The authors concluded that stress cardiac MRI should be considered as a first-line noninvasive imaging tool for the evaluation of female individuals with chest pain, corroborating recent guidelines (23).

Figure 10:

Heydari et al sought to evaluate sex-specific prognostic performance in a multicenter Stress CMR Perfusion Imaging in the United States [SPINS] Study registry of 2349 patients. The primary outcome measure was a composite of cardiovascular death and nonfatal myocardial infarction. At the 5.4-year median follow-up, female individuals with normal stress cardiac MRI findings had a low annualized rate of primary composite outcome similar to male individuals (P value was nonsignificant). In contrast, female individuals with abnormal cardiac MRI findings were at higher risk for the primary outcome compared with female individuals with normal cardiac MRI findings. Presence of abnormal stress cardiac MRI findings was an independent predictor for the primary outcome measure. There was no effect modification for sex. Female individuals had lower rates of invasive coronary angiography and downstream costs at 90 days following cardiac MRI. There was no effect of sex on diagnostic image quality. (Reprinted, with permission, from reference 22.)

Stress cardiac MRI typically involves rest and stress perfusion sequences using a short-acting vasodilator. Removing rest-perfusion from the imaging protocol can save at least 5 to 7 minutes of scan time. Swoboda et al (24) hypothesized that when interpreting stress perfusion cardiac MRI, stress-only analysis has equivalent diagnostic accuracy to the stress-rest analysis. The authors also aimed to identify the optimum threshold of infarct transmurality at which segments with both infarction and stress-induced subendocardial hypoperfusion should be considered ischemic. A total of 666 patients from the Clinical Evaluation of MAgnetic Resonance imaging in Coronary heart disease (or, CE-MARC) trial were included. Each patient underwent a complete cardiac MRI examination with stress perfusion, rest perfusion, LGE, and quantitative ICA. The optimal stress-LGE analysis classified all segments with a stress perfusion defect as ischemic unless they had greater than 75% infarct transmurality (area under the receiver operating characteristic curve, 0.843; sensitivity, 75.6%; specificity, 93.1%). This method had higher diagnostic accuracy when compared with the stress-rest method (area under the receiver operating characteristic curve, 0.834; sensitivity, 73.6%; specificity, 93.1%; P value for difference = .02). The occurrence of MACE was assessed over a median follow-up period of 6.5 years. The presence of inducible ischemia by stress-LGE or stress-rest analysis was similar, and both were strong predictors of MACE (hazard ratio: 2.65; P < .001, for both). The authors concluded that rest-perfusion imaging can safely be removed from cardiac MRI protocols and that segments without transmural infarction should be considered ischemic.

Dilated Cardiomyopathy

For some nonischemic dilated cardiomyopathies, LGE is localized only at the right ventricular insertion points (IP-LGE) (Fig 11). IP-LGE is generally considered a benign finding; however, there are only limited data from large dilated cardiomyopathy cohorts about the prognostic effect of IP-LGE. Claver et al (25) sought to evaluate the baseline characteristics and prognostic implications of LGE limited to the right ventricular insertion points, as well as LGE present at both the right ventricular insertion points and left ventricle. In a retrospective observational multicenter cohort study involving 1165 patients with dilated cardiomyopathy, the primary end point included appropriate defibrillator therapies, sustained ventricular tachycardia, resuscitated cardiac arrest, or sudden death. During a median follow-up of 36 months, 74 patients (6%) reached the primary end point. The first arrhythmic event was appropriate implantable cardioverter defibrillator (ICD) therapy in 33 patients, sustained monomorphic ventricular tachycardia before any ICD implantation in 26 patients, resuscitated cardiac arrest in eight patients, and sudden death in seven patients. The authors concluded that patients with IP-LGE have baseline characteristics and prognosis similar to those of LGE-negative patients. In contrast, patients with IP-LGE have milder left ventricular dysfunction and significantly lower risk of ventricular arrhythmias or sudden death when compared with LGE-positive patients. Among LGE-positive patients, the presence of IP-LGE at right ventricular insertion points did not identify a subgroup of patients with different clinical characteristics or different prognosis.

Figure 11:

Some nonischemic dilated cardiomyopathies (DCMs) have late gadolinium enhancement (LGE) localized only at the right ventricular (RV) insertion points (IP-LGE). Claver et al sought to evaluate the prognostic implications of IP-LGE, as well as LGE present at both the RV insertion points and left ventricle, in 1165 patients with DCM. Representative cardiac MR images in a patient with IP-LGE: (A) cine left ventricular short-axis view, (B) LGE magnitude, and (C) phase-sensitive inversion recovery, all synchronized at the same phase of the cardiac cycle. The LGE sequences show a prominent area of inferior IP-LGE. The study found that insertion point LGE did not affect outcomes among patients with DCM. (Adapted, with permission, from reference 25.)

Sudden Cardiac Death in Dilated Cardiomyopathy

ICD therapy is established as the most effective prophylactic strategy adopted for primary and secondary prevention of sudden cardiac death in dilated cardiomyopathy. New prognostic strategies are needed to improve the delivery of ICD while withholding device implantation in those at low risk of sudden cardiac death. The CarDiac MagnEtic Resonance for Primary Prevention Implantable CardioVerter DefibrillAtor ThErapy (DERIVATE-ICM) registry aims to evaluate the net reclassification improvement for the indication of ICD implantation using cardiac MRI as compared with standard of care based on transthoracic echocardiography left ventricular ejection fraction (LVEF) evaluation in consecutive patients with dilated cardiomyopathy from several centers across Europe and the United States. Pontone et al (26) evaluated 861 patients from the registry with dilated cardiomyopathy with chronic heart failure and LVEF of less than 50% at transthoracic echocardiography. During a median follow-up of 1054 days, MACE occurred in 88 patients (10.2%). Left ventricular end-diastolic volume index (hazard ratio: 1.007 [95% CI: 1.000, 1.011], P = .05), cardiac MRI–derived LVEF (hazard ratio: 0.972 [95% CI: 0.945, 0.999], P = .045), and LGE (hazard ratio: 1.010 [95% CI: 1.002, 1.018], P = .015) were found to be independent predictors of MACE. A multiparametric cardiac MRI–weighted predictive derived score better identified patients at high risk for MACE compared with a transthoracic echocardiography LVEF cutoff of 35%, with a net reclassification improvement of 31.7% (P = .007). By comparing the standard of care evaluation with a cardiac MRI–guided strategy for ICD therapy, the DERIVATE-ICM registry demonstrates that LGE can provide better stratification of arrhythmic risk, especially in patients who do not meet implantation criteria.

Hypertrophic Cardiomyopathy and Obstruction

Apical left ventricular aneurysms in hypertrophic cardiomyopathy (HCM) are associated with adverse outcomes. Studies have shown that aneurysms are common in patients with apical mid HCM. It is unclear however, why some patients are predisposed to aneurysm formation. Sherrid et al (27) focused on the leading hypotheses, namely mid left ventricular obstruction, and sought to ascertain its frequency in apical mid HCM. The authors analyzed echocardiographic and cardiac MRI examinations of patients with aneurysms from three dedicated programs and compared them with 63 healthy controls and 47 controls with apical mid HCM who did not have aneurysms. Of 108 with apical aneurysms included in the study, 103 patients (95%) had mid left ventricular obstruction with mid left ventricular complete systolic emptying (Fig 12).

Figure 12:

Sherrid et al evaluated the imaging and physiologic features of mid left ventricle (LV) obstruction and its frequency in hypertrophic obstructive cardiomyopathy apical aneurysms. The upper panel shows a hypertrophied LV from end diastole to end systole, obstructing the papillary muscle (PM) (red arrow). Relatively apical insertion of the PM causes obstruction at the apical-to-mid level, with a resulting small aneurysm (white arrow). The subvalvular apparatus (SVA) is long (orange double arrow). The schematic on the left side of the lower panel shows mid-LV obstruction, short SVA (double arrow), and large apical aneurysm with hypertrophic involvement of the PM. The right-hand schematic shows apical-mid obstruction, long SVA (double arrow), and a small apical aneurysm. (Reprinted, with permission, from reference 27.)

Among patients with obstruction, 84% exhibited a midsystolic Doppler signal void, indicating complete flow cessation, while only 19% showed Doppler systolic gradients of 30 mm Hg or greater. Five patients (5%) had relative hypokinesia in the mid left ventricle without obstruction. Compared with all control groups, patients with mid left ventricular obstruction aneurysms had smaller short-axis systolic areas (P < .007), greater percentage change in short-axis area (P < .005), larger papillary muscle areas (P < .003), and higher ratios of diastolic papillary muscle areas to short-axis diastolic areas (P < .005). Those with aneurysms had 22% larger short-axis papillary muscle areas than patients with elevated left ventricle velocities but no aneurysms (median: 3.00 cm² [IQR: 2.38–3.70 cm²] vs 2.45 cm² [IQR: 1.81–2.95 cm²], P = .004). The authors suggest that mid left ventricular obstruction may serve as a potential therapeutic target to prevent or delay subsequent aneurysm formation, but further studies are needed to validate these findings.

LGE confers an increased risk of MACE in patients with HCM. Although the predominant pattern of LGE in HCM is intramural, multiple studies have documented subendocardial LGE. However, the clinical significance of subendocardial LGE remains underappreciated. In addition, LGE involving the right ventricular insertion point is commonly described, with little known of its prognostic value. In a single-center retrospective study, Yang et al (28) examined 497 consecutive patients with HCM with LGE confirmed with cardiac MRI. Subendocardium-involved LGE was defined as LGE involving subendocardium not corresponding to a coronary vascular distribution. Patients with ischemic heart disease that might contribute to subendocardial LGE were excluded. End points included a composite of heart failure–related events, arrhythmic events, and stroke.

Among the 497 patients, subendocardium-involved LGE was observed in 184 (37.0%), and right ventricular insertion point LGE was present in 414 (83.3%). Extensive LGE (≥15% of left ventricular mass) was found in 135 patients. Over a median follow-up of 57.9 months, 66 patients (13.3%) reached composite end points. Those with extensive LGE had a significantly higher annual incidence of adverse events (5.1% vs 1.9% per year, P < .001; see Fig 13). In patients with extensive LGE, the risk of composite end points increased with greater LGE extent, while this trend was not seen in patients with nonextensive LGE (<15%). For those with extensive LGE, the extent of LGE significantly correlated with composite end points (hazard ratio: 1.05, P = .03), even after adjusting for factors like LVEF less than 50%, atrial fibrillation, and ventricular tachycardia. In patients with nonextensive LGE, subendocardium-involved LGE, rather than LGE extent, was independently linked to adverse outcomes (hazard ratio: 2.12, P = .03). Right ventricular insertion point LGE was not significantly associated with poor outcomes.

Figure 13:

The predominant pattern of late gadolinium enhancement (LGE) in HCM is intramural, but other patterns such as subendocardial and right ventricular insertion point (RVIP) can be seen. Yang et al examined 497 consecutive patients with HCM for subendocardial and RVIP LGE patterns by using cardiac MRI. (A) A four-chamber LGE sequence (left) shows subendocardial enhancement along the interventricular septum yet no obstructive coronary lesion at invasive coronary angiography (right). (B) A short-axis LGE sequence with RVIP LGE. Subendocardium-involved LGE and RVIP LGE were observed in 184 (37.0%) and 414 (83.3%), respectively, of the cohort. Red circle: endocardial border, green circle: epicardial border, yellow patches: LGE, blue circle: normal reference myocardium. (C) Kaplan-Meier curves show event-free rates for various combinations of LGE extent and subendocardial pattern. In patients with nonextensive LGE, subendocardium-involved LGE rather than LGE extent was independently associated with adverse outcomes (HR: 2.12, P = .03). HCM = hypertrophic cardiomyopathy, HR = hazard ratio. (Adapted, with permission, from reference 28.)

The authors conclude that in patients with HCM and extensive LGE (≥15%), the risk of cardiovascular events increases with percentage increase in LGE extent. In patients with HCM and nonextensive LGE (<15%), the subendocardium-involved LGE pattern, an underrecognized phenotype in HCM, is associated with adverse outcomes, especially for heart failure–related events. Finally, while right ventricular insertion point LGE is commonly observed in HCM, it does not appear to have prognostic value.

Myocarditis

COVID-19.— Although the pulmonary disease associated with SARS-CoV-2 has been well documented, prior studies have described heterogeneous cardiac manifestations of the virus, including myocarditis, myocardial infarction, and cardiogenic shock. Vidula et al (29) retrospectively analyzed 1047 patients from 18 international sites with confirmed COVID-19 infection who underwent cardiac MRI. LGE was present in 403 patients (38.8%), most commonly with midmyocardial (22.4%) and subepicardial (10.0%) patterns (Fig 14). Chest discomfort, abnormal electrocardiogram, elevated natriuretic peptide level, and elevated troponin level were associated with acute myocarditis patterns at univariable analysis. Variables associated with acute ischemic patterns included chest discomfort, abnormal electrocardiogram, known coronary disease, hospitalization, elevated natriuretic peptide level, and elevated troponin level. At multivariable analysis, troponin level elevation remained significantly associated with acute myocarditis patterns (odds ratio: 4.98 [95% CI: 1.76, 14.05], P = .003). The article highlights the various LGE patterns, including infarct patterns, that can be ascribed to COVID-19.

Figure 14:

Vidula et al retrospectively analyzed 1047 patients, from 18 international sites, with confirmed COVID-19 infection who underwent cardiac MRI. (A) Example cardiac MR images in a patient with acute myocarditis linked to a recent COVID-19 infection show subepicardial late gadolinium enhancement (LGE) in the midinferior and inferoseptal walls (left image, indicated by white arrow). There are also elevated native T2 times (middle image, white arrow) and native T1 times (right image, white arrow) in the midinferior wall. (B) Example cardiac MR images in a patient after a complicated COVID-19 hospitalization reveal nonacute, nonischemic injury. There is extensive midmyocardial LGE in the septum (left image, white arrow) and subepicardial LGE in the inferior and inferolateral walls visible on the short-axis view (left image, red arrow). The midmyocardial LGE in the septum is also seen on the four-chamber view (middle image, white arrow), and subepicardial LGE in the inferior wall appears on the two-chamber view (right image, white arrow). (Reprinted, with permission, from reference 29.)

New Insights from Cardiac MRI into Valvular Disease

Specific global longitudinal strain patterns have been previously described for both diagnostic and prognostic purposes in patients with severe aortic stenosis. Specifically, relative apical sparing pattern (RASP) has been proposed as a specific marker of aortic stenosis, potentially differentiating it from cardiac amyloidosis. Abecasis et al (30) assessed the prevalence of RASP in a group of patients with severe symptomatic aortic stenosis referred for surgical aortic valve replacement. The authors sought to evaluate the clinical significance of RASP, its possible relation to amyloid deposition, and persistence after surgery. In this prospective study, 150 participants with severe symptomatic aortic stenosis referred for surgical repair underwent cardiac MRI. RASP was defined as average apical longitudinal strain/(average basal longitudinal strain + average mid longitudinal strain) greater than 1 as determined with echocardiography. Aortic valve replacement was performed in 119 participants. RASP was identified in 23 (15.3%) of these participants. There was no suspicion of amyloidosis at preoperative cardiac MRI (native T1 value, 1053 msec [IQR, 1025–1076 msec]; extracellular volume, 28% [IQR, 25%–30%]), and no participant had amyloid deposition at histopathology. Compared with participants without RASP, those with RASP had significantly higher transvalvular gradients, despite no differences in flow, and higher preoperative left ventricular mass and positive remodeling as assessed with both echocardiography and cardiac MRI. The RASP group had lower LVEF at cardiac MRI evaluation, with higher left atrial volumes and significantly more prevalent impairment of right ventricular free wall longitudinal strain. At tissue characterization, this strain pattern was associated with higher absolute left ventricular LGE mass. The authors concluded that RASP does not necessarily indicate cardiac amyloidosis. RASP may also be found in severe symptomatic aortic stenosis without amyloidosis, representing advanced left ventricle disease.

Societal Statements

The European Association of Cardiovascular Imaging, in collaboration with the European Heart Rhythm Association, provided a two-part, highly comprehensive, up-to-date, and evidence-based clinical consensus statement on the use of imaging in patients undergoing cardiac device implantation (31,32). A combination of echocardiography, cardiac MRI, cardiac CT, and cardiac PET is recommended for predevice imaging. The statement covers several devices, including conventional pacemakers, implantable defibrillators, and resynchronization devices. It emphasizes the central role imaging plays in assessing the need for cardiovascular implantable cardiac devices. For example, conventional guidelines proposed an LVEF threshold for implanting cardioverter defibrillators. The clinical consensus statement allows for nuances in this assessment, recommending assessment of scar burden in patients with HCM and preserved LVEF, for example (31). The consensus statement also examines the need for preprocedural imaging in patients in whom device implantation is considered difficult. Patients with complex congenital heart disease, anatomic variants, or acquired valvulopathies often require multimodality imaging prior to implantation to avoid lead misplacement or entrapment and paradoxical embolism (31). Part two of the statement focuses on post–device implantation imaging, including assessment of postprocedural complications (eg, pneumothorax, cardiac perforation, pacemaker syndrome) and the most appropriate imaging modality to employ based on clinical deterioration. Safety of cardiac MRI after device implantation, as well as a stepwise implantable device identification algorithm to allow device identification based on chest radiography, are also discussed (32).

The 2023 European Society of Cardiology Guidelines for the management of cardiomyopathies has called for increasing use of cardiac MRI, particularly in the diagnosis of nondilated left ventricular cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy, myocarditis, and hemochromatosis (33). Previous limitations to cardiac MRI imposed by implantable cardiac devices have been obviated by the use of specialized sequences, reducing the rate of uninterpretable studies (33).

The American Heart Association compiled a scientific statement in contemporary cardio-oncology. The statement aimed to outline current evidence for cardiovascular imaging in cancer treatment and survivorship and to suggest new methods for optimal application of cardiovascular imaging in future clinical trials and registries (34).

Recommendations have been provided by the Society for Cardiovascular Magnetic Resonance on the use of cardiac MRI in patients suspected of having or after COVID-19 cardiac involvement (35). The article details the various cardiac manifestations of COVID-19 and provides recommendations for cardiac protocols to assess COVID-19, as well as recommendations for clinical reporting of cardiac MRI findings.

Finally, an expert consensus document on cardiac CT nomenclature has been updated by multiple imaging societies, yielding a standardized set of medical terms commonly used in clinical and research activities related to cardiac CT (36).

Conclusion

This review has attempted to highlight the most noteworthy and practice-changing cardiovascular imaging research from 2023 across a range of radiologic modalities of interest to both radiologists and referring clinicians. It is our hope that the findings from these studies will encourage further research in the field and ultimately help improve patient diagnoses, treatments, and outcomes.