Study population
From January 2016 to January 2020, 179 Adult Chinese Han race essential hypertensive patients with or without T2DM [HTN(T2DM+) and HTN(T2DM-), respectively] who underwent CMR at our institution were consecutively included. Hypertension was defined as a clinical systolic blood pressure (SBP) ≥ 140 mmHg and/or a diastolic blood pressure (DBP) ≥ 90 mmHg or a history of antihypertensive medications. The diagnosis of T2DM was based on the current American Diabetes Association guideline recommendations [15]. The exclusion criteria (Fig. 1) included patients with symptoms of heart failure (n = 3), LVEF < 50% (n = 8), known coronary artery disease (n = 10), myocardial infarction (n = 4), moderate to severe valvular disease (n = 3), cardiomyopathy (n = 3), bundle branch block (n = 5), atrial fibrillation (n = 2), serious liver and lung dysfunction (n = 6), estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73 m2 (n = 8) and poor image quality (n = 2). In addition, patients unmatched for age and sex (n = 15) were also excluded. Finally, 110 patients including 70 (35 men and 35 women; mean age, 55.0 ± 14.1 years) and 40 (20 men and 20 women; mean age, 55.7 ± 9.8 years) age- and sex-matched patients with HTN (T2DM-) and HTN (T2DM+), were eligible for this study. Another 37 healthy individuals (18 men and 19 women; mean age, 54.2 ± 10.5 years) matched for age and sex were selected from our healthy volunteer database to serve as the control group, and they underwent the same CMR examination.
This study was approved by the Biomedical Research Ethics Committee of our hospital and conducted in accordance with the Declaration of Helsinki (2013 EDITION).
CMR protocol
All the CMR examinations were performed using a 3.0 T whole-body scanner (Trio Tim; Siemens Medical Solutions, Erlangen, Germany) in the supine position. Data acquisition was performed with a standard ECG-triggering device that monitored each subject’s dynamic ECG changes during the end-inspiratory breath hold period. A balanced steady-state free precession (bSSFP) sequence (repetition time [TR]: 39.34 ms, echo time [TE]: 1.22 ms, flip angle: 40°, slice thickness: 8 mm, field of view [FOV]: 250 × 300 mm, and matrix size: 208 × 139) was used to acquire 8–12 continuous cine images from the base to the apex in the short-axis view, as well as vertical LV two- and four-chamber cine images in the long-axis view. For perfusion imaging, a dose of 0.2 mL/kg gadobenate dimeglumine (MultiHance 0.5 mmol/ml; Bracco, Milan, Italy) was injected into the right antecubital vein with a power injector (Stellant, MEDRAD, Indianola, PA, USA) at a flow rate of 2.5–3.0 mL/s, followed by 20 ml of saline. Rest first-pass perfusion images were acquired in three standard short-axis slices (basal, middle, and apical) and in one four-chamber view slice by inversion recovery prepared echo-planar imaging sequence (TR/TE: 163.0/1.12 ms, flip angle: 10°, slice thickness: 8 mm, FOV: 360 mm × 270 mm, and matrix size: 256 × 192). To exclude myocardial infarction, late gadolinium enhancement (LGE) images were acquired by segmented-turbo-FLASH–phase-sensitive inversion recovery (PSIR) sequences (TR/TE: 750 ms/1.18 ms; flip angle: 40°, slice thickness: 8 mm, FOV: 400 × 270 mm, and matrix size: 256 × 148) 10 − 15 min after contrast administration.
CMR data Analysis
CMR images were evaluated using offline commercially available software (cvi42, v. 5.10.2; Circle Cardiovascular Imaging, Calgary, Canada) by two radiologists with more than 3 years of CMR experience, who were blinded to the clinical data.
The endocardial and epicardial contours of the LV myocardium on the short-axis cine images were manually traced at the end-diastolic and end-systolic phases in the cmr42 short-3D module. Then, LV mass at end-diastole, LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), LVEF, stroke volume and cardiac index were computed automatically. The trabeculae and papillary muscles were excluded from the LV mass and included in the LV cavity. LV mass, LVEDV and LVESV indexed for body surface area (BSA) (LVMI, LVEDVI and LVESVI, respectively) were calculated using the Mosteller formula [16]. In addition, LV remodeling index, calculated as LVM/LVEDV, was included for analysis.
The LV global radial (GRPS), circumferential (GCPS) and longitudinal peak strain (GLPS) were obtained by manually delineating the endocardium and epicardium of the cine images at the end-diastole from the short-axis and long-axis two- and four-chamber slice views in the tissue tracking module. Strain was depicted as relative lengthening, shortening and thickening of the myocardium from end diastole (reference phase).
For the evaluation of first-pass myocardial perfusion (Fig. 2), the endocardium and epicardium and a region of interest drawn in the LV chamber were manually determined in the first-pass perfusion images (basal, middle and apical). Then, signal intensity-time curves were generated for the blood pool and each myocardial segment based on the 16-segment heart model. Consequently, semiquantitative segmental perfusion indices including the upslope, time to maximum signal intensity (TTM), and max signal intensity (MaxSI) were acquired automatically, and the global first-pass myocardial perfusion indices for each subject were calculated by averaging the regional values of the 16 myocardial segments. In addition, the presence of LGE was visually evaluated by the two radiologists with consensus.
Reproducibility of LV myocardial strain and perfusion
Intra- and inter-observer variabilities for the LV global myocardial strain and perfusion indices were analyzed in 30 random cases including 20 HTN patients and 10 controls. To determine the intra-observer variability, one observer (XM. L) evaluated the same subjects on two separate measurements one month apart. For the inter-observer variability evaluation, a second investigator (L. J) who was blinded to the first observer’s results and clinical data reanalyzed the measurements.
Statistical analysis
Categorical variables are presented as frequencies (percentages) and were compared using Chi-square tests. Continuous variables were evaluated for normality distribution by the Shapiro-Wilk test and are expressed as the mean ± standard deviation (SD). One-way analysis of variance (one-way ANOVA) was used to compare the baseline characteristics among the normal and HTN groups. Comparisons of the CMR-derived parameters between different groups were evaluated by analysis of covariance (ANCOVA) after adjusting for age, sex, body mass index (BMI) and heart rate followed by Bonferroni’s post hoc test. Pearson’s correlation coefficient was used to determine the correlation between LV myocardial strain and first-pass myocardial perfusion indices in HTN. Backwards stepwise multivariable linear regression analyses were performed to determine the predictors for LV strains and myocardial perfusion indices in the whole population and patients with HTN. Inter- and intra-observer agreements were determined by the evaluation of intraclass correlation coefficients (ICCs). All analyses were performed in SPSS version 21 (IBM, Armonk, New York, USA), and a two-tailed p < 0.05 was considered significant.