Sequential Changes in Right Ventricular Strain During the Development of Pulmonary Hypertension Assessed With Magnetic Resonance Feature Tracking: An Animal Study

Abstract

Objective

To evaluate the sequential changes in right ventricular (RV) strain and histological vascular remodeling and their correlations during pulmonary hypertension (PH) development in a rat model by using cardiac magnetic resonance (CMR)-feature tracking (FT).

Materials and Methods

Separate cohorts of six-week-old male Wistar rats (10 per group) were allocated to a baseline group, PH induction groups assessed at 1, 2, 3, or 4 weeks after monocrotaline injection, and vehicle-injected control groups assessed at 2 and 4 weeks. Each rat underwent a single magnetic resonance imaging (MRI) examination and was subsequently euthanized for histological analysis. Cine images were acquired using a 9.4T MRI scanner, and strain was analyzed using the FT technique. Histological analysis was used to quantify pulmonary arterial wall thickness percentage (PA %WT) and RV collagen density. Overall group differences were analyzed using the Kruskal–Wallis test, which was followed by the Mann–Whitney U test with Bonferroni correction for pairwise comparisons with the baseline group. Correlations between the variables were assessed using Spearman’s correlation coefficients.

Results

All CMR-derived RV parameters showed significant differences across the baseline and PH induction groups (all P < 0.001). RV free wall longitudinal strain (RVFWLS) showed the earliest impairment at 2 weeks, with a progressive decline thereafter. The RV free wall circumferential strain (RVFWCS) and RV ejection fraction were reduced beginning at 3 weeks, whereas the RV end-diastolic volume index increased significantly only at 4 weeks. The PA %WT increased progressively from week 2, whereas the RV collagen density did not change significantly. RVFWLS and RVFWCS showed strong positive correlations with PA %WT (r = 0.885 and r = 0.757, respectively; both P < 0.001), which were primarily driven by parallel changes across the post-induction time points. The control groups showed no significant temporal changes in the CMR or histological measurements.

Conclusion

In this preclinical model of PH, staged CMR and histological assessments demonstrated sequential RV remodeling in parallel with PA remodeling, with RVFWLS impairment preceding overt RV dysfunction and enlargement.

INTRODUCTION

Pulmonary hypertension (PH) is characterized by progressive changes in the pulmonary vasculature, which are accompanied by adaptive right ventricular (RV) responses to an increasing vascular load [1]. With sustained disease progression, compensatory RV remodeling may eventually become insufficient, causing overt RV dysfunction and failure, which remain the principal determinants of mortality in patients with PH [2,3,4]. As pulmonary vascular remodeling and RV functional adaptation evolve in parallel, interest in evaluating the pulmonary artery (PA) and right ventricle as integrated cardiopulmonary units has increased [5].

Cardiac magnetic resonance (CMR) is the reference standard for assessing RV morphology and function. CMR-based feature tracking (FT) is a clinically feasible postprocessing technique that enables the quantification of myocardial deformation from routine cine images. Previous studies have demonstrated impaired FT-derived RV strain even in patients with PH and preserved RV ejection fraction (RVEF), and RV strain has been shown to be associated with right ventricle–PA interactions assessed using invasive hemodynamic measurements [6,7,8]. These findings suggest that RV strain may provide a sensitive imaging marker of right ventricle–PA interactions in PH. Early non-invasive detection of altered cardiopulmonary interaction is particularly relevant in high-risk populations, such as those with connective tissue disease or genetic predisposition [9].

However, the relative timing of the changes in RV strain, functional parameters, and pulmonary vascular remodeling during PH development remains insufficiently characterized. We hypothesized that these changes would occur in a distinct temporal sequence, with the RV strain demonstrating early alterations.

Therefore, the present study aimed to evaluate the sequential changes in RV strain and histological vascular remodeling and their correlation across successive time points during PH development in a rat model of monocrotaline-induced PH.

MATERIALS AND METHODS

Animal Model

This study was approved by the Institutional Animal Care and Use Committee of the Yonsei University College of Medicine (approval number: 2021-0103) and was conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Six-week-old male Wistar rats were allocated to the baseline group; PH induction groups assessed at 1, 2, 3, and 4 weeks; or control groups assessed at 2 and 4 weeks. Because histological analysis required tissue harvesting, separate cohorts were studied at each time point. Ten rats were included in each group on the basis of the resource equation method for exploratory animal studies [10].

To induce PH, monocrotaline (40 mg/kg; Sigma Chemical, St. Louis, MO, USA) was administered as a single subcutaneous injection to the rats in the PH induction groups at baseline. The predefined time points were selected to represent different stages of pulmonary vascular remodeling.

To account for age-related changes, control rats received a single subcutaneous injection of the vehicle solution used for monocrotaline preparation (1 M HCl, distilled water, 1 M NaOH, and 0.9% saline adjusted to pH 7.4) at baseline, and underwent imaging and histological assessments 2 or 4 weeks later.

During the experiment, seven monocrotaline-injected rats were excluded: four died before their scheduled euthanasia (one at 2 weeks, one at 3 weeks, and two at 4 weeks), and three were excluded due to poor image quality (one at 2 weeks and two at 4 weeks). A total of 77 rats were used to achieve a predefined target sample size of 10 rats per group, of which 70 rats (10 baseline, 40 PH induction, and 20 control rats) completed the study protocol and were included in the final analysis.

CMR Imaging and Analysis

Each rat underwent a single CMR scan at its designated time point. Imaging was performed under 2% isoflurane anesthesia using a 9.4T small animal MRI scanner (Bruker BioSpin, Billerica, MA, USA) equipped with an 86-mm volume transmit coil and a 4-channel phased-array surface coil. Body weight was measured immediately before each CMR scan to calculate the body surface area [11]. Cine images were acquired in two- and four-chamber views and short-axis stacks from the RV base to the apex by using a fast low-angle shot sequence with a self-gated technique (IntraGate, Bruker BioSpin, ParaVision 5.1), which was adopted to compensate for the rats’ rapid heart rates (approximately 500 beats per minute). Imaging parameters were as follows: repetition time, 5.3 ms; echo time, 2.1 ms; flip angle, 15°; field of view, 50 × 50 mm; matrix, 128 × 128; slice thickness, 2 mm; number of slices, 8–9; and 25 cardiac phases.

All images were analyzed using a commercially available software (cvi42; Circle Cardiovascular Imaging, Calgary, Canada). Two experienced research assistants with three and five years of experience in cardiovascular image analysis who were proficient with the software independently performed the image analyses while blinded to group allocation. For volumetric analysis, the endocardial borders of the right ventricle were manually traced on end-diastolic and end-systolic short-axis images to calculate the RV end-diastolic volume (RVEDV), RV end-systolic volume, and RVEF. For strain analysis, both the endocardial and epicardial borders of the RV free wall were manually delineated on reference four-chamber and short-axis cine images at end-diastole (Fig. 1). FT was then performed using a built-in software algorithm, which automatically propagated the contours over the cardiac cycle. All tracking results were visually inspected, and manual corrections were applied when necessary to ensure accuracy. RV free wall longitudinal strain (RVFWLS) and RV free wall circumferential strain (RVFWCS) were obtained from the tracked contours, excluding the interventricular septum to avoid the influence of left ventricular contraction.

Fig. 1. Example of RV strain measurement in a rat 2 weeks after monocrotaline injection. The endocardial (yellow) and epicardial (white) borders of the RV free wall were manually delineated on the end-diastole frame of four-chamber (A) and short-axis (B) cine images, and these contours were automatically propagated throughout the cardiac cycle. Corresponding strain curves demonstrating an RV free wall longitudinal strain of -19.51% (C) and circumferential strain of -22.28% (D). RV = right ventricular.

Histological Analysis

After CMR, the rats were euthanized, and their hearts were resected, sectioned at approximately 2-mm thickness at the midway point between the apex and base of the right ventricle in the short-axis plane, and fixed in 4% formaldehyde for 24 hours. The lungs were inflated with 4% formaldehyde through the trachea at a constant pressure, fixed for 24 hours, and sectioned transversely at the level of the main PA bifurcation into slices approximately 2 mm thick. Subsequently, the specimens were embedded in paraffin, and 5-µm sections were prepared. The sections were stained with hematoxylin and eosin to assess general histology and with Masson’s trichrome to evaluate the right ventricle and PA. The slides were digitally scanned at x400 magnification (ImageScope v10.0; Aperio Technologies, Vista, CA, USA).

Using ImageJ software (National Institutes of Health, Bethesda, MD, USA), the number of pixels corresponding to collagen and myocardial tissue was quantified from the middle one-third of the RV free wall, and collagen density was calculated as the percentage of collagen pixels relative to the total number of myocardial pixels. From each lung specimen, PAs with an outer diameter of approximately 200 µm were identified, and ten vessels that appeared closest to the true transverse sections were selected. The outer and inner diameters were measured as mean perpendicular distances across the outer adventitial and luminal borders, respectively. Pulmonary arterial wall thickness percentage (PA %WT) was calculated as (outer diameter - inner diameter)/outer diameter ×100 [12,13,14,15]. Because this measurement is based on the entire vessel wall rather than the medial layer alone, PA %WT was used as a morphometric index of PA remodeling encompassing the intimal, medial, and adventitial layers [16]. The median value of ten measurements was used to represent the PA %WT for each rat.

Statistical Analysis

Continuous variables were presented as mean ± standard deviation. The interobserver agreement for strain measurements was assessed using the intraclass correlation coefficient, which was interpreted as follows: <0.5, poor; 0.50–0.75, moderate; 0.75–0.90, good; and >0.9, excellent [17]. After assessing the interobserver agreement, the mean of the two observers’ measurements was used for subsequent analyses. The overall differences between the baseline and four PH induction groups, as well as between the baseline and two control groups, were assessed using the Kruskal–Wallis test, with statistical significance defined as P < 0.05. For variables with significant overall differences, pairwise comparisons between the baseline group and each PH induction or control group were performed using the Mann–Whitney U test. To account for multiple comparisons, Bonferroni-corrected significance thresholds of P < 0.0125 (0.05/4) for comparisons between the baseline and PH induction groups and P < 0.025 (0.05/2) for comparisons between the baseline and control groups were pre-specified. For descriptive purposes, the earliest change was operationally defined as the first post-induction time point showing a statistically significant difference from the baseline. Spearman’s rank coefficient was used to assess correlations among CMR-derived parameters, RV collagen density, and PA %WT. All statistical analyses were performed using SPSS software (version 29.0; IBM Corp., Armonk, NY, USA).

RESULTS

CMR-Derived RV Parameters in PH Induction Groups

CMR-derived RV parameters in the baseline and PH induction groups are summarized in Table 1. Body weight increased across the baseline and post-induction time points (baseline, 219.40 ± 13.18 g; 1 week, 249.50 ± 8.66 g; 2 weeks, 285.80 ± 11.56 g; 3 weeks, 320.20 ± 10.85 g; and 4 weeks, 332.00 ± 7.44 g).

Table 1. CMR-derived RV parameters in PH induction groups in comparison with the baseline.

	Baseline group (n = 10)	1-week group (n = 10)	2-week group (n = 10)	3-week group (n = 10)	4-week group (n = 10)	P
RVFWLS, %	-23.72 ± 0.87	-23.38 ± 1.09	-19.89 ± 1.93*	-17.42 ± 1.59*	-15.12 ± 1.16*	<0.001
RVFWCS, %	-22.04 ± 1.26	-22.41 ± 1.85	-22.03 ± 2.28	-18.82 ± 1.87*	-16.08 ± 1.21*	<0.001
RVEF, %	64.23 ± 3.01	64.70 ± 3.16	62.80 ± 2.70	59.24 ± 2.88*	52.75 ± 3.69*	<0.001
RVEDV index, mL/m2	10.48 ± 0.51	10.69 ± 0.55	10.91 ± 0.55	10.98 ± 0.58	12.54 ± 0.63*	<0.001

Values are presented as means ± standard deviations. P-values were calculated using the Kruskal–Wallis test.

^*Significant difference from pairwise comparison with the baseline group (Bonferroni-corrected P < 0.0125). CMR = cardiac magnetic resonance, RV = right ventricular, PH = pulmonary hypertension, RVFWLS = right ventricular free wall longitudinal strain, RVFWCS = right ventricular free wall circumferential strain, RVEF = right ventricular ejection fraction, RVEDV = right ventricular end-diastolic volume

All CMR-derived RV parameters showed significant overall differences across the baseline and PH induction groups (all P < 0.001) (Table 1). RVFWLS remained unchanged at 1 week (baseline, -23.72% ± 0.87% vs. 1 week, -23.38% ± 1.09%; P = 0.403), and showed a reduction at 2 weeks (-19.89% ± 1.93%, P < 0.001), with further reductions at 3 weeks (-17.42% ± 1.59%, P < 0.001) and at 4 weeks (-15.12% ± 1.16%, P < 0.001) (Figs. 2, 3A).

Fig. 2. Representative cases of a control rat (A) and a monocrotaline-injected rat (B) 2 weeks after baseline. The upper row shows four-chamber images at end-diastole and end-systole. The lower row shows corresponding color maps of RV free wall longitudinal strain derived from feature-tracking analysis. Although RV size and global morphology appear similar between the two animals, the monocrotaline-injected rat demonstrates reduced RV free wall longitudinal strain compared with the control rat (-19.7% vs. -23.5%). This difference is visually reflected in the corresponding end-systolic strain map. Corresponding cine image loops throughout the cardiac cycle are provided as Supplementary Videos 1, 2, 3, 4. RV = right ventricular.

Fig. 3. Temporal changes in cardiac magnetic resonance-derived RV parameters across baseline and pulmonary hypertension induction groups. A: RVFWLS. B: RVFWCS. C: RVEF. D: RVEDV index. Data are presented as box-and-whisker plots, with the mean value of each group connected by lines to illustrate temporal trends. P-values indicate pairwise comparisons with the baseline group (Bonferroni-corrected significance threshold, P < 0.0125). RV = right ventricular, RVFWLS = right ventricular free wall longitudinal strain, RVFWCS = right ventricular free wall circumferential strain, RVEF = right ventricular ejection fraction, RVEDV = right ventricular end-diastolic volume.

RVFWCS remained unchanged through 2 weeks (baseline, -22.04% ± 1.26% vs. 1 week, -22.41% ± 1.85% [P = 0.529]; vs. 2 weeks, -22.03% ± 2.28% [P = 0.928]). A significant reduction in RVFWCS was first observed at 3 weeks (-18.82% ± 1.87%, P < 0.001), with a further reduction at 4 weeks (-16.08% ± 1.21%, P < 0.001) (Fig. 3B). RVEF showed a reduction at 3 weeks (baseline, 64.23% ± 3.01% vs. 3 weeks, 59.24% ± 2.88%; P = 0.003), with further reduction at 4 weeks (52.75% ± 3.69%, P < 0.001) (Fig. 3C). The RVEDV index remained unchanged through 3 weeks and showed an increase at 4 weeks (baseline, 10.48 ± 0.51 mL/m² vs. 4 weeks, 12.54 ± 0.63 mL/m², P < 0.001) (Fig. 3D).

Interobserver agreement was good for both strain measurements, with intraclass correlation coefficients of 0.844 (95% confidence interval [CI]: 0.741–0.909) for RVFWLS and 0.794 (95% CI: 0.663–0.877) for RVFWCS.

Histological Findings in PH Induction Groups

The histological measurements are summarized in Table 2. After monocrotaline injection, lung histological assessments demonstrated features consistent with pulmonary vascular remodeling, particularly in rats analyzed 3 and 4 weeks post-induction. Small-to-medium-sized PAs showed varying degrees of wall thickening and luminal narrowing. The perivascular regions exhibited inflammatory cell infiltration, and thickening of the alveolar septa was occasionally observed (Fig. 4). Histological analysis revealed no remarkable structural alterations in the RV myocardium following monocrotaline injection (Fig. 5). Although RV cavity enlargement was observed at 4 weeks, myocyte hypertrophy and interstitial changes were not prominent in this group. The PA %WT increased significantly beginning at 2 weeks, from 12.58% ± 0.58% at baseline to 23.86% ± 1.26% at 2 weeks, 35.74% ± 1.60% at 3 weeks, and 45.92% ± 3.28% at 4 weeks (overall P < 0.001) (Fig. 6A). In contrast, the RV collagen density, a quantitative marker of myocardial fibrosis, showed no significant difference across the baseline and PH induction groups (P = 0.201) (Fig. 6B).

Table 2. Histological measurements in the PH induction groups in comparison with the baseline.

	Baseline group (n = 10)	1-week group (n = 10)	2-week group (n = 10)	3-week group (n = 10)	4-week group (n = 10)	P
PA %WT	12.58 ± 0.58	12.74 ± 0.63	23.86 ± 1.26*	35.74 ± 1.60*	45.92 ± 3.28*	<0.001
RV collagen density, %	3.72 ± 0.38	3.70 ± 0.33	3.90 ± 0.26	3.87 ± 0.33	4.02 ± 0.27	0.201

Values are presented as means ± standard deviations. P-values were calculated using the Kruskal–Wallis test.

^*Significant difference from pairwise comparison with the baseline group (Bonferroni-corrected P < 0.0125).

PH = pulmonary hypertension, PA %WT = pulmonary arterial wall thickness percentage, RV = right ventricular

Fig. 4. Representative histological images of pulmonary arteries. Masson’s trichrome–stained lung sections showed increasing pulmonary arterial wall thickness across baseline and PH induction groups: 12.85% in the baseline group (A) and 12.93%, 23.78%, 35.40%, and 45.12% in the 1-, 2-, 3-, and 4-week PH induction groups (B-E), respectively. Perivascular inflammatory cell infiltration and thickening of alveolar septa were also observed (scale bar = 200 µm). PH = pulmonary hypertension.

Fig. 5. Representative histological images of the right ventricle. In Masson’s trichrome–stained heart sections, blue-stained areas indicate myocardial fibrosis, which was limited to small focal regions across all groups: 3.75% in the baseline group (A), and 4.09%, 3.84%, 4.12%, and 4.25% in the 1-, 2-, 3-, and 4-week PH induction groups (B-E), respectively. Myocyte hypertrophy and interstitial fibrosis were not prominent within 4 weeks of PH induction. Right ventricular enlargement was observed in the 4-week PH induction group. PH = pulmonary hypertension.

Fig. 6. Temporal changes in histological findings across the baseline and pulmonary hypertension induction groups. A: PA %WT. B: RV collagen density. Data are presented as box-and-whisker plots, with the mean value of each group connected by lines to illustrate temporal trends. In (A), P-values indicate pairwise comparisons with the baseline group (Bonferroni-corrected significance threshold, P < 0.0125); in (B), the P-value indicates the overall comparison across groups (Kruskal–Wallis test). PA %WT = pulmonary arterial wall thickness percentage, RV = right ventricular.

Results in Control Groups

Comparisons between the baseline and control groups are presented in Table 3. Body weight increased across the baseline and control groups (baseline, 219.40 ± 13.18 g; 2-week control group, 288.20 ± 15.46 g; 4-week control group, 342.90 ± 13.10 g).

Table 3. CMR-derived RV parameters and histological measurements in the control groups in comparison with the baseline.

	Baseline group (n = 10)	2-week control group (n = 10)	4-week control group (n = 10)	P
RVFWLS, %	-23.72 ± 0.87	-23.30 ± 1.35	-23.50 ± 1.51	0.714
RVFWCS, %	-22.04 ± 1.26	-22.42 ± 1.72	-21.77 ± 2.20	0.627
RVEF, %	64.23 ± 3.01	64.56 ± 3.57	63.40 ± 2.97	0.736
RVEDV index, mL/m2	10.48 ± 0.51	10.42 ± 0.65	10.50 ± 0.47	0.897
PA %WT	12.58 ± 0.58	12.49 ± 0.90	12.90 ± 0.24	0.132
RV collagen density, %	3.72 ± 0.38	3.89 ± 0.25	3.95 ± 0.28	0.215

Values are presented as means ± standard deviations. P-values were calculated using the Kruskal–Wallis test.

CMR = cardiac magnetic resonance, RV = right ventricular, RVFWLS = right ventricular free wall longitudinal strain, RVFWCS = right ventricular free wall circumferential strain, RVEF = right ventricular ejection fraction, RVEDV = right ventricular end-diastolic volume, PA %WT = pulmonary arterial wall thickness percentage

Both RVFWLS and RVFWCS remained stable across the baseline and the 2- and 4-week control groups, with no significant overall differences (P = 0.714 and P = 0.627, respectively) (Table 3). Similarly, no significant overall differences in the RVEF or RVEDV indices were observed across the baseline and control groups (P = 0.736 and P = 0.897, respectively). No significant differences in PA %WT (P = 0.132) or RV collagen density (P = 0.215) were observed across the baseline and control groups.

Correlation of RV Strain Values With Histological Findings

In the pooled analyses across all time points, both strain parameters demonstrated strong positive correlations with PA %WT (r = 0.885 for RVFWLS and r = 0.757 for RVFWCS; both P < 0.001). Inspection of the scatter plots showed that the RV strain and PA %WT exhibited concurrent changes across groups rather than consistent correlations within individual time points (Fig. 7). Within each time point, the PA %WT values were relatively clustered within a narrow range, whereas the RV strain values showed a wider dispersion, indicating greater inter-individual variability. Across the post-induction time points, both PA %WT and RV strain values shifted accordingly, as reflected in their mean values.

Fig. 7. Scatterplots with Spearman’s correlation coefficients between the RVFWLS and PA %WT (A), and between the RVFWCS and PA %WT (B). Red, yellow, green, blue, and purple dots represent data from the baseline group and the 1-, 2-, 3-, and 4-week pulmonary hypertension induction groups, respectively. RVFWLS = right ventricular free wall longitudinal strain, PA %WT = pulmonary arterial wall thickness percentage, RVFWCS = right ventricular free wall circumferential strain.

In comparison, the RVFWLS showed a weak correlation with RV collagen density (r = 0.317, P = 0.025), whereas the RVFWCS did not show a significant correlation (r = 0.260, P = 0.068). Both RVFWLS and RVFWCS showed significant correlations with the RVEDV index (r = 0.652 and 0.653, respectively; both P < 0.001) and negative correlations with RVEF (r = -0.681 and -0.647, respectively; both P < 0.001). Similar to PA %WT, these correlations were not consistently present within individual time points but were primarily attributable to differences across groups (Supplementary Fig. 1). Time-point–stratified Spearman correlation analyses are provided in Supplementary Table 1.

DISCUSSION

In this preclinical study, we characterized the changes in RV free wall strain using CMR-FT in a monocrotaline-induced rat model of PH. The key observations were the temporal sequence of RV and PA remodeling across post-induction time points. Impairment of RVFWLS was first observed at the 2-week post-induction time point, and was accompanied by an increase in PA %WT, whereas a reduction in RVEF was observed later. In the pooled analyses, both RVFWLS and RVFWCS were significantly correlated with PA %WT.

Because PH typically has an insidious onset and is often diagnosed years after symptom onset [18], it is inherently difficult to longitudinally assess early disease stages in humans. In addition, the direct assessment of histological vascular remodeling is not feasible in clinical settings. In this context, experimental animal models provide an opportunity to examine the changes in RV deformation in relation to pulmonary vascular remodeling at different stages of disease development. The monocrotaline rat model is widely used as an experimental model of PH because of its reproducibility and well-characterized pathophysiology [19,20,21]. Monocrotaline induces pulmonary vascular remodeling, which is characterized by the proliferation of PA smooth muscle cells [22], with progressive increments in pressure overload. Although the direct temporal translation between this accelerated disease model and the more gradual course of human PH is limited, the stage-dependent changes observed in this study provide insights into the evolution of RV deformation and PA remodeling during the development of PH. In particular, RVFWLS showed earlier changes than RVEF in this experimental setting. From a clinical perspective, these findings have potential implications for conditions in which PH develops gradually, such as connective tissue disease, congenital heart disease, and other high-risk conditions [9]. In such contexts, RV strain may have potential as a non-invasive imaging marker for detecting early alterations in RV function or right ventricle–PA interactions. However, further validation in longitudinal human studies is required prior to clinical application.

An important pathophysiological implication of this study is the concurrent evolution of RV strain impairment and PA remodeling. The absence of a significant increase in RV collagen density despite early RV strain deterioration further supports a load-driven mechanism rather than structural myocardial remodeling. From a physiological standpoint, RV strain is known to be influenced by loading conditions, particularly in the setting of increased pulmonary vascular load [8,23,24], and the concurrent changes observed in RV strain and PA %WT across post-induction time points are consistent with a relationship mediated by increasing vascular load. However, because the pooled correlations were primarily driven by differences across post-induction time points rather than by consistent within-time-point relationships, they should be interpreted as reflecting parallel changes over time rather than a direct cross-sectional relationship.

The earlier reduction in RVFWLS in comparison with RVFWCS observed in this study may be related to differences in myocardial fiber mechanics. Longitudinal shortening has been reported to be more sensitive than circumferential shortening to changes in loading conditions [24,25], which may explain why RVFWLS changes were observed at relatively earlier stages in this model. In contrast, circumferential shortening may become more prominent at later stages, potentially reflecting adaptive changes in myocardial contraction patterns under sustained pressure overload [26,27]. However, these interpretations are speculative and require further investigation.

In this study, only the RV free wall strain was analyzed to isolate right ventricle-specific deformation and minimize the influence of left ventricular mechanics, although the interventricular septum contributes to RV systolic performance [28]. The radial strain was not analyzed because the CMR-FT-derived radial strain is less reproducible and more dependent on the contouring accuracy and specific software algorithms [29]. Additional considerations are required when interpreting the RV strain measurements. Although RV strain has emerged as a sensitive imaging parameter, standardized reference values and diagnostic thresholds have not yet been established. Previous studies using speckle-tracking echocardiography have suggested a lower normal limit of approximately -20% for global RV longitudinal strain [30,31], although substantial variability exists across CMR-based measurement techniques [32]. Intervendor variability in the FT software may also affect strain quantification [33,34], and consistency in analytical platforms is recommended for longitudinal assessments.

This study had several limitations. First, the lack of invasive hemodynamic measurements precluded direct assessment of PA pressure and its relationship with RV strain. Second, instead of a longitudinal follow-up, separate groups were analyzed at predefined time points to represent the disease stages. Although this cross-sectional design was chosen for feasibility and to minimize repeated anesthesia, it limited the assessment of within-subject temporal changes. Third, although reproducible, the monocrotaline model does not fully replicate the heterogeneity of human PH, and may involve nonspecific toxic effects on the heart and lungs [19,21].

In conclusion, this preclinical study delineated the temporal sequence of CMR-derived RV changes and stage-matched histological changes during PH development. RV strain impairment emerged concurrently with pulmonary vascular remodeling and preceded overt RV dysfunction and enlargement. These findings provide a basis for further investigation of RV strain as a non-invasive marker of early PH-related RV changes, particularly in longitudinal human studies.

Availability of Data and Material

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.

Supplement

The Supplement is available with this article at https://doi.org/10.3348/kjr.2026.0774.

Video 1

Four-chamber cine MRI of a control rat at 2 weeks.

Video 2

Four-chamber cine MRI with feature-tracking analysis of a control rat at 2 weeks.

Video 3

Four-chamber cine MRI of a monocrotaline-injected rat at 2 weeks.

Video 4

Four-chamber cine MRI with feature-tracking analysis of a monocrotaline-injected rat at 2 weeks.