Value of Right and Left Ventricular T1 and T2 Blood Pool Mapping in Patients with Chronic Thromboembolic Hypertension before and after Balloon Pulmonary Angioplasty

Background: Parametric imaging has taken a steep rise in recent years and non-cardiac applications are of increasing interest. Therefore, the aim of our study was to assess right (RV) and left ventricular (LV) blood pool T1 and T2 values in patients with chronic thromboembolic pulmonary hypertension (CTEPH) compared to control subjects and their correlation to pulmonary hemodynamic. Methods: 26 patients with CTEPH (mean age 64.8 years ± 12.8 SD; 15 female), who underwent CMR and right heart catheterization (RHC) before and 6-months after balloon pulmonary angioplasty (BPA), were retrospectively included. Ventricular blood pool values were measured, compared to control subjects (mean age 40.5 years ± 12.8 SD; 16 female) and correlated to invasive measures (CI, mPAP, PVR). Results: In both, control subjects and CTEPH patients, RVT1 and RVT2 were significantly reduced compared to LVT1 and LVT2. Compared to control subjects, RVT2 was significantly reduced in CTEPH patients (p = 0.0065) and increased significantly after BPA (p = 0.0048). Moreover, RVT2 was positively correlated with CI and negatively correlated with mPAP and PVR before (r = 0.5155, r = −0.2541, r = −0.4571) and after BPA (r = 0.4769, r = −0.2585, r = −0.4396). Conclusion: Ventricular blood pool T2 mapping might be novel non-invasive CMR imaging marker for assessment of disease severity, prognosis, follow-up and even therapy monitoring in PH.


Introduction
Due to the outstanding success of parametric imaging techniques in the diagnosis of various cardiac diseases, parametric imaging-native T1 and T2 mapping and also quantification of extracellular volume (ECV)-has already found its way into diagnostic recommendations [1]. This has meanwhile favored, that parametric imaging is subject of multiple studies in patients with different etiologies of pulmonary hypertension (PH). So far, pathological elevations of cardiac native T1 time and extracellular volume (ECV) compared to healthy volunteers, promising correlations to functional and hemodynamic parameters and influences on patient outcome and survival have already been demonstrated [2][3][4][5][6][7][8]. Even more, parametric imaging seems to be suitable to investigate therapy effects-for example, of balloon pulmonary angioplasty (BPA)-on cardiac function and pulmonary 2 of 10 hemodynamic in chronic thromboembolic pulmonary hypertension (CTEPH) [9]. The ongoing research with all these previous efforts and published results make it clear that parametric imaging as a component of non-invasive cardiac imaging has still arrived beside volumetric measures in patients with PH.
In addition to purely cardiac fields of application, non-cardiac tissues are also increasingly being examined by parametric imaging. For example, a very recent study by Guo et al. investigated the value of hepatic T1 mapping in patients with PH and demonstrated, that hepatic T1 values were predictive for adverse cardiovascular events [10]. Furthermore, Tilman et al. investigated right ventricular and left ventricular T2 blood pool values in patients with left-to-right shunting and showed that the "right-to-left-ventricular blood pool T2 ratio-RVT2/LVT2" might be a novel imaging biomarker for detection of shunts without the need for additive phase-contrast acquisitions [11].
In particular, the approach to investigate native T1 and T2 blood pool values and ratios is completely new and seems very interesting, because previous studies have shown that blood T2 is sensitive to the level of blood oxygenation and quantitative T2 Mapping as a novel, non-invasive method enables estimation of blood oxygen saturation [12,13]. Since blood oxygen saturation is reduced in patients with PH, it would be interesting to find out to what extent parametric imaging is able to distinguish between PH patients and healthy subjects as well as in the context of therapies. Therefore, the aim of our study was to investigate left and right ventricular native T1 and T2 blood pool values and their ratios in healthy controls and in patients with CTEPH before and after BPA in correlation to hemodynamic parameters.

Patient Population
A total of 26 consecutive CTEPH patients (15 female; mean age 64.8 ± 12.8 years ± standard deviation (SD)) were included in this retrospective cohort study from February 2014 to September 2017. Table 1 presents the comorbidities of the CTEPH patients. Cardiac magnetic resonance (CMR) imaging was performed in all patients for clinical reasons as part of the routine examinations before and after BPA. The primary diagnosis of CTEPH was based on the results of the imaging and hemodynamic workup, which included ventilationperfusion scintigraphy (V/Q-SPECT), computed tomography pulmonary angiography (CTPA) or dual-energy computed tomography (DECT), right-heart-catheterization (RHC) measurements and conventional pulmonary angiography. Values are expressed as mean ± SD or as n (%).
To reduce contrast media and magnetic field risks for healthy volunteers, the first 26 patients (16 female; mean age of 40.5 ± 12.8 years SD), who had undergone cardiac MRI for exclusion of myocardial inflammation between October 2022 and December 2022, were selected as control subjects. The control subjects were only included in the control group if both cardiac MRI examination and patient follow-up were completely unremarkable (normal heart size and normal heart function, absence of wall motion abnormalities, no sign of valvular heart disease, no signs of myocardial or pericardial inflammation, absence of pulmonary edema or pleural effusions, normal diameters of pulmonary trunk and aorta). Exclusion criteria and contraindications for CMR examinations were: kidney failure, incompatible metal or cochlear implants, known gadolinium intolerance and claustrophobia.

CMR Analysis
LV and RV native T1 and T2 blood pool values were measured in regions of interest (ROI) at the basal or midventricular SA section. All ROIs were drawn carefully within the LV and RV to avoid measuring of partial volume-averaging artefacts and registration errors with gradual T1 changes at myocardial borders, due to including of trabeculae and papillary muscles. The lower limit of the measured blood pool ROI areas was defined as 100 mm 2 and the upper limit as 300 mm 2 to guarantee comparable and valid measurements for all CTEPH patients and controls.
The T1 and T2 measurements were performed by two board-certified and experienced radiologists (12 and 27 years of cardiovascular imaging experience, respectively), and all T1 and T2 maps were of diagnostic quality (no patient had to be excluded). Both investigators were blinded to the patient demographics. At baseline, the first investigator performed all native T1 and T2 blood pool measurements and repeated the measurements after 14 days to assess intraobserver variability. The second investigator performed all native T1 and T2 blood pool measurements to determine interobserver variabilities. Postprocessing was performed by using the cardiovascular imaging software version 42 (Circle Cardiovascular Imaging, Calgary, AB, Canada).

Right Heart Catheterization
Standardized RHC was performed in all patients via the right internal jugular vein (6F sheath and standard Swan-Ganz catheter) within the pre-and postprocedural (6 month after final BPA procedure) diagnostic workup.

Statistical Analysis
The statistical analyses were obtained by using PRISM statistical software (Graphpad Software Version 9.5, San Diego, CA, USA). Patient characteristics were expressed by mean ± SD. All data were checked for normal distribution using the Shapiro-Wilk test. In cases of normally distributed data, Student's t-test was used, and for not-normally distributed data, the Wilcoxon signed-rank test (non-parametric) was applied. The correlation strengths were tested using the Spearman correlation coefficient r and interpreted according to Hinkle et al. [15]. According to Hinkle r > 0.3 was considered as a low correlation, r > 0.5 as a moderate correlation, r > 0.7 as a strong correlation, and r > 0.9 as a very strong correlation. To assess intra-and interobserver agreement The intra-class concordance correlation coefficient (ICC) was used to assess intraobserver and interobserver agreement. An ICC > 0.8 was defined as an excellent agreement. All results were tested at a 5% significance level. An alpha error of being less as 0.05 was accepted as statistically significant. Table 2 presents the RV and LV T1 and T2 blood pool values for CTEPH patients and control subjects. RVT1 and RVT2 were significantly lower than LVT1 and LVT2 in both, CTEPH patients (p = 0.0063 and p < 0.0001) and control subjects (p = 0.0002 and p < 0.0001). Figure 1 shows representative T1 and T2 Maps in a patient with CTEPH and T1 and T2 Maps in a control subject. Compared to control subjects, RVT2 was significantly lower in CTEPH patients (p = 0.0065), whereas LVT2, RVT1 and LVT1 differed not significantly compared to control subjects. In contrast to RVT1/LVT1, RVT2/LVT2 differed also significantly between CTEPH patients and control subjects (p = 0.0006). Table 3 presents the RV and LV T1 and T2 blood pool values of the CTEPH patients compared to the control subjects. Table 4 presents the functional and hemodynamic response to BPA therapy and Table 5 presents RV and LV T1 and T2 blood pool values of CTEPH patients before and after BPA. BPA leaded to significant increases of RVT2 and RVT2/LVT2 (p = 0.0048 and p = 0.0036), whereas LVT2, RVT1, LVT1 and RVT1/LVT1 differed not significantly after BPA. Figure 2 shows RV and LV T2 values in a CTEPH patient before and after BPA.     Table 4 presents the functional and hemodynamic response to BPA therapy an ble 5 presents RV and LV T1 and T2 blood pool values of CTEPH patients before and BPA. BPA leaded to significant increases of RVT2 and RVT2/LVT2 (p = 0.0048 an 0.0036), whereas LVT2, RVT1, LVT1 and RVT1/LVT1 differed not significantly after Figure 2 shows RV and LV T2 values in a CTEPH patient before and after BPA.     Interestingly, RVT2 showed a significant positive moderate correlation to CI before BPA (r = 0.5155, p = 0.007) and a significant positive weak correlation after BPA (r = 0.4769, p = 0.0138). Moreover, RVT2 also showed significant negative weak correlations to PVR before (r = −0.4571, p = 0.0189) and after BPA (r = −0.4396, p = 0.0246), whereas only nonsignificant weak correlations to mPAP before (r = −0.2541, p = 0.2014) and after (r = −0.2585, p = 0.2022) BPA were present. Table 6 presents the correlations of RVT2 and Table 7 RVT2/LVT2 ratio to CI, mPAP and PVR before and after BPA in CTEPH patients.  Interestingly, RVT2 showed a significant positive moderate correlation to CI before BPA (r = 0.5155, p = 0.007) and a significant positive weak correlation after BPA (r = 0.4769, p = 0.0138). Moreover, RVT2 also showed significant negative weak correlations to PVR before (r = −0.4571, p = 0.0189) and after BPA (r = −0.4396, p = 0.0246), whereas only non-significant weak correlations to mPAP before (r = −0.2541, p = 0.2014) and after (r = −0.2585, p = 0.2022) BPA were present. Table 6 presents the correlations of RVT2 and Table 7 RVT2/LVT2 ratio to CI, mPAP and PVR before and after BPA in CTEPH patients.

Discussion
Myocardial T1 and T2 times show excellent reproducibility and are reliably used in daily clinical routine in many cardiac diseases, for example, in myocardial inflammation [1], fibrosis [16][17][18], amyloidosis [19,20], iron overload [21] and Morbus Fabry [22]. In addition to the analysis of myocardial T1 and T2 times, T1 blood pool values before and after contrast material administration have also been well studied as in integral part of ECV quantifications [23]. In contrast, T2 blood pool values have not been studied so far and data is sparse. A first study on T2 blood pool values by Tilman et al. showed that LVT2 and RVT2 blood pool values and their ratio are suitable to diagnose left-to-right atrial shunting [11]. To the best of our knowledge, this is the first study in the field which investigated T1 and T2 blood pool values in patients with CTEPH compared to control subjects and in correlation to pulmonary hemodynamic.
The four most significant findings of our study are: (1) LV and RV native T1 and T2 blood pool values measured on basal SA sections are reproducibly in both control subjects and CTEPH patients. To explain and understand these results, one needs to understand pathophysiological and technical MRI aspects. Relaxation times of blood are a function of hemoglobin quantity and oxygen saturation [24,25] and the magnetic susceptibility of hemoglobin depends on its oxygenation condition-oxyhemoglobin has a high signal on T2, while deoxyhemoglobin has a low signal. This means blood T2 times increases the higher the oxygen saturation and the intracellular hemoglobin content is, respectively; it could be shown that T2 alterations of blood oxygen saturation affects the blood T2 value [12].
In healthy people, the arterial oxygen saturation in the left heart is higher than the venous oxygen saturation in the right heart. This is concordant to our finding that LVT2 blood values are higher than RVT2 blood values in both control subjects and CTEPH patients. Compared to the control subjects, the RVT2 blood values are significantly decreased in CTEPH patients. This can be explained by the lower blood oxygen saturation and increased deoxyhemoglobin due to the increased PVR and right heart failure in patients with PH. Compared to that, the oxygen saturation in the left ventricle seems not affected as LVT2 blood values are similar between CTEPH patients and control subjects, again very well fitting to the pathophysiology of precapillary PH. Furthermore, RVT2 blood values were significantly positive correlated to CI and significantly negative correlated to PVR before and after BPA, whereas a weak not significant negative correlation to mPAP was present. These correlations impressively demonstrate that therapy-associated improvements in mPAP, PVR, and oxygen saturation are accompanied by concomitant changes in T2 values. At least, blood pool T1 mapping seem less affected by changes of oxygen saturation as they do not significantly differ between the CTEPH patients and the control subjects. This is in line with results presented by Lin et al., who showed that blood T1 times were independent of oxygen saturation [26]. The slight non-significant increases in blood pool T1 values in the CTEPH patients compared to the control subjects are most likely due to chance or combinations of influencing variables such as hematocrit and temperature, for example. Influencing variables and combinations of variables were investigated by Liu et al. in human neonatal blood at 3 Tesla [27].
The main limitations of this study are: First, it was a single-centre study with a relatively small number of included patients. Second, our study only included CTEPH patients and, therefore, only limited inferences on blood pool T2 values, oxygen saturation and pulmonary hemodynamic in other PH subgroups appear to be permissible. Third, postcontrast mapping, which might provide additional information, was not a subject of this study. Fourth, hemodynamic of the control subjects is not eligible as RHC is not obtained in healthy people. Fifth, hematocrit values, which also affect the measured native T1 and T2 blood pool times beside oxygenation, were not assessed.

Conclusions
Based on our results, we conclude that right ventricular blood pool T2 mapping might be a novel and additional non-invasive CMR imaging marker and increment for assessment of disease severity, prognosis, follow-up and even therapy monitoring in PH. Data Availability Statement: The datasets used/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.