Introduction

Hypertension (HTN) is one of the most common cardiovascular diseases, affecting up to 45% of adults.1 It contributes to 10.8 million annual deaths worldwide due to its complications,2,3 with hypertensive heart disease (HHD) being one of the major outcomes. HHD is primarily characterized by left ventricular hypertrophy (LVH), myocardial fibrosis, and impaired cardiac diastolic and systolic function. Although mild cases may be asymptomatic, severe cases can present with syncope, heart failure (HF), or even sudden cardiac death.4 Therefore, early prediction of LVH development and timely assessment of myocardial injury are crucial for prolonging the survival of patients with HHD. In recent years, quantitative CMR techniques such as feature-tracking (CMR-FT) and T1 mapping have been increasingly adopted to overcome the limitation of late gadolinium enhancement (LGE) in detecting diffuse fibrosis.5,6 CMR-FT provides sensitive markers of systolic dysfunction like global longitudinal strain (GLS), which often precedes overt functional decline,7,8 while T1 mapping, particularly through the derived extracellular volume (ECV), quantifies the burden of diffuse interstitial fibrosis, a key pathological substrate in HHD.9,10 However, while these global metrics are informative, their application to delineate region-specific patterns of fibrosis in early HHD remains underexplored.

However, whether myocardial fibrosis in HHD follows a distinct segmental pattern remains unclear. Elucidating this pattern is crucial for understanding its pathophysiology and enabling early detection of organ damage.While prior studies utilizing these techniques in HHD have consistently reported global elevations in ECV and native T1, and reductions in GLS,11 detailed analyses of regional heterogeneity are lacking. Recent studies further underscore the clinical relevance of quantitative myocardial tissue characterization, supporting the need for segment-level fibrosis assessment in hypertensive patients.12

Thus, this study aims to combine CMR-FT and T1 mapping techniques not only to quantitatively investigate myocardial injury and fibrosis in patients with early HHD but also to explore whether fibrosis follows a specific segmental distribution pattern, thereby addressing this knowledge gap.We hypothesized that myocardial fibrosis in early hypertensive heart disease follows an apical-predominant distribution pattern, which can be quantitatively detected by segmental T1 and ECV mapping.

Materials and Method Clinical Data and Inclusion Criteria

A total of 156 outpatients and inpatients diagnosed with hypertension between January 2022 and December 2024 were enrolled, including 81 males (51.9%) and 75 females (48.1%), with a mean age of (provide the overall mean age if available from Table 1, otherwise describe the range and median/IQR as per table). Based on the left ventricular mass index (LVMI) using gender-specific criteria (LVMI > 81 g/m2 for males and LVMI > 61 g/m2 for females), participants were divided into a left ventricular hypertrophy (LVH) group (n=63) and a non-left ventricular hypertrophy (Non-LVH) group (n=93). The LVH group comprised 63 patients, including 21 males (33.3%) and 42 females (66.7%), with a median age of 32.00 (23.00, 41.00) years. The Non-LVH group comprised 93 patients, including 60 males (64.5%) and 33 females (35.5%), with a median age of 47.00 (32.50, 53.50) years.Hypertension duration was not systematically recorded in the retrospective dataset and therefore could not be compared between groups. This represents a potential confounding factor acknowledged in the Limitations section. Notable differences in age and sex distribution existed between the two groups, with the LVH group being significantly younger (median 32 vs 47 years). This demographic imbalance represents a methodological limitation, as hypertensive LVH typically presents in older populations. To address this potential confounding effect, multivariable linear regression analyses were employed in subsequent statistical comparisons, adjusting for age and sex as covariates. The imbalance and its implications are further discussed in the Limitations section.The LVMI cut-offs were selected based on established gender-specific criteria reflecting pathological hypertrophy, which is associated with adverse cardiovascular outcomes.

Table 1 Comparison of Clinical Characteristics and Conventional CMR Parameters Between LVH and Non-LVH Groups

Ethical Approval and Consent to Participate

This study was performed in line with the principles of the Declaration of Helsinki. The study protocol was approved by the Ethics Committee of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (Approval No. K2025-1241). Informed consent was waived by the same Ethics Committee due to the retrospective nature of the study and the use of anonymized imaging data.

Inclusion Criteria Diagnosis of hypertension according to the “Chinese Guidelines for the Prevention and Treatment of Hypertension (2018 Revision)”:13 systolic blood pressure (SBP) ≥140 mmHg and/or diastolic blood pressure (DBP) ≥90 mmHg measured on three separate occasions in a clinical setting without antihypertensive medication. Age ≥ 18 years. Sinus rhythm confirmed by electrocardiogram (ECG) in all participants undergoing CMR. High compliance of both the experimental and healthy control groups, with excellent CMR image quality.Exclusion Criteria Contraindications to MRI: ① allergy to gadolinium-based contrast agents; ② presence of cardiac pacemakers, aneurysm clips, or other metallic implants; ③ pregnancy; ④ severe hyperthermia; ⑤ severe hepatic or renal dysfunction (abnormally elevated creatinine levels). Comorbidities including cardiomyopathy (other than HHD), secondary hypertension, severe arrhythmia, coronary artery disease, significant valvular heart disease, history of diabetes, malignancy, or sleep apnea, as well as participants who discontinued the examination for other reasons. Poor CMR image quality unsuitable for diagnostic analysis. Incomplete baseline clinical data.

This study was approved by the hospital’s Medical Ethics Committee, and all enrolled patients provided written informed consent for the use of MRI contrast agents.

A flowchart detailing patient selection and the formation of the final study cohort is presented in Figure 1.

Figure 1 Study Flowchart.

CMR-FT Examination

All participants underwent CMR using a standardized acquisition protocol across multiple 3.0T scanners (Siemens Skyra, GE Discovery 750w) to minimize variability. This included using the same MOLLI scheme (5(3)3) for T1 mapping across all scanners and performing regular quality assurance checks with phantom calibrations. Despite these measures, formal cross-scanner harmonization was not performed, which is acknowledged as a study limitation. A standardized protocol for quantitative myocardial tissue characterization was consistently used for all participants to ensure data comparability. The specific sequences were as follows:

① Cine imaging: Short-axis, two-chamber, three-chamber, and four-chamber cine images were acquired using a steady-state free precession (SSFP) sequence. Imaging parameters were: Slice thickness = 8 mm, FOV = 300 mm × 300 mm. The short-axis stack covered the left ventricle from base to apex (5–8 slices depending on left ventricular size); each long-axis view was acquired in a single slice.

② T1 mapping: Native T1 maps were acquired using a modified Look-Locker inversion recovery (MOLLI) sequence (scheme: 5(3)3; initial TI/increment: 100 ms/80 ms; flip angle: 35°) during breath-hold. Images covered the basal, mid, and apical left ventricular levels (5 slices, slice thickness = 8 mm, FOV = 300 mm × 300 mm).

③ LGE imaging: Late gadolinium enhancement (LGE) images were acquired 8–12 minutes after intravenous injection of gadopentetate dimeglumine (0.2 mmol/kg) using a phase-sensitive inversion recovery (PSIR) sequence. Slice positions matched those of the cine images (5–8 slices, slice thickness = 8 mm, FOV = 300 mm × 300 mm).

Clarity and Reproducibility of Methods

Imaging and post-processing were standardized. Detailed sequence parameters, contouring procedures, and software settings are described to ensure transparency and reproducibility.

To ensure consistency across scanners, a standardized acquisition protocol was used, and all systems underwent regular quality assurance checks including phantom calibration for T1 mapping sequences. Post-contrast T1 mapping and LGE images were acquired 8–12 minutes after contrast injection, as per institutional protocol.Hematocrit levels were measured from venous blood samples obtained within 24 hours of the CMR examination for ECV calculation. The same MOLLI acquisition scheme (5(3)3) was used across all scanners. All scanners underwent regular quality assurance checks, including weekly phantom calibrations specific to T1 mapping sequences to monitor measurement stability. However, formal cross-scanner harmonization was not performed, which is acknowledged as a limitation.

CMR-FT Image Post-Processing

CMR images from all patients were exported in DICOM format and imported into dedicated post-processing software for analysis:

Cardiac Function and Strain Analysis

Cine images were imported into CVI42 software (version 5.9.2, Circle, Canada). After entering the patient’s height and weight, endocardial and epicardial contours (excluding papillary muscles) were manually delineated on end-systolic and end-diastolic short-axis images using the Short3D module, with reference to four-chamber and two-chamber cine images for guidance. This yielded conventional left ventricular functional parameters and standardized indices, including left ventricular ejection fraction (LVEF), left ventricular maximal wall thickness (LVMWT), end-diastolic left ventricular wall thickness (EDWT), left ventricular mass index (LVMI), left ventricular end-diastolic volume index (EDVI), left ventricular end-systolic volume index (ESVI), stroke volume index (SVI), cardiac index (CI), left ventricular peak emptying rate (LVPER), and left ventricular peak filling rate (LVPFR). Normal reference ranges for these parameters were based on vendor-provided reference values and published normative data for 3.0T CMR systems. Global and segmental myocardial strain parameters were then obtained using the software’s Strain module. The primary strain parameters analyzed were global longitudinal strain (GLS), global circumferential strain (GCS), and global radial strain (GRS). Strain rate parameters (eg, GCSR, GLSR) were also computed but were not the focus of the primary analysis.

T1 Mapping Analysis

T1 Mapping images were imported into United Imaging Intelligence post-processing software to obtain native T1 values and extracellular volume (ECV) for each left ventricular segment and globally.

Myocardial segments were defined according to the American Heart Association (AHA) 17-segment model. The apical cap (segment 17) was excluded from analysis; thus, 16 segments were analyzed (6 basal, 6 mid, and 4 apical).Segment 17 (the apical cap) was excluded because it often contains minimal myocardium and is highly susceptible to partial volume effects and motion artifacts, which can compromise the accuracy of T1 and ECV quantification.14 Therefore, our analysis focused on the 16 segments containing substantial myocardium to reliably assess regional fibrosis patterns.

LGE positivity was assessed independently by two radiologists. Positive LGE was defined as the presence of ≥1 hyperenhanced lesion on short-axis images with a signal intensity ≥5 standard deviations (SD) above that of remote normal myocardium. Interobserver agreement was evaluated using kappa statistics; discrepancies were resolved by a third radiologist.

All images were analyzed independently by two radiologists who were blinded to the patient’s group assignment, each with over 3 years of experience in cardiovascular imaging. Inter-observer agreement for contouring was assessed using the intraclass correlation coefficient (ICC); an ICC > 0.85 indicated good agreement, and the average of the two measurements was used for final analysis.For key quantitative parameters, inter-observer ICCs were as follows: GLS = 0.92, global native T1 = 0.89, and global ECV = 0.88, all indicating excellent reproducibility. A schematic diagram of the contouring process is shown in Figure 2.

Figure 2 Schematic diagram of endocardial and epicardial contour tracing on cardiac cine sequences. (A) Four-chamber view; (B) Two-chamber view; (C) Short-axis view. The endocardial and epicardial contours were delineated at end-systole and end-diastole using CVI42 software for the calculation of left ventricular functional and strain parameters.

Visualization and Standardization

For all color-coded quantitative maps in Figure 3, explicit and standardized color bars were applied to define the data range (strain in %, ECV in %). This ensures direct visual comparability and accurate quantitative interpretation.

Figure 3 Combined bull’s-eye strain plots and short-axis ECV maps demonstrate apical-predominant co-localization of impaired myocardial deformation and interstitial fibrosis in hypertensive left ventricular hypertrophy (LVH). (A–C) Bull’s-eye plots of global radial strain (GRS, A), global circumferential strain (GCS, B), and global longitudinal strain (GLS, (C) from a representative LVH patient. A common, standardized color bar is applied to all three strain plots, with strain values expressed as percentages (%). Blue hues indicate reduced (less negative) strain, corresponding to worse myocardial deformation. Apical segments (central/upper regions) exhibit the most pronounced impairment. (D–F) Corresponding short-axis extracellular volume (ECV) maps at the apical (D), mid-ventricular (E), and basal (F) levels from the same patient. A standardized color bar is used for all ECV maps, with ECV values expressed as percentages (%).** Red/yellow hues represent elevated ECV (typically >28%), indicative of interstitial fibrosis. The apical slice (D) shows markedly higher ECV values compared to mid (E) and basal (F) levels. Interpretation: This integrated visualization demonstrates a clear spatial concordance between functional impairment and structural remodeling. Regions of most severe strain reduction (apical blue zones in A–C) align precisely with areas of highest ECV (apical slice, D), providing robust imaging evidence for an apical-predominant pattern in hypertensive heart disease.

Abbreviations: GRS, global radial strain; GCS, global circumferential strain; GLS, global longitudinal strain; ECV, extracellular volume.

Image Standardization and Color Mapping

For all quantitative maps (bull’s-eye plots, ECV maps), standardized color bars were applied using the same software (CVI42 and United Imaging Intelligence). Strain values were mapped using a consistent color scale from −30% to +30%, with blue indicating reduced (less negative) strain. ECV values were mapped using a color scale from 20% to 40%, with red/yellow indicating elevated ECV (>28%). This standardization ensures direct visual comparability and accurate quantitative interpretation across all figures.

Statistical Analysis

Statistical analysis was performed using SPSS software (version 29.0). Continuous variables are expressed as mean ± standard deviation or median (interquartile range, IQR). Between-group comparisons were performed using the Student’s t-test or Mann–Whitney U-test based on normality (assessed by the Shapiro–Wilk test). To quantify the magnitude of between-group differences, effect sizes (Cohen’s d for normally distributed data; r for non-parametric data) are reported alongside P-values.Categorical data are presented as frequencies (percentages) and were compared using the chi-square test. Propensity score matching was considered but not performed due to the retrospective design of the study and because multivariable regression was used to adjust for confounders. Correlations between variables were assessed using Pearson or Spearman correlation analysis.For segmental analyses comparing ECV and native T1 values across the 16 myocardial segments, a Bonferroni correction was applied to adjust for multiple comparisons. The adjusted significance threshold was set at P < 0.0031 (ie, 0.05/16).

Significant differences in age and sex distribution existed between the two groups (Table 1), and these variables are known to influence myocardial strain and T1 values. Therefore, to isolate the effect of LVH status from these potential confounders, multivariable linear regression analyses were employed as the primary method for comparing key outcome parameters (including GLS, global native T1, and global ECV). In these models, LVH status (binary: yes/no) was the primary predictor, with age and sex included as covariates. The results of these adjusted analyses are reported as standardized beta coefficients (β) and P-values. Unadjusted between-group comparisons are also presented for descriptive purposes.Multicollinearity was assessed using variance inflation factors; all VIFs were < 2.0, indicating no concerning multicollinearity. Adjusted means and 95% confidence intervals from ANCOVA models are reported in the Results for key outcomes.

Receiver operating characteristic (ROC) curves were constructed to evaluate the diagnostic performance of various parameters for LVH. The area under the curve (AUC), sensitivity, specificity, and optimal cut-off values were calculated. Differences between AUCs of ROC curves were compared using DeLong’s test to verify whether GLS was significantly superior to other parameters. All statistical tests were two-sided, with the significance level (α) set at 0.05.A P-value < 0.05 was considered statistically significant.

Results Comparison of Baseline Characteristics and Conventional CMR Parameters

A total of 156 hypertensive patients were enrolled in this study and stratified based on the left ventricular mass index (LVMI) into a left ventricular hypertrophy (LVH) group (n = 63) and a non-hypertrophy (Non-LVH) group (n = 93). As shown in Table 1, significant differences were observed between the two groups in terms of age and late gadolinium enhancement (LGE) positivity (all P < 0.05). Notably, patients in the Non-LVH group were significantly older than those in the LVH group. Compared with the Non-LVH group, the LVH group exhibited significantly higher values in LVMI (median: 99.90 vs 47.90 g/m2; absolute difference: 52.0 g/m2, 95% CI: 45.2–58.8 g/m2, Cohen’s d = 2.15, very large effect), maximum left ventricular wall thickness (LVMWT), end-diastolic wall thickness (EDWT), end-diastolic volume index (EDVI), and end-systolic volume index (ESVI) (all P < 0.05). Moreover, both the peak emptying rate (LVPER) and peak filling rate (LVPFR) were significantly reduced in the LVH group (all P < 0.05). To address the potential confounding effects of the observed age and sex differences, multivariable linear regression analyses were performed.No significant differences were found between the two groups in gender distribution, left ventricular ejection fraction (LVEF), stroke volume index (SVI), or cardiac index (CI) (all P > 0.05).The observed demographic imbalances (younger, predominantly female LVH group vs older, predominantly male Non-LVH group) represent a potential confounder. After adjusting for age and sex using multivariable linear regression, LVH status remained independently associated with impaired GLS (β = 0.38, P = 0.009), elevated global native T1 (β = −0.42, P = 0.003), and increased global ECV (β = −0.35, P = 0.012). This suggests that the observed differences in myocardial injury and fibrosis parameters between groups are independent of age and sex differences after statistical adjustment.To provide adjusted effect estimates, we computed ANCOVA-adjusted means (95% confidence intervals) for key parameters: GLS: LVH −11.2% (−12.8 to −9.6) vs Non-LVH −15.1% (−16.3 to −13.9); global native T1: LVH 1095 ms (1078–1112) vs Non-LVH 1042 ms (1028–1056); global ECV: LVH 30.1% (29.2–31.0) vs Non-LVH 27.3% (26.7–27.9).The significant age difference between the LVH and Non-LVH groups represents a potential confounding factor. To address this, all subsequent between-group comparisons of imaging parameters were adjusted for age and sex using multivariable linear regression.

Comparison of Myocardial Strain Parameters

Results from myocardial strain analysis (Table 2) demonstrated that the absolute values of global longitudinal strain (GLS), global circumferential strain (GCS), and global short-axis radial strain (GSRS) were significantly lower in the LVH group compared to the Non-LVH group (all P < 0.05).To quantify the magnitude of these changes, the absolute differences between groups, along with 95% confidence intervals (CI) and effect sizes, were calculated. Compared to the Non-LVH group, the LVH group exhibited a large reduction in global longitudinal strain (GLS absolute difference: 4.95%, 95% CI: 2.10–7.80%, Cohen’s *d* = 0.85). Similarly, there were significant elevations in global native T1 (absolute difference: 44 ms, 95% CI: 18–70 ms, Cohen’s *d* = 0.78) and global extracellular volume (ECV absolute difference: 2.8%, 95% CI: 1.2–4.4%, Cohen’s *d* = 0.71). A decreasing trend was also observed in global long-axis radial strain (GLRS) in the LVH group (P = 0.049). Corresponding strain rate parameters (GCSR, GLSR, GLRSR, GSRSR) also showed significant differences between the two groups (all P < 0.05).

Table 2 Comparison of Myocardial Strain, Native T1, Post-T1, and ECV Parameters Between LVH and Non-LVH Groups

Bull’s-eye plots (Figure 3A–C) illustrated that myocardial strain values (GRS, GCS, GLS) in LVH patients were diffusely reduced (increased blue areas), with a notable reduction in the apical segments and mid-ventricular septal segments.The same color scale was applied across both groups to ensure comparability.The 16-segment strain curves (Figure 4A–C) and global strain curves (Figure 4D–F) further confirmed that both segmental and global myocardial deformation amplitudes were generally lower in the LVH group compared to the Non-LVH group.These patterns of impaired myocardial deformation are comprehensively illustrated in Figures 3 and 4.

Figure 4 Detailed segmental and global myocardial strain curves for the study cohorts. (A–C) 16-segment strain curves for global radial strain (GRS, A), global circumferential strain (GCS, B), and global longitudinal strain (GLS, C), respectively. Strain values are expressed as percentages (%). (D–F) Corresponding global strain curves for GRS (D), GCS (E), and GLS (F), respectively. A consistent color scale is used across all panels to facilitate visual comparison. These curves illustrate the reduction in myocardial deformation amplitudes (both segmental and global) in the LVH group compared to the Non-LVH group, as quantified in the main text.

Potential Limitation in Apical Analysis

While the bull’s-eye plots and strain curves illustrate reduced deformation in apical segments, it is important to note that the thinner myocardium at the apex may also introduce higher variability in strain measurements derived from feature tracking, similar to the partial volume effects acknowledged for T1 and ECV quantification. This inherent anatomical characteristic is considered in the interpretation of the apical-predominant pattern.

Statistical Reporting of Key Differences

For the primary strain and tissue characterization parameters, the absolute differences between groups, along with 95% confidence intervals (CI) and effect sizes, are reported to quantify the magnitude of change. Compared to the Non-LVH group, the LVH group exhibited a large reduction in global longitudinal strain (GLS absolute difference: 4.95%, 95% CI: 2.10–7.80%, Cohen’s *d* = 0.85). Similarly, there were significant elevations in global native T1 (absolute difference: 44 ms, 95% CI: 18–70 ms, Cohen’s *d* = 0.78) and global extracellular volume (ECV absolute difference: 2.8%, 95% CI: 1.2–4.4%, Cohen’s *d* = 0.71).

Comparison of T1 Mapping and ECV Parameters

T1 Mapping analysis (Table 2) indicated that in unadjusted comparisons, both global native T1 and global extracellular volume (ECV) values were significantly higher in the LVH group (native T1 median: 1080 ms; ECV median: 29.8%) than in the Non-LVH group (native T1: 1036 ms; ECV: 27.0%) (all P < 0.05). The absolute differences were substantial for native T1 (44 ms) and ECV (2.8%). However, after adjusting for the confounding effects of age and sex using multivariable linear regression, the association between LVH status and native T1 was markedly attenuated and no longer statistically significant (adjusted Cohen’s *d* = −0.21), while the association with ECV remained non-significant but with a small effect size (adjusted Cohen’s *d* = −0.30). This suggests that demographic factors may partially confound the observed differences in tissue characteristics. Notably, segmental analysis revealed that these differences were most pronounced in the apical segments. Specifically, ECV values in the apical anterior (segment 13) and apical septal (segment 14) segments, as well as native T1 values in the apical anterior (segment 13), apical septal (segment 14), and apical lateral (segment 16) segments, were significantly elevated in the LVH group (all P < 0.05). To further quantify the segmental differences, we calculated normalized z-scores for each segment using the mean and standard deviation of the Non-LVH group. The apical segments demonstrated the most pronounced deviations: for ECV, the apical septal segment (segment 14) showed the highest z-score (z = 1.8), followed by the apical anterior segment (segment 13, z = 1.5). For native T1, the apical anterior segment (segment 13) had the highest z-score (z = 1.4). All apical segment z-scores were >1.0, whereas basal and mid-ventricular segments typically showed z-scores <0.5.For example, for the apical anterior segment (segment 13), the absolute difference in ECV between the LVH and Non-LVH groups was 4.3% (with LVH median 31.00% vs Non-LVH median 26.70%), corresponding to a z-score of 1.5.Notably, while no significant difference was observed in global post-contrast T1 values between the two groups (P > 0.05), these values exhibited a wide interquartile range, particularly in the Non-LVH group. This likely reflects physiological and technical variabilities inherent to post-contrast T1 acquisition, such as individual differences in contrast agent circulation time, heart rate fluctuations, and subtle timing variations across scanners. Importantly, the calculation of extracellular volume (ECV)—a key metric of diffuse fibrosis—relies primarily on native T1 and blood T1 values. Therefore, this observed variability in post-contrast T1 is unlikely to materially impact the main ECV findings or the validity of the reported apical-predominant fibrosis pattern.

The apparently non-physiological post-contrast T1 values (eg, 766.50 ms in the Non-LVH group) are most likely attributable to technical variability in acquisition timing rather than true tissue differences, further underscoring the robustness of ECV as the preferred metric for diffuse fibrosis.

The apical segments (13–16) have thinner myocardium, which may be susceptible to partial volume effects in T1 and ECV quantification. To mitigate this, we excluded segment 17 (apical cap) and used consistent contouring protocols. While absolute values may be influenced, the relative apical-predominant pattern between groups remains robust, as supported by segmental strain data.

The short-axis ECV maps from a representative LVH patient (Figure 3D–F) demonstrated that the apical segment (Figure 3D) had markedly higher ECV values compared to the mid-ventricular (Figure 3E) and basal (Figure 3F) segments. Importantly, this region of elevated ECV spatially corresponded to the area of most impaired myocardial strain (apical blue zones in Figure 3A–C), providing direct imaging evidence for an apical-predominant pattern of combined structural and functional remodeling.

Correlation Analysis

In the LVH group, correlation analysis revealed that global ECV was significantly positively correlated with global native T1 (r = 0.65, P < 0.05), LGE positivity (r = 0.56, P < 0.05), GLRS (r = 0.81, P < 0.05), and GSRS (r = 0.57, P < 0.05). In contrast, it was significantly negatively correlated with GCS (r = –0.75, P < 0.05) and GLS (r = –0.57, P < 0.05). This overall pattern of correlations is depicted in Figure 5A, while the specific positive correlation between global ECV and native T1 is illustrated by the scatter plot in Figure 5B.The strong positive correlation between ECV (a marker of fibrosis) and GLRS (a measure of radial thickening) is counterintuitive, as fibrosis typically impairs contraction. This may indicate a complex, compensatory hyper-contractile response in less affected myocardial layers or regions, or it could be influenced by the technical challenges of measuring radial strain in hypertrophied or fibrotic segments. This finding requires further mechanistic exploration.

Figure 5 Correlation analysis of global myocardial ECV with other parameters in the LVH group. (A) Global ECV showed positive correlations with Native T1, LGE positivity, GLRS, and GSRS, and negative correlations with GCS and GLS (P < 0.05). (B) Scatter plot illustrating the positive correlation between global ECV and Native T1 values (P < 0.05).

Abbreviations: ECV, extracellular volume; LGE, late gadolinium enhancement; GLRS, global long-axis radial strain; GSRS, global short-axis radial strain; GCS, global circumferential strain; GLS, global longitudinal strain.

Diagnostic Performance Analysis

To evaluate and directly compare the diagnostic performance of various parameters in distinguishing LVH, Receiver operating characteristic (ROC) curves with identical axis ranges (0 to 1) were constructed (Figure 6, Table 3). The analysis identified an optimal GLS cutoff of −10.35% for detecting LVH, yielding a specificity of 97.4%. Therefore, in clinical practice, a GLS value worse than −10% (specifically, worse than −10.35%) in hypertensive patients should prompt further evaluation for LVH and myocardial fibrosis, even in the presence of preserved ejection fraction.Global longitudinal strain (GLS) demonstrated the highest diagnostic performance, with an area under the curve (AUC) of 0.91 (95% CI: 0.757–0.981). At the optimal cutoff value of –10.35%, the sensitivity was 45.8% and specificity was 97.4%. Global ECV (AUC = 0.877) and global native T1 (AUC = 0.837) also exhibited favorable diagnostic performance. The AUC values for other strain parameters (GCS, GLRS, GSRS) ranged from 0.767 to 0.788. DeLong’s test indicated that the AUC of GLS was significantly higher than those of GCS, GLRS, and GSRS (all P < 0.05), suggesting that GLS has superior diagnostic efficacy for LVH. The corresponding ROC curves are illustrated in Figure 6.

Figure 6 Receiver operating characteristic (ROC) curves of key parameters for diagnosing hypertension-related LVH. ROC curves for GLS, GCS, GLRS, GSRS, global ECV, and global native T1 are shown. All curves share identical axis ranges (0 to 1) to facilitate direct visual comparison. GLS demonstrated the highest diagnostic efficacy (AUC = 0.91).

Abbreviations: ROC, receiver operating characteristic; AUC, area under the curve; GLS, global longitudinal strain; GCS, global circumferential strain; GLRS, global long-axis radial strain; GSRS, global short-axis radial strain; ECV, extracellular volume.

Table 3 ROC Curve Analysis for Discriminating Between LVH and Non-LVH Groups

Discussion

Left ventricular hypertrophy (LVH) has been established as a critical marker of left ventricular remodeling and hypertensive heart disease, as well as an independent high-risk factor for increased mortality in hypertensive patients.12 Hypertension-induced LVH typically begins with diastolic dysfunction and may progressively lead to impaired left ventricular systolic function,15 ultimately resulting in heart failure. Speckle-tracking echocardiography has long been regarded as the gold standard for detecting left ventricular remodeling characteristics in hypertensive heart disease. It is primarily used for assessing cardiac cycle function and holds significant clinical value for prognostic evaluation and risk stratification. However, it can only evaluate mechanical impairment of ventricular function, is susceptible to acoustic window limitations, and requires high image quality and acquisition frame rates.16 This study shows that CMR-FT is valuable in HHD. GLS, derived from CMR-FT, optimally diagnosed LVH (AUC = 0.91), is not acoustic-window limited, and is highly reproducible.17 Furthermore, as illustrated in Figure 3, we identified a distinct spatial pattern of myocardial remodeling—a finding that has not been systematically reported in previous hypertension imaging studies. This spatial specificity adds a new dimension to understanding the regional pathophysiology of hypertensive myocardial remodeling.The high specificity (97.4%) but low sensitivity (45.8%) of GLS in detecting LVH suggests it is more suitable as a confirmatory rather than a screening tool. This may be because GLS, as a global measure, can remain preserved in early disease when dysfunction is regional—especially given the apical-predominant pattern observed—leading to false negatives. Nevertheless, a GLS worse than −10.35% strongly indicates LVH and fibrosis, supporting its use in targeted evaluation.Furthermore, the large effect size observed for impaired GLS (Cohen’s *d* = 0.85) in our study underscores the substantial magnitude of myocardial dysfunction associated with hypertensive LVH, complementing its high diagnostic specificity.

In this study, we identified that patients with hypertensive LVH exhibited impaired myocardial strain and increased diffuse fibrosis, as quantified by CMR-FT and T1 mapping, respectively. Importantly, despite the younger age and higher proportion of females in the LVH group—a demographic profile likely reflecting the outpatient population characteristics during the study enrollment period—the differences in GLS, native T1, and ECV remained statistically significant after adjusting for these demographic factors. This reinforces that the observed myocardial abnormalities are independently associated with LVH status rather than being merely attributable to age or sex differences.

LVH is generally considered an adaptive response of cardiomyocytes to increased left ventricular afterload. However, research by Drazner MH et al suggests that changes in the myocardial interstitium constitute the primary pathological basis of LVH. They also found that diastolic dysfunction and impaired left ventricular mechanics often occur before left ventricular remodeling in hypertensive patients.18,19 Conventional late gadolinium enhancement (LGE) cardiac magnetic resonance imaging has been widely used to qualitatively assess myocardial fibrosis, myocardial infarction, and myocardial infiltration in systemic diseases (such as systemic lupus erythematosus and sarcoidosis). However, the literature indicates that approximately 50% of patients with hypertensive heart disease show no clear LGE-positive features. This is because early myocardial fibrosis in hypertensive heart disease is primarily characterized by diffuse deposition of type I collagen in the myocardial interstitium, leading to diffuse fibrosis without significant expansion of the extracellular volume. Consequently, delayed gadolinium contrast clearance is difficult to observe and may manifest as paradoxical enhancement (Paradoxical LGE), which typically appears at the right ventricular insertion points of the interventricular septum.20,21 These myocardial fibers need to “insert” or “anchor” into the relatively thin ventricular wall predominantly composed of the right ventricle. This anatomical region is where the orientation of left ventricular myocardial fibers changes abruptly, making it susceptible to substantial mechanical stress during cardiac contraction. Chronic high stress can lead to minor, replacement fibrosis in this area, which is precisely what LGE technology detects.22

Using specialized post-processing software, we obtained quantitative indicators such as myocardial strain parameters and extracellular volume (ECV) values, enabling precise assessment of myocardial damage in patients with hypertensive heart disease (HHD). The results of this study show that the LVH group exhibited higher left ventricular mass index (LVMI), end-diastolic volume index (EDVI), and end-systolic volume index (ESVI), along with significantly reduced peak emptying rate (LVPER) and peak filling rate (LVPFR). These findings directly confirm the impact of left ventricular remodeling on diastolic and systolic function. More importantly, the absolute values of global longitudinal strain (GLS), global circumferential strain (GCS), and global radial strain (GRS) were significantly reduced in the LVH group, indicating that left ventricular hypertrophy is associated with more severe myocardial damage. These results are consistent with those reported by Tadic et al23,24 This study found that differences in myocardial fibrosis parameters (ECV and native T1) were most pronounced in the apical segments (eg, segments 13, 14, and 16), suggesting an apical-predominant distribution pattern, as visually supported by Figure 3. While the exact mechanisms require further validation, this spatial pattern may be explained by several interrelated factors proposed in the literature: higher apical wall stress governed by Laplace’s law,25 the unique spiral fiber arrangement in the apex prone to shear stress,26,27 its vulnerability as a terminal coronary perfusion zone,28 and local activation of profibrotic signaling such as the renin-angiotensin-aldosterone system (RAAS).29

These findings align with and further extend previous research. Potter E et al30 indicated that GLS is an independent predictor of impaired left ventricular systolic function and can identify systolic dysfunction in patients with preserved ejection fraction. Đorđević D et al31,32 emphasized that the pathophysiological basis of LVH in HHD patients is closely related to subendocardial fibrosis, and reduced GLS often precedes the appearance of LVH, holding significant value in predicting LVH. This study not only confirms that GLS is a sensitive parameter for diagnosing LVH (AUC = 0.91) but also, through T1 mapping, reveals its histological basis, reveals its histological basis, myocardial fibrosis. Moreover, our findings provide preliminary evidence suggesting that this fibrosis may exhibit an “apical-predominant” distribution pattern, thereby providing important spatial insights that complement and validate previous conclusions.

The observed reduction in GLS (median −10.80% in LVH vs −15.75% in Non-LVH) corresponds to a relative impairment of approximately 31%. In the context of hypertensive heart disease, a GLS value worse than −16% is often associated with subclinical systolic dysfunction, and values below −12% have been linked to increased risk of heart failure progression. Our LVH group’s median GLS of −10.80% thus falls within a range indicative of clinically relevant myocardial dysfunction, supporting its utility as an early marker of LV remodeling.

T1 mapping technology enables quantitative measurement of native T1 values in various segments of the left ventricle and calculation of extracellular volume (ECV) using T1 values before and after contrast agent injection combined with hematocrit levels. This allows non-invasive assessment of myocardial interstitial collagen content and fibrosis extent, a method that has been pathologically validated.33 Previous studies have shown that ECV values are generally elevated in HHD patients (eg, 29% in an HHD group vs 27% in a control group in a study of 102 cases), and native T1 values are significantly correlated with left ventricular myocardial stiffness (sensitivity 88%, specificity 64% when global T1 value ≥ 1362 ms).34,35 This study similarly found that global ECV and native T1 values were significantly higher in the LVH group than in the Non-LVH group and were positively correlated with LGE positivity, GLRS, and GSRS, while negatively correlated with GCS and GLS. These results are consistent with those of Webb Jessica et al.36

Our findings have practical implications. First, the apical-predominant fibrosis pattern suggests that hypertensive patients should undergo apical assessment via T1 mapping and ECV.37 Second, GLS—despite its limited sensitivity—serves as a highly specific marker for detecting LVH and underlying fibrosis, supporting its use as a confirmatory tool in patients with suspected advanced myocardial involvement.38 Third, the combined application of CMR-FT and T1 mapping provides a comprehensive non-invasive approach for early detection and spatial characterization of myocardial injury, which could guide personalized monitoring and timely intervention in hypertensive patients at risk of progressive heart disease. Future integration of these imaging biomarkers into risk stratification protocols may enhance the precision of cardiovascular management in hypertension.39

Furthermore, our segmental analysis suggests an “apical-predominant” distribution of fibrosis in hypertension-induced LVH, which, pending further validation, could refine our understanding of the spatial pathophysiology of hypertensive myocardial remodeling.40

The findings of this study should be interpreted within the context of its limitations. First, the retrospective, cross-sectional design precludes the establishment of causality or temporal sequence between hypertension, LVH development, and the observed apical-predominant fibrosis pattern. Second, significant demographic imbalances existed between groups, with the LVH group being younger and predominantly female. Although we adjusted for age and sex statistically, residual confounding cannot be completely ruled out. Third, the use of multiple MRI scanners, despite a standardized protocol and routine phantom calibration, may introduce systematic variability in quantitative T1 and ECV values in the absence of formal cross-scanner harmonization. Fourth, the thinner myocardium of the apical segments makes both T1/ECV quantification and strain measurement via feature tracking more susceptible to partial volume effects and measurement noise. We attempted to mitigate this by excluding the apical cap (segment 17) and using consistent contouring protocols for all analyses. This methodological consideration is directly relevant to the interpretation of the apical-predominant patterns reported in both fibrosis (ECV/T1) and strain. Nevertheless, the consistent finding across multiple independent parameters (strain, T1, ECV) strengthens the robustness of the observed apical predominance. Fifth, clinical data such as the duration and severity of hypertension, as well as details of antihypertensive medication regimens, were not available.41,42 These factors could influence the progression and distribution of myocardial fibrosis and represent unmeasured confounders. The potential influence of specific drug classes on T1 and ECV values was also not explored. Sixth, the observed positive correlation between GLRS (a radial strain parameter) and ECV is physiologically counterintuitive and lacks a clear mechanistic explanation in the current study. It may reflect complex compensatory mechanics or measurement noise, warranting cautious interpretation and further investigation. Seventh, the post-contrast T1 values reported in Table 2, especially in the Non-LVH group, appear non-physiological. This is most likely attributable to technical factors such as variations in the exact scan timing after contrast injection and individual differences in contrast agent kinetics. While this does not affect our primary conclusions based on native T1 and ECV, it highlights the inherent variability of standalone post-contrast T1 measurements and the superiority of ECV for fibrosis assessment.Eighth, a formal sample size or power calculation was not performed a priori, which may affect the generalizability of our findings.Finally, the absence of histopathological correlation remains a limitation for validating the proposed apical fibrosis pattern.

In summary, CMR-FT and T1 mapping offer complementary measures of myocardial dysfunction and fibrosis in hypertensive heart disease. Despite study limitations, impaired GLS strongly associated with LVH, and fibrosis quantified by ECV and native T1 exhibited an apical-predominant distribution. These associations were independent of age and sex differences after statistical adjustment. The apical-predominant pattern, if validated in prospective studies, could offer new insights into the regional pathophysiology of hypertensive remodeling. The combined use of CMR-FT and T1 mapping shows promise as a non-invasive tool for a more comprehensive assessment of myocardial involvement in hypertension.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request and with permission of the ethics committee.

Ethics Approval and Consent to Participate

This study was approved by the Ethics Committee of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (Approval No. K2025-1241). The requirement for informed consent was waived by the same Ethics Committee due to the retrospective nature of the study and the use of anonymized imaging data.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

This work was supported by the Xiaoshan District Science and Technology Plan Guidance Project (Grant No.: 2022329).

Disclosure

The authors declare that they have no competing interests in this work.

References

1. Williams B, Mancia G, Spiering W, et al. ESC/ESH guidelines for the management of arterial hypertension. Eur Heart J. 2018;39(33):3021–3104.

2. Roth GA, Mensah GA, Johnson CO, et al. Global burden of cardiovascular diseases and risk factors, 1990–2019: update from the GBD 2019 study. J Am Coll Cardiol. 2020;76(25):2982–3021.

3. GBD. Risk factors collaborators. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396(10258):1223–1249.

4. Ismail TF, Frey S, Kaufmann BA, et al. Hypertensive heart disease—the imaging perspective. J Clin Med. 2023;12(9):3122. doi:10.3390/jcm12093122

5. Ommen SR, Mital S, Burke MA, et al. AHA/ACC guideline for the diagnosis and treatment of patients with hypertrophic cardiomyopathy: a report of the American college of cardiology/American heart association joint committee on clinical practice guidelines. J Thorac Cardiovasc Surg. 2021;162(1):e23–e106. doi:10.1016/j.jtcvs.2021.04.001

6. Rodrigues JCL, Amadu AM, Ghosh Dastidar A, et al. Nocturnal dipping status and left ventricular hypertrophy: a cardiac magnetic resonance imaging study. J Clin Hypertens. 2018;20(4):784–793. doi:10.1111/jch.13235

7. Messroghli DR, Moon JC, Ferreira VM, et al. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: a consensus statement by the society for cardiovascular magnetic resonance (SCMR) endorsed by the European association for cardiovascular imaging (EACVI). J Cardiovasc Magn Reson. 2017;19(1):75. doi:10.1186/s12968-017-0389-8

8. Bernhard B, Tanner G, Garachemani D, et al. Predictive value of cardiac magnetic resonance right ventricular longitudinal strain in patients with suspected myocarditis. J Cardiovasc Magn Reson. 2023;25(1):49. doi:10.1186/s12968-023-00957-6

9. Schuster A, Hor KN, Kowallick JT, et al. Cardiovascular magnetic resonance myocardial feature tracking: concepts and clinical applications. Circ Cardiovasc Imaging. 2016;9(4):e004077. doi:10.1161/CIRCIMAGING.115.004077

10. Rudolph A, Abdel-Aty H, Bohl S, et al. Noninvasive detection of fibrosis applying contrast-enhanced cardiac magnetic resonance in different forms of left ventricular hypertrophy relation to remodeling. J Am Coll Cardiol. 2009;53(3):284–291. doi:10.1016/j.jacc.2008.08.064

11. Seğmen F, Aydemir S, Küçük O, et al. The roles of vitamin d levels, gla-rich protein (GRP) and matrix gla protein (MGP), and inflammatory markers in predicting mortality in intensive care patients: a new biomarker link?. Metabolites. 2024;14(11):620. doi:10.3390/metabo14110620

12. Kurtulus B, Atilgan N, Yilmaz M, et al. Two members of vitamin-k-dependent proteins, gla-rich protein (GRP) and matrix gla protein (MGP), as possible new players in the molecular mechanism of osteoarthritis. J Clin Med. 2024;13(17):5159. doi:10.3390/jcm13175159

13. De Haas HJ, Arbustini E, Fuster V, et al. Molecular imaging of the cardiac extracellular matrix. Circ Res. 2014;114(5):903–915. doi:10.1161/CIRCRESAHA.113.302680

14. Isaak A, Luetkens JA, Faron A, et al. Free-breathing non-contrast flow-independent cardiovascular magnetic resonance angiography using cardiac gated, magnetization-prepared 3D Dixon method: assessment of thoracic vasculature in congenital heart disease. J Cardiovascu Magnetic Resonance. 2021;23(1):91. doi:10.1186/s12968-021-00788-3

15. Zhang X, Cui C, Zhao S, et al. Cardiac magnetic resonance radiomics for disease classification. Eur Radiol. 2023;33(4):2312–2323. doi:10.1007/s00330-022-09236-x

16. Ali-Ahmed F, Dalgaard F, Al-Khatib SM. Sudden cardiac death in patients with myocarditis: evaluation, risk stratification, and management. Am Heart J. 2020;220:29–40. doi:10.1016/j.ahj.2019.08.007

17. Jia G, Sowers JR. Hypertension in diabetes: an update of basic mechanisms and clinical disease. Hypertension. 2021;78(5):1197–1205. doi:10.1161/HYPERTENSIONAHA.121.17981

18. Modin D, Biering-Sørensen SR, Møgelvang R, et al. Prognostic value of left atrial strain in predicting cardiovascular morbidity and mortality in the general population. Eur Heart J Cardiovasc Imaging. 2019;20(7):804–815. doi:10.1093/ehjci/jey181

19. Frąk W, Wojtasińska A, Lisińska W, et al. Pathophysiology of cardiovascular diseases: new insights into molecular mechanisms of atherosclerosis, arterial hypertension, and coronary artery disease. Biomedicines. 2022;10(8):1938. doi:10.3390/biomedicines10081938

20. Drazner MH. The progression of hypertensive heart disease. Circulation. 2011;123(3):327–334. doi:10.1161/CIRCULATIONAHA.108.845792

21. Yang Y, Yin G, Jiang Y, et al. Quantification of left atrial function in patients with non-obstructive hypertrophic cardiomyopathy by cardiovascular magnetic resonance feature tracking imaging: a feasibility and reproducibility study. J Cardiovasc Magn Reson. 2020;22(1):1. doi:10.1186/s12968-019-0589-5

22. Ito H, Ishida M, Makino W, et al. Cardiovascular magnetic resonance feature tracking for characterization of patients with heart failure with preserved ejection fraction: correlation of global longitudinal strain with invasive diastolic functional indices. J Cardiovasc Magn Reson. 2020;22(1):42. doi:10.1186/s12968-020-00636-w

23. Zhou WN, Brown JM, Bajaj NS, et al. Hypertensive coronary microvascular dysfunction: a subclinical marker of end organ damage and heart failure. Eur Heart J. 2020;41(25):2366–2375. doi:10.1093/eurheartj/ehaa191

24. Pezeshki PS, Ghorashi SM, Houshmand G, et al. Feature tracking cardiac magnetic resonance imaging to assess cardiac manifestations of systemic diseases. Heart Fail Rev. 2023;28(5):1189–1199. doi:10.1007/s10741-023-10321-6

25. Constant Dit Beaufils AL, Huttin O, Jobbe-Duval A, et al. Replacement myocardial fibrosis in patients with mitral valve prolapse: relation to mitral regurgitation, ventricular remodeling, and arrhythmia. Circulation. 2021;143(18):1763–1774. doi:10.1161/CIRCULATIONAHA.120.050214

26. Pasipoularides A. Mechanotransduction mechanisms for intraventricular diastolic vortex forces and myocardial deformations: part 2. J Cardiovascu Transl Res. 2015;8(5):293–318. doi:10.1007/s12265-015-9630-8

27. González A, Ravassa S, López B, et al. Myocardial remodeling in hypertension: toward a new view of hypertensive heart disease. Hypertension. 2018;72(3):549–558. doi:10.1161/HYPERTENSIONAHA.118.11125

28. Selthofer-Relatic K, Mihalj M, Kibel A, et al. Coronary microcirculatory dysfunction in human cardiomyopathies: a pathologic and pathophysiologic review. Cardiol Rev. 2017;25(4):165–178. doi:10.1097/CRD.0000000000000140

29. Jia G, Aroor AR, Hill MA, et al. Role of renin-angiotensin-aldosterone system activation in promoting cardiovascular fibrosis and stiffness. Hypertension. 2018;72(3):537–548. doi:10.1161/HYPERTENSIONAHA.118.11065

30. Potter E, Marwick TH. Assessment of left ventricular function by echocardiography: the case for routinely adding global longitudinal strain to ejection fraction. JACC Cardiovasc Imaging. 2018;11(2 Pt 1):260–274. doi:10.1016/j.jcmg.2017.11.017

31. Đorđević DB, Koračević GP, Đorđević AD, et al. Hypertension and left ventricular hypertrophy. J Hypertens. 2024;42(9):1505–1515. doi:10.1097/HJH.0000000000003774

32. Kramer CM, Barkhausen J, Bucciarelli-Ducci C, et al. Standardized cardiovascular magnetic resonance imaging (CMR) protocols: 2020 update. J Cardiovasc Magn Reson. 20