(Background) Hypertension remains an important cause of myocardial dysfunction and heart failure worldwide. Myocardial fibrosis, characterized by an elevation in collagen synthesis, is increased in systemic arterial hypertension. Histologic evidence of myocardial fibrosis is present in the early stages of hypertension. Non-invasive quantitation of myocardial fibrosis in hypertension would provide a quantitative parameter for the early diagnosis of target organ damage.
Cardiovascular Magnetic Resonance (CMR) with late gadolinium enhancement (LGE) has been widely used to detect focal myocardial fibrosis. However, diffuse fibrosis cannot be adequately evaluated by this technique due to the lack of normal reference myocardium. An alternative method is quantitation of myocardial T1 relaxation time (T1 mapping). Myocardial T1 mapping uses multiple images at different inversion times over several heartbeats to derive a T1 recovery curve. Look-Locker ‘LL’ and Modified Look-Locker ‘MOLLI’ sequences are the most appropriate sequences for myocardial T1 mapping. Animal studies have identified cardiac iron deposition to be involved in the development of myocardial interstitial fibrosis induced by angiotensin II with subsequent accumulation of type I and III collagen. T2* relaxometry has been clinically validated to estimate myocardial iron content in iron overload syndromes. Therefore, T2* mapping could be added to T1 mapping to further characterize myocardial fibrosis in hypertension.
(Main objective and hypothesis) The main objective of this study was investigation of T1 and T2* relaxometry using CMR in patients with hypertension. The hypothesis was that CMR relaxometry could quantify and assess hypertensive myocardial fibrosis at an earlier stage than other established techniques such as echocardiography, pulse wave velocity and strain imaging.
(Specific Aims) The work first established the native T1 and T2* values of normal myocardium in healthy volunteers with results correlated to echocardiography. A second investigation comprised longitudinal evaluation of the inter-study and inter-scan variability of myocardial relaxometry between two scanners (at 1.5T) over three years. Next, alterations of relaxometry in two groups of hypertensive patients was explored; the first group with normal and the second with abnormal diastolic function. Comparison of functional variables with echocardiography included myocardial deformation (strain) imaging. A final study investigated differences between LL and MOLLI T1 mapping sequences in hypertensive patients.
(Methods) Forty-one healthy volunteers underwent transthoracic echocardiography and a CMR study at 1.5T (Siemens Avanto) include SSFP imaging for LV function and mass, native T1 (LL and MOLLI 3,3,5) and T2* maps, myocardial tagging using SPAMM, aortic pulse wave velocity with phase contrast imaging. A second scan was performed by a different operator three years later using a different 1.5T scanner (Siemens Aera).
Two groups of hypertensive patients were enrolled with the study. The first group, consisting of 30 patients with recently diagnosed hypertension (duration ≤ 1 year) and normal systolic and diastolic function by echocardiography, underwent a comprehensive CMR protocol at 1.5T. This included SSFP imaging for LV function and mass, pre- and post-contrast T1 maps (using both LL and MOLLI 3,3,5 techniques) and T2* maps, myocardial tagging using SPAMM, aortic pulse wave velocity, adenosine stress perfusion and late gadolinium enhancement imaging. Myocardial T1 and T2* relaxation times were calculated from basal, midcavity and apical single short-axis slices using a standard AHA-16 segment model. The second group consisting of twenty patients with established hypertension (duration ≥ 1 year) with normal LVEF but abnormal diastolic function by echocardiography underwent the same scan protocol. Results for two hypertensive groups were compared with the control group.
(Result) Native T1 and T2* in healthy volunteers were comparable to the published literature. Reproducibility of T1 and T2* was high between observers with very low bias and high interclass correlation, and were consistent and reproducible over time and across scanners. Pre- and post-contrast of T1 and T2* maps of early hypertensive patients were significantly different compared with controls. Relaxometry of hypertensive patients with abnormal diastolic function was significantly different to controls and to the hypertensive patients with normal diastolic function. This suggests that impairment of diastolic function may correlate with fibrotic accumulation and implies a dose-dependent effect based on increased exposure to systemic hypertension.
Comparison MOLLI to LL techniques, the pre-contrast T1 was longer in MOLLI than LL for both controls and hypertensive patients. There was no difference in post-contrast T1 between the two. However, myocardial T1 mapping with LL and MOLLI sequences was increasingly discordant at higher T1 values.
(Conclusion) T1 and T2* mapping are able to detect subclinical changes of myocardial substrate in hypertensive patients with otherwise normal ejection fraction, normal LGE, and normal diastolic function. This provides the earliest non-invasive detection of end-organ cardiac damage. Increasingly abnormal T1 and T2* values are also associated with diastolic dysfunction and imply a dose-dependent relationship to hypertension. Myocardial relaxometry mapping methods hold promise for earlier detection of hypertensive heart disease, allowing earlier intervention and offer a quantitative target for monitoring the effects of antihypertensive medication.