Features of cardiovascular magnetic resonance native T1 mapping in maintenance hemodialysis patients and their related factors

Abstract

Purpose

Increased myocardial T1 values on cardiovascular MRI (CMRI) have been shown to be a surrogate marker for myocardial fibrosis. The use of CMRI in patients on hemodialysis (HD) remains limited. This research aimed to explore the characteristics of native T1 values in HD patients and identify factors related to T1 values.

Methods

A total of thirty-two patients on HD and fourteen healthy controls were included in this study. All participants underwent CMRI. Using modified Look-Locker inversion recovery (MOLLI) sequence, native T1 mapping was achieved. Native CMRI T1 values were compared between the two groups. In order to analyze the relationship between T1 values and clinical parameters, correlation analysis was performed in patients on HD.

Results

Patients on HD exhibited elevated global native T1 values compared to control subjects. In the HD group, the global native T1 value correlated positively with intact parathyroid hormone (iPTH) (r = 0.418, p = 0.017) and negatively with triglycerides (r= −0.366, p = 0.039). Moreover, the global native T1 value exhibited a positive correlation with the left ventricular end-diastolic volume indexed to body surface area (BSA; r = 0.528, p = 0.014), left ventricular end-systolic volume indexed to BSA (r = 0.506, p = 0.019), and left ventricular mass indexed to BSA (r = 0.600, p = 0.005). A negative correlation was observed between the global native T1 value and ejection fraction (r = 0.-0.551, p = 0.010).

Conclusion

The global native T1 value was prolonged in HD patients compared with controls. In the HD group, the global T1 value correlated strongly with iPTH, triglycerides, and cardiac structural and functional parameters.

Keywords:

• Hemodialysis
• cardiovascular magnetic resonance
• native T1 mapping
• ambulatory blood pressure monitoring
• myocardial fibrosis

Introduction

Populations diagnosed with end-stage renal disease (ESRD) are at a higher risk of cardiovascular (CV) mortality, primarily caused by sudden cardiac death and recurrent heart failure due to uremic cardiomyopathy. Various cardiomyopathies can result in myocardial dysfunction due to myocardial fibrosis, eventually leading to myocardial remodeling and poor outcomes. The identification and reversal of uremic cardiomyopathy are crucial for reducing the mortality of patients on hemodialysis (HD), and cardiovascular MRI (CMRI) is an effective method for diagnosing heart conditions. In the past, contrast-enhanced CMRI was used to demonstrate myocardial fibrosis in individuals with ESRD. Due to the potential risk of gadolinium use in patients with kidney disease, the use of contrast-enhanced CMRI in individuals with ESRD has been limited. Thus, the imaging and quantification of myocardial fibrosis in these patients is challenging [1,2]. A technique of measuring myocardial fibrosis without the need for contrast agent would be highly desirable.

The association between non-contrast-enhanced native T1 values and histologically assessed myocardial fibrosis burden in patients with aortic stenosis (AS) was found to be significant [3]. A number of studies have discovered that the modified Look-Locker inversion recovery (MOLLI) sequence exhibits high reproducibility and test-retest stability [4–6]. There is also much evidence for the accuracy of MOLLI-derived T1 mapping indices in relation to the reference standard represented by LV myocardial histology [7,8].

A variety of cardiac conditions resulting in diffuse myocardial fibrosis have been investigated using native T1 mapping, and elevated T1 values have been revealed in patients with heart disease compared to control subjects in various studies [9–11]. Within our investigation, MOLLI T1 mapping was employed to evaluate native T1 values in patients on HD, aiming to identify the factors that influence the global T1 value.

Methods

Study subjects

The healthy control group did not have any cardiovascular disease or cardiovascular risk factors, and were not regularly taking medications. Every participant provided informed consent, and the study protocol received approval from the center’s ethics committee.

Blood samples were collected from fasting individuals on a midweek interdialytic day. Standard methods were used to measure several indices including hemoglobin, serum lipids, albumin, calcium, phosphorus, high sensitivity C-reactive protein (hs-CRP), intact parathyroid hormone (iPTH), homocysteine, and N-terminal fragment pro-brain natriuretic peptide (NT-proBNP).

In patients on HD, ambulatory blood pressure (ABP) monitoring was conducted for 44 h after the midweek hemodialysis session. ABP readings were taken every 30 min from 6 am to 6 pm and every 60 min from 6 pm to 6 am, using Spacelabs 90207 devices (Spacelabs Healthcare, Issaquah, WA, USA) in the nonaccess arm.The recordings started right after the completion of hemodialysis and concluded just before the subsequent dialysis session. From these recordings, the 44-h average systolic blood pressure (SBP), 44-h average diastolic blood pressure (DBP), 44-h systolic blood pressure variability (SBPV), and 44-h diastolic blood pressure variability (DBPV) were calculated.

LV diastolic function was assessed by the ratio of early to late diastolic transmitral flow velocity (E/A), the ratio of early transmitral peak velocity to early diastolic peak annular velocity (E/e′), and left atrial volume index. The echocardiographic examination was performed the day after dialysis. A CMRI would be scheduled on the same day.

CMRI protocols

CMRI was performed on a 3 T CMRI system (Trio; Siemens, Erlangen, Germany) with an 8-channel cardiovascular phased-array torso coil. All participants underwent CMRI using standard protocols [12]. All HD patients received hemodialysis regularly (thrice weekly). On cine CMRI, multiple short-axis slices were acquired from the mitral orifice to the LV apex. The scanning parameters were as follows: repetition time (TR) 48 ms, echo time (TE) 1.5 ms, flip angle (FA) 40°, field of view (FOV) 320 × 320, matrix 256 × 173, slice thickness 6 mm, and gap 0 mm. T1 mapping images were acquired for the same short-axis slices. Myocardial T1 mapping was performed with an electrocardiography (ECG)-triggered MOLLI sequence with the following parameters: TR 314.85 ms, TE 1.12 ms, FA 35°, FOV 300 × 327, matrix 256 × 192, slice thickness 8 mm, gap 0 mm.

Image analysis

The anonymized images were analyzed on a Siemens Healthcare (syngoMR) workstation in a randomly determined order. For LV function and mass analysis, endocardial and epicardial contours were traced on each pre and post contrast image pair and the area enclosed by each contour was computed. Using the software, parameters such as left ventricular end-diastolic volume indexed to body surface area (LVEDVI), left ventricular end-systolic volume indexed to body surface area (LVESVI), left ventricular mass indexed to body surface area (LVMI), ejection fraction (EF), and cardiac index (CI), were calculated.

The myocardium was divided into 16 segments according to the American Heart Association (AHA) standard [13]. Representative basal, mid-cavity, and apical slices from the short-axis views were selected for analysis. Regions of interest (ROIs) were delineated by user-defined semiautomated border delineation (Argus, Siemens Healthcare). The ROIs were standardized to be of similar size and shape. The global T1 value was obtained by averaging the T1 values of every myocardial segment [14].

Statistical analysis

Statistical analyses were performed using SPSS version 20.0 software (IBM Inc, Chicago, IL). Continuous variables are presented as mean ± standard deviation (SD). Changes in variables between the two groups were compared using unpaired t tests. To compare multiple groups, analysis of variance (ANOVA) was used. Levene’s test was used to test the homogeneity of variance. Chi-square tests were used to compare categorical variables. The Pearson product-moment correlation coefficient was used to measure linear correlations between two variables. In all analyses, p < 0.05 was taken to indicate significance.

Results

Baseline data

There were 32 patients on HD and 14 healthy controls included in the study. The detailed information about the participants can be found in Table 1. The gender composition was similar between the two groups (p = 0.766), but the patients on HD were significantly older than the healthy controls (57.9 ± 11.3 years vs. 43.1 ± 10.9 years, p < 0.001).

Table 1. Baseline Demographic characteristics and clinical data for the patients on HD and healthy controls.

Variable	HD patients(n = 32)	Healthy controls(n = 14)	p Value
Gender (Female,%)	68.6	64.2	0.766
Age(years)	57.9 ± 11.3	43.1 ± 10.9	<0.001
Dialysis parameters
Duration of dialysis (months)	83.3 ± 57.7
Dry weight(kg)	61.7 ± 11.9
Ultrafiltration volume	2.4 ± 0.9
Laboratory examination
Hemoglobin(g/L)	120.4 ± 13.8	135.4 ± 10.2	<0.001
Hematocrit(%)	37.1 ± 4.2	45.1 ± 3.8	<0.001
Albumin(g/L)	36.3 ± 2.9	44.6 ± 5.4	<0.001
Serum creatinine(umol/L)	946.5 ± 188.3	79.3 ± 12.3	<0.001
Blood urea nitrogen (mmol/L)	27.3 ± 6.5	5.1 ± 1.6	<0.001
Uric acid (umol/L)	376.0 ± 70.3	305.2 ± 30.4	<0.001
Cholesterol(mmol/L)	4.4 ± 0.9
High density lipoprotein(mmol/L)	1.1 ± 0.3
Low density lipoprotein(mmol/L)	2.4 ± 0.8
Triglycerides(mmol/L)	1.8 ± 1.3
Lipoprotein a(mg/dL)	21.4 ± 20.8
Intact parathyroid hormone(pg/mL)	481.6 ± 313.6
Homocysteine(umol/L)	26.0 ± 9.2
High sensitivity C-reactive protein(mg/L)	3.5 ± 3.6
Calcium(mmol/L)	2.2 ± 0.2
Phosphorus(mmol/L)	2.1 ± 0.5
Calcium-phosphorus product	4.8 ± 1.4
Sodium(mmol/L)	137.8 ± 2.6
Potassium(mmol/L)	5.4 ± 0.8
N-terminal fragment pro-brain natriuretic peptide (pg/ml)	10657 ± 11859.7

Data are presented as percentage [number of participants] or mean ± SD, as appropriate.

Native myocardial T1 determined by the MOLLI sequence

In comparison with the controls, the patients on HD had higher global native T1 value (1208.9 ± 90.9 ms vs. 1134. 5 ± 28.1 ms, p = 0.045), with the difference between the two groups remaining after adjustment for age (p = 0.045; Figure 1).

Figure 1. Comparison of the global native T1 value between the HD group (n = 32) and the healthy control group (n = 14).

The native myocardial T1 values of the LV apical, mid-cavity, and basal segments in the HD group were 1213.1 ± 10.3 ms, 1200.6 ± 89.8 ms, and 1213.1 ± 89.1 ms, respectively. These findings suggest that native myocardial T1 values are similar across different segments of the left ventricle (p = 0.831; Figure 2).

Figure 2. Comparison of native T1 values of left ventricular segments in the HD group (n = 32).

After adjusting for age, the native T1 values of the mid-cavity and basal segments were significantly higher in the HD group than in the healthy control group (mid-cavity segments, 1200.6 ± 89.8 ms vs.1121.4 ± 30.4 ms, p = 0.021; basal segments, 1213.1 ± 89.1 ms vs. 1119.1 ± 45.4 ms, p = 0.015), whereas those of the apical segments were similar in the two groups (1213.1 ± 10.3 ms for the HD group vs. 1163.0 ± 39.4 ms for the healthy control group, p = 0.249).

Associations between the global native T1 value and ABP and laboratory and cardiac parameters in patients on HD

The correlation coefficients of the global native T1 value for these parameters are presented in Table 2. The global native T1 value correlated positively with iPTH (r = 0.418, p = 0.017; Figure 3A) and negatively with triglycerides (r = −0.366, p = 0.039). However, we did not find significant correlations between the global native T1 value and other laboratory parameters.

Figure 3. Graphs illustrate the correlation between the global native T1 value and laboratory, cardiac parameters in patients on HD. The correlation was analyzed between the global native T1 value and A, intact parathyroid hormone, B, left ventricular end-diastolic volume indexed to body surface area, C, left ventricular mass indexed to body surface area, and D, left ventricular ejection fraction.

Table 2. Correlation analysis of the global native T1 value.

Variable	Pearson coefficient	p Value
Duration of dialysis	−0.125	0.495
Dry weight	−0.020	0.961
Ultrafiltration volume	0.099	0.588
Hemoglobin	−0.091	0.619
Hematocrit	−0.089	0.628
Albumin	−0.281	0.120
Serum creatinine	−0.156	0.395
Blood urea nitrogen	−0.251	0.165
Uric acid	−0.224	0.217
Cholesterol	0.130	0.480
High sensitivity C-reactive protein	0.241	0.184
Low density lipoprotein	0.215	0.238
Triglycerides	−0.366	0.039
Lipoprotein a	0.067	0.717
Intact parathyroid hormone	0.418	0.017
Homocysteine	0.074	0.689
High sensitivity C-reactive protein	−0.043	0.816
Calcium	0.211	0.247
Phosphorus	0.109	0.553
Calcium-phosphorus product	0.161	0.378
Sodium	0.354	0.057
Potassium	−0.015	0.936
N-terminal fragment pro-brain natriuretic peptide	0.212	0.243
44-h average SBP	0.204	0.320
44-h average DBP	0.316	0.208
44-h systolic blood pressure variability	0.259	0.402
44-h diastolic blood pressure variability	0.135	0.662
Left ventricular end-diastolic volume indexed to body surface area	0.528	0.014
Left ventricular end-systolic volume indexed to body surface area	0.506	0.019
Left ventricular mass indexed to body surface area	0.600	0.005
Cardiac index	−0.210	0.357
Ejection fraction	−0.551	0.010
Ratio of early to late diastolic transmitral flow velocity (E/A)	0.008	0.965
Ratio of early transmitral peak velocity to early diastolic peak annular velocity (E/e′)	0.253	0.162
Left atrial volume index	0.279	0.123

A total of 28 hemodialysis patients underwent valid 44-h ABP monitoring and provided BP values for use in the analysis. No significant correlation was found between the global native T1 value and the 44-h average SBP (r = 0.204, p = 0.320), 44-h average DBP (r = 0.316, p = 0.208), 44-h SBPV (r = 0.259, p = 0.402), and 44-h DBPV (r = 0.135, p = 0.662).

CMRI can also provide cardiac structural and functional parameters, including LVEDVI, LVESVI, CI, LVMI, EF. In the HD group, there were positive correlations between the global native T1 value and LVEDVI (r = 0.528, p = 0.014), LVESVI (r = 0.506, p = 0.019), and LVMI (r = 0.600, p = 0.005), and a negative correlation between the global native T1 value and EF (r = −0.551, p = 0.010) (Figure 3(B–D)). However, no correlation was found between the global native T1 value and CI (r = −0.210, p = 0.357).

Discussion

Cardiovascular disease accounts for the majority of fatalities in patients with ESRD. Additionally, myocardial fibrosis, commonly observed in various cardiac diseases, could potentially serve as a reliable indicator of cardiac death [15–17]. Strictly speaking, there are two categories of cardiac fibrosis: replacement fibrosis and interstitial fibrosis.The underlying factors, characteristics, and medical relevance of the two types show variations [18]. Replacement fibrosis occurs in specific areas of the heart as a result of damage to the myocardial tissue. Its main function is to preserve the integrity and function of the heart after injury, and it is frequently observed after myocardial infarction. The sudden loss of numerous heart muscle cells can trigger an inflammatory response, leading to the activation of reparative myofibroblasts that eventually generate fibrous tissue. In various cardiac diseases, fibrosis mainly affects the interstitial tissue and develops gradually without causing substantial loss of cardiomyocytes. Myocardial interstitial fibrosis can also be promoted by systemic hypertension, aging, obesity, and diabetes [19]. Replacement and interstitial fibrosis often coexist and are a constant feature of pathologic cardiac remodeling [20]. Unlike the process of replacing damaged tissue with fibrotic cells, interstitial fibrosis has the potential to be reversed and is therefore a promising target for therapeutic interventions [21,22]. Patients with myocardial infarction and known cardiac disease were excluded from our study; thus, any changes to the myocardial interstitium may largely depend on the degree of uremia-related fibrosis. Endomyocardial biopsy is the accepted standard for assessing myocardial fibrosis.However, due to its invasive nature, it is not ideal for regular and frequent monitoring of patients. Cardiac magnetic resonance-derived T1 mapping has been validated as a reliable noninvasive method for the precise identification and measurement of diffuse interstitial myocardial fibrosis [23]. However, the assessment of native T1 values in patients on HD and comparisons with such values in the healthy population are insufficient.

In our study using the MOLLI sequence on a 3 T scanner, we discovered that the HD group exhibited higher values for both the global native T1 value and the native T1 values of the mid-cavity and basal segments.This finding suggests that patients on HD may experience more severe myocardial fibrosis compared to individuals who are healthy. Patients undergoing hemodialysis had comparable average native T1 values across the basal, mid-cavity, and apical segments, indicating the likelihood of extensive myocardial fibrosis. The results were consistent with Eliane et al.’s findings, which indicated that septal and midseptal T1 values T1 values were considerably higher in the HD group than the control group [1]. The possibility of using native T1 mapping as a novel approach to measure cardiac tissue abnormalities in patients on HD was proposed. The results of our study are consistent with the aforementioned perspectives.

Myocardial fibrosis is common in patients on HD, with many factors being involved in the process, including volume and pressure overload, malnutrition, anemia, uremic toxins, high catecholamine levels, hyperparathyroidism [15], hyperhomocysteinemia [24], and aging. However, our study did not find significant correlation between the global native T1 value and phosphorus, calcium-phosphorus product, hs-CRP, and homocysteine. It is possible that this lack of correlation is due to the limited number of participants and significant variations in these parameters affected by dialysis and diet.

Myocardial fibrosis is significantly affected by the presence of hypertension. Prolonged pressure overload activates the expression of procollagen genes and triggers the synthesis of collagen proteins, which ultimately results in the accumulation of excessive collagen and the development of fibrosis [25]. Nevertheless, we discovered that there was no correlation between the global native T1 value and the average or variation of the 44-h ABP. This result may be due to characteristics of the blood pressure in patients on HD: a volume-related progressive increase during the interdialytic interval and a rapid decrease during hemodialysis. The cross-sectional ABP may not be able to fully reflect the long-term control of BP, and therefore the influence of BP on myocardial fibrosis needs further investigation.

The development of myocardial fibrosis in patients on HD is influenced by numerous significant factors. PTH has been acknowledged to play a role in the development of interstitial fibrosis in myocardial tissue [26]. Moreover, in adult rats, an increased infusion of PTH was associated with the development of myocardial hypertrophy [27]. Correlations have been discovered between an increase in PTH and left ventricular myocardial mass assesed by height (LVMH) in both healthy individuals and those already diagnosed with cardiovascular disease [28]. In our study, the global native T1 value correlated positively with iPTH (r = 0.418, p = 0.017), suggesting a possible association between hyperparathyroidism and myocardial fibrosis. Myocardial fibrosis plays a critical role in LV hypertrophy, and interventional research is necessary to determine the impact of secondary hyperparathyroidism on myocardial fibrosis.

We also found that the global native T1 value correlated negatively with triglycerides (r = −0.366, p = 0.039). However, the relationship between abnormal lipid metabolism and cardiac fibrosis is unclear, and further investigations are needed to confirm whether triglycerides has a protective effect on myocardial injury or fibrosis.

Another finding of our study was that the global native T1 value correlated with cardiac systolic and diastolic function in patients on HD. The global native T1 value correlated positively with LVMI, LVEDVI, and LVESVI, and negatively with EF. These findings may indicate associations between myocardial histological changes and cardiac structural and functional changes in patients on HD. LV hypertrophy is a condition characterized by an enlargement of the left ventricle of the heart. It is commonly associated with increased pressure or volume in the heart, as well as a state of increased uremic toxicity. LV hypertrophy is considered to be an independent risk factor for cardiogenic death in patients on HD. In patients with ESRD, myocardial hypertrophy is accompanied by an increase in collagen fibers and other stromal molecules, and continuous accumulation of collagen may lead to ventricular dysfunction. LVMI can be used to evaluate LV hypertrophy, and a positive correlation was observed between the global native T1 value and LVMI in patients on HD (r = 0.600, p = 0.005), which may indicate that the more severe the LV hypertrophy is, the more severe the myocardial fibrosis is. Eliane et al. [1] found that T1 values were notably prolonged among individuals undergoing hemodialysis. Moreover, these prolonged T1 values demonstrated a positive correlation with LVMI.

LV enlargement is a reliable indicator of cardiovascular outcomes, and it is widely acknowledged that this enlargement serves as a compensatory response to LV systolic dysfunction [29,30]. Diastolic dysfunction usually precedes systolic dysfunction in the LV. Our results showed a positive correlation between the global native T1 value and LVEDVI (r = 0.528, p = 0.014), suggesting that the degree of LV enlargement is related to myocardial fibrosis. In our study, only three patients (9.3%) had decreased cardiac systolic function, which is consistent with other reports [31]. In general, systolic dysfunction is typically not the first indication of heart injury and tends to occur as a result of left ventricular hypertrophy and dilation. However, once cardiac systolic dysfunction occurs, the prognosis is poor. The initial LV ejection fraction upon initiating dialysis is an influential prognostic factor for both cardiovascular and all-cause mortality [32]. In our study, the global native T1 value correlated negatively with cardiac systolic function, which suggests that myocardial systolic dysfunction may be accompanied by more severe myocardial fibrosis.

Patients with diabetes were excluded from our study because a long-term hyperglycemic state has significant effects on myocardial cells and interstitium, manifesting as cardiomyocyte hypertrophy, degeneration, focal necrosis, and microcirculatory disturbance. In order to ensure an accurate assessment of myocardial damage, we excluded individuals who had been diagnosed with diabetes from the study, as hyperglycemia could potentially affect the results.

Although our study has demonstrated that the global native T1 value correlated with iPTH, triglycerides, and cardiac structural and functional parameters in HD patients, there are several limitations that should be acknowledged. First, the number of patients enrolled into the study was relatively small, it could not make generalizations to other hemodialysis patients. In addition, the association between laboratory value and the global native T1 value could not indicate a cause-and-effect relationship due to the cross-sectional setting of the study. Therefore, the influence of such parameters on native T1 value should be further investigated by longitudinal, long-term studies in a larger group of patients. Second, we were unable to obtain a cardiac biopsy. Therefore, T1 values were not compared to the results obtained from cardiac biopsy, while cardiac biopsy continues to be regarded as the gold standard for tissue characterization or function assessment.

Conclusions

The mapping of native T1 values via CMRI has unique advantages for assessing cardiac damage in patients on HD, thereby introducing a potential innovative approach to quantifying anomalies in cardiac tissue among these patients. Other factors, such as iPTH and triglycerides, may also be involved in the occurrence and development of myocardial fibrosis. Additional research is necessary to explore the potential correlation between T1 values and prognosis in patients on HD, as well as to determine whether interventions targeting T1 value reduction can lead to enhanced outcomes.

Acknowledgements

We thank Liwen Bianji (Edanz) (www.liwenbianji.cn) for editing the language of a draft of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.