HomeRadiology: Cardiothoracic ImagingVol. 5, No. 3 Previous CommentaryFree AccessHistopathologic Validation of Stress T1 Mapping in Myocardial Ischemia: Another Step toward Clinical Translation?Matthew K. Burrage , Vanessa M. FerreiraMatthew K. Burrage , Vanessa M. FerreiraAuthor AffiliationsFrom the University of Oxford Centre for Clinical Magnetic Resonance Research, Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, England (M.K.B., V.M.F.); and Faculty of Medicine, University of Queensland, Brisbane, Australia (M.K.B.).Address correspondence to M.K.B. (email: [email protected]).Matthew K. Burrage Vanessa M. FerreiraPublished Online:Jun 22 2023MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In See also article by Zhuang et al in this issue.Matthew K. Burrage, MD, PhD, is an advanced cardiovascular imaging and heart failure cardiologist and senior lecturer at the University of Queensland in Brisbane, Australia. He completed his DPhil at the University of Oxford on stress T1 mapping and novel cardiovascular MRI applications for the assessment of cardiovascular disease.Download as PowerPointOpen in Image Viewer Vanessa M. Ferreira, MD, PhD, is a British Heart Foundation associate professor of cardiovascular medicine, deputy director of the Oxford Clinical Centre for Magnetic Resonance Research (OCMR), and honorary consultant cardiologist in the Division of Cardiovascular Medicine, Radcliffe Department of Medicine at the University of Oxford. Her research focuses on cardiovascular MRI (CMR) myocardial tissue characterization, clinical translation of myocardial T1 mapping, and developing contrast material–free CMR technologies.Download as PowerPointOpen in Image Viewer Coronary artery disease (CAD) remains a leading cause of morbidity and mortality worldwide. Clinical manifestations typically result from dysregulation of coronary physiology in the presence of flow-limiting epicardial stenosis. Compensatory mechanisms in the downstream coronary microvasculature maintain myocardial perfusion by ensuring adequate myocardial blood flow (MBF) and volume via increased microvascular dilatation and capillary recruitment. However, these compensatory mechanisms may become depleted with progressive stenosis and overwhelmed during times of physiologic or pharmacologic stress, with subsequent loss of coronary vasodilatory reserve; this results in a diminished ability to adequately increase myocardial perfusion, oxygen supply-demand mismatch, and signs and symptoms of myocardial ischemia.There is increasing interest in using noninvasive imaging tests to identify impaired myocardial perfusion reserve in patients at risk; this may allow risk stratification, prognostication, and early therapeutic intervention to improve individual outcomes. Although first-pass perfusion at cardiovascular MRI (CMR) can directly assess global and focal reductions in MBF during vasodilator stress and rest, MBF alone may not reflect all aspects of myocardial ischemia. Myocardial blood volume (MBV), on the other hand, represents the total volume of capacitance vessels in both the coronary macro- and microcirculations and closely reflects the level of microvascular autoregulation; MBV may better reflect changes in myocardial oxygen consumption and may be a more comprehensive and specific marker of cellular and myocardial ischemia (1). Additionally, while stress perfusion CMR is well validated for detecting obstructive CAD, it requires an intravenous gadolinium-based contrast agent, which may not be suitable for all patients. There is significant interest in developing contrast material–free CMR techniques to assess cardiovascular disease (2,3).Stress T1 mapping is a contrast agent–free CMR technique that can assess coronary vasodilatory reserve in ischemic and nonischemic cardiovascular disease (4,5). The basic principle is that T1 values are sensitive to increased tissue free water content, including increases in MBF and MBV. These changes are detectable via T1 mapping, given that blood T1 increases the measured myocardial T1 via partial volume effects (6). The normal myocardium at rest is supplied by the epicardial coronary arteries and downstream microvasculature. Vasodilator stress causes coronary vasodilation, with at least a twofold rise in MBV typically seen in healthy individuals with normal vasodilatory reserve (6). This normal increase in MBF, MBV, and water content appears to be associated with a 6%–8% rise in myocardial T1 during vasodilatory stress (termed stress T1 reactivity, or ΔT1) (5–7). In patients with obstructive CAD, a significant coronary artery stenosis induces compensatory vasodilation of the downstream microvasculature, with increased resting MBV and impaired coronary vasodilatory reserve; the ability for further coronary vasodilation during stress conditions will be significantly diminished or abolished, translating to blunted or absent stress T1 reactivity (5).Stress T1 mapping has shown feasibility to distinguish between normal, ischemic, infarcted, and remote myocardium, without the need for gadolinium-based contrast agents. These four classes of myocardium appear to have distinctive rest and stress T1 profiles that allow their differentiation (5,6). Molecular mechanisms of adenosine stress T1 mapping have been investigated in mice, showing that adenosine ΔT1 appears to be mediated through the A2A (endothelial-dependent) and A2B (endothelial-independent) receptors (8). However, few histopathologic validations of this technique have been performed.In this issue of Radiology: Cardiothoracic Imaging, Zhuang and colleagues (9) add to the current literature by performing a miniature-swine histopathologic validation study of stress T1 mapping for the assessment of myocardial ischemia. Ten adult male Chinese miniature swine with progressive, mechanically induced left anterior descending coronary artery stenosis underwent rest and adenosine triphosphate stress T1 mapping, alongside first-pass perfusion and late gadolinium enhancement (LGE) imaging. Findings were compared with two healthy control swine who underwent the same CMR protocol. Infarcted myocardium was defined at histopathologic examination (>50% transmurality), while ischemic myocardium was defined according to regions of visible hypoperfusion on stress first-pass perfusion CMR images in areas downstream of a significant obstructive coronary stenosis but without pathologic evidence of infarction.Results were striking, with clear differences in rest and stress tissue characteristics, allowing differentiation of ischemic, infarcted, and remote unaffected myocardium. Diagnostic performance of stress T1 reactivity, combined with rest T1, was excellent in detecting ischemic (area under the receiver operating characteristic curve [AUC], 0.89) and infarcted (AUC, 0.97) myocardium, with high reproducibility. Significant correlations were seen between histopathologically derived collagen volume fractions and stress T1 reactivity and other tissue characteristics. This represents the first histopathologic validation of stress T1 mapping to assess myocardial ischemia, and the authors are to be commended for their work.However, a few important caveats must be considered before stress T1 mapping can be considered a contrast material–free panacea for imaging myocardial ischemia. First, T1 mapping sequence selection is critically important, with methods that are less sensitive to heart rate effects being highly desirable (6). T1 mapping sequences, such as the modified Look-Locker inversion recovery (MOLLI) 5(3)3 sequence, tend to underestimate T1 values at higher heart rates, as there may be insufficient time for longitudinal magnetization to sufficiently recover between heartbeats. This results in underestimation of stress T1 values during fast heart rates, and thus, seemingly impaired stress T1 reactivity, which could lead to misdiagnosis. The MOLLI 5s(3s)3s sequence, which is counted in seconds rather than heartbeats, as used here, gives greater time for magnetization recovery and would therefore seem better suited for stress T1 mapping applications; however, for human applications, this may be limited by the longer breath holds and resultant breathing motion artifacts, which can negatively impact the quality and diagnostic accuracy of stress T1 maps.The shortened MOLLI T1 mapping sequence offers several advantages for stress T1 mapping in human patients. First, it has significantly shorter breath holds (approximately 5 seconds per T1 map acquisition at a heart rate >100 beats per minute, compared with 12–15 seconds for MOLLI 5s(3s)3s); this makes it significantly easier for human patients to breath hold during hemodynamic stress and provides a higher yield of good quality T1 maps. Shortened MOLLI has also virtually eliminated heart rate dependency via its conditional reconstruction algorithm and rejection of data acquisitions where magnetization has not sufficiently recovered; this is particularly advantageous for stress applications, where heart rate increases significantly. Shortened MOLLI has also been shown to have less intraindividual variability than MOLLI variants, with a significantly greater stress T1 response and effect size (7). Further general technical considerations include the set trigger delay for acquisition; systolic-phase acquisition, with a trigger delay of 0 msec, can mitigate the risk of mistriggering during fast heart rates (7).Second, although touted as a contrast material–free approach, human stress T1 mapping validation studies to date still rely on contrast material–based first-pass perfusion and LGE CMR images as reference to identify areas of hypoperfusion and infarction. Without using LGE or perfusion images as reference, a blinded segmental T1 map analysis approach may be less sensitive to detect the T1 changes; this may be due to factors such as “dilution” of the ischemic subendocardium by the less-affected subepicardium within a myocardial segment or risk of blood pool contamination next to areas of subendocardial ischemia. A more direct way of visualization, similar to that used in clinical practice with pixelwise perfusion maps, would be highly beneficial for stress T1 mapping.Third, not all cases of impaired stress T1 reactivity signify obstructive CAD. Coronary microvascular dysfunction (CMD) is also an important consideration, which may be seen in patients without obstructive CAD and in cardiac conditions such as severe aortic stenosis, characterized by depleted coronary vasodilatory reserve not due to CAD (4). This is likely more prevalent in real-world practice in comorbid human populations with cardiovascular risk factors, as compared with the pure ischemia model used in this animal study.Finally, although histopathologic validation of stress T1 mapping was performed in this study by assessing myocardial fibrosis and the collagen volume fraction, it would have been informative to further validate the pathophysiologic and compensatory changes in the microvasculature downstream of ischemic and remote myocardium. Assessment of microvascular density from histopathology to determine the degree of capillary recruitment or rarefaction and inspection of the microvascular architecture could have provided additional insights into how these may contribute to the imaging changes seen in stress T1 reactivity and MBF in CAD. Blunted stress T1 reactivity has shown strong correlations with microvascular density in animal models of CMD (10). It is not surprising that there is blunted T1 reactivity in areas of high fibrosis and scarred or infarcted myocardium, where the increased tissue water content coincides with collagen deposition and fibrosis; these may account for a significantly elevated resting myocardial T1.In humans with chronic CAD, the remote myocardium shows blunted stress T1 reactivity, presumably secondary to a degree of CMD, in combination with a degree of compensatory resting vasodilation in the setting of cardiovascular risk factors (5). Further histopathologic interrogation of the remote myocardium may provide more insights, although this may not have been directly applicable in this animal model of acute ischemia isolated to a single coronary territory; further work is required to elucidate the potential mechanisms of a blunted stress T1 response in remote myocardium in humans with chronic CAD. Nonetheless, this invasive animal validation study of stress T1 mapping against histopathologic examination of infarction and myocardial fibrosis provides important clinical evidence of the imaging signals seen in stress T1 mapping and represents an important step toward its further clinical translation for assessing patients with cardiovascular disease.Disclosures of conflicts of interest: M.K.B. No relevant relationships. V.M.F. Grants from British Heart Foundation (BHF), National Institute of Heath Research (NIHR), NIHR Oxford Biomedical Research Centre; payment or honoraria from Massive Open Online Course (MOOC) Cardiac Imaging 2020; support for attending meetings and/or travel from European Society of Cardiology (ESC); patents planned, issued, or pending: Hann E, Piechnik SK, Popescu IA, Zhang Q, Werys K, Ferreira VM: “Method and Apparatus for quality prediction,” PCT/GB2020/050249, publication date August 13, 2020 (Publication number WO/2020/161481); Zhang Q, Piechnik SK, Ferreira VM, Hann E, Popescu IA: “Method and apparatus for enhancing medical images,” PCT/GB2020/052117, published March 11, 2021 (Publication number WO/2021/044153); Zhang Q, Piechnik SK, Ferreira VM, Werys K, Popescu IA: “A method for identity validation and quality assurance of quantitative magnetic resonance imaging protocols,” PCT/GB2020/051189, publication date November 26, 2020 (Publication number WO/2020/234570); Society for Cardiovascular Magnetic Resonance (SCMR), Vice Secretary Treasurer, SCMR Executive Committee, SCMR Board of Trustees.