Application of cardiac imaging in hypertensive patients

Introduction Hypertension (HTN) is the most common modifiable cause of death from cardiovascular disease (CVD), afflicting an estimated 30% of the world's population, half of whom are unaware they have the disease. In the USA, the overall cost for HTN was $90.5 million in 2010. In Australia in 2011‚Äö-13 one-third of adults had HTN, two thirds of whom had uncontrolled high blood pressure. HTN may lead to structural changes in the heart, which is called hypertensive heart disease (HHD). According to the American College of Cardiology (ACC) guideline for the management of chronic heart failure, HHD is classified into stage B heart failure. HTN causes left ventricular hypertrophy (LVH), a decrease of cardiac function, and may lead to fibrosis. However, the main focus of current cardiac clinical imaging is on left ventricular (LV) mass and left atrial (LA) dilatation, which reflect only macro changes of cardiac muscle. Functional diagnosis and tissue characteristics have not been mentioned in the latest guidelines for HTN diagnosis and there is still a gap in the literature that requires filling with high-quality studies. There are at least three reasons for this gap: (1) the sensitivity of conventional techniques is low and cannot follow the microscopic progression of the disease (sensibility); (2) conventional techniques may be affected by a number of factors and artefacts (reduced reliability); and (3) some advanced techniques require extra technical tasks (applicability). In this thesis, I am going to introduce several recent advanced validated imaging techniques which have high sensitivity and reliability but are still promisingly applicable to routine diagnosis of HHD. This thesis will combine three main focuses: (1) reviewing current states of the techniques; (2) evaluating the reproducibility of new techniques; and (3) applying these techniques to HHD. Aims This thesis aims to: (1) review and assess the usage of T1 mapping in HHD; (2) review and define normal ranges of feature tracking; (3) compare reproducibility of strain by tissue motion tracking in MRI (magnetic resonance imaging) and echocardiography; (4) assess the accuracy of volumes by strain analyses; (5) assess if LA function is independent of LV function in HHD; and finally, (6) identify which afterload is more strongly associated with LV anatomy and function. Methods Two systematic reviews and meta-analyses were performed to identify normal ranges in the literature and assess the applicability of the two techniques, T1 mapping and tissue motion tracking, in HHD. An important study platform for this research was LOWCBP. This is a multi-centre, randomised, open-label, blinded endpoint trial involving 308 patients being treated for uncomplicated HTN with controlled office blood pressure (OBP) (< 140/90 mmHg) but elevated central BP (‚Äöv¢‚Ä¢ 0.5SD above age and sex-specific normal values). Participants were randomised for intervention with spironolactone (25 mg/d) or usual care and are being followed over 24 months, with the primary outcome being LV mass index (using cardiac magnetic resonance imaging). Office and central BP is measured in the clinic and at home over seven days and by 24-hour ambulatory monitoring. Firstly, 54 patients were selected to assess the reproducibility of novel techniques. Secondly, 100 patients were selected to evaluate a new method for volume measurement. Thirdly, 72 appropriate patients were included in a study that used advanced methods to assess the association between LA and LV function in HHD. Finally, 108 patients were eligible for a study on the association of afterload with cardiac activities. Results The first meta-analysis showed that T1 mapping can categorise cardiac diseases into three groups: those readily identified (i.e. amyloidosis and hypertrophic cardiomyopathy); those difficult to distinguish from normal (i.e. HTN); and remaining disease entities that can be separated from the normal, but with significant overlap with other cardiac diseases. In addition, abnormal native T1 values can be shorter time (e.g. Fabry, Iron overload) or pronounced prolongation (e.g. amyloid), while extracellular volume was only increased irrespective of the underlying pathophysiology. In the second meta-analysis on normal ranges of strains by tissue motion tracking, 659 healthy subjects were included from 18 papers for MRI feature tracking. Pooled means of LV global longitudinal strain (LVGLS), LV global radial strain (LVGRS), and right ventricular global longitudinal strain (RVGLS) were determined. Metaregression showed that variation of LVGCS was associated with field strength. Variations of LVGLS, LVGRS, and RVGLS were not associated with any of age, sex, software, field strength, sequence, LV ejection fraction, or LV size. LVGCS appeared as the most robust in MRI feature tracking (FT). Among the MRI-derived strain techniques, the normal ranges were mostly concordant in LVGLS and LVGCS but varied substantially in LVGRS and RVGLS. In the third study, the results have shown that LVGLS and RVGLS in echocardiography were more reproducible than those in MRI. This would be because echocardiographic images have higher temporal resolution, and software can follow minor motion of cardiac tissues. In the fourth study, the volumes by strain analyses showed a great agreement with gold standard MRI volumes. In addition, the volumes by speckle tracking and FT showed a closer agreement with the gold standard compared to the conventional Simpson's biplane method. This method may also provide an opportunity to reduce the time taken for image analysis. In the fifth study, strain analyses proved that LA function is not simply another side of the same coin with LV function. In fact, LA contractile strain is independent of LVGLS in both MRI and echocardiography. Therefore, LA function assessment should be required in further studies about HHD. Finally, in the last study, I have pointed out the stronger association between 24-hour BP and LV mass, compared to OBP. OBP was associated with instantaneous parameters such as LV function. Conclusion This thesis has introduced novel advanced cardiac imaging techniques that may apply to HTN. Although T1 mapping was not applicable to HTN, tissue motion tracking appeared to be beneficial for high accurate cardiac diagnosis in hypertensive endorgan damage. In addition, volumes derived by the technique deviated less from the gold standard and the technique was less time-consuming, compared to the biplane method of disks on echocardiography; however, it requires further improvement. From the results of this study, this technique was reproducible and ready for clinical application. Thanks to the technique, additional independent variables were identified and allowed to assess cardiac function during hypertensive progression. Finally, combined with current imaging methods, the association between different components of BP and cardiac anatomy and function were represented. Unlike several assumptions about BP, each BP was found to have its own merits and was associated with different parameters of cardiac activity. Accordingly, advanced cardiac imaging gives us more profound insights into cardiac activities with more sensitive, qualitative, reproducible parameters and fewer assumptions. These parameters have made clear the association between different components of BP and different aspects of cardiac activity. The 24-hour BP measure would be preferable because it reflects the accumulated effects of HTN.