Partition coefficients for gadolinium chelates in the normal myocardium: Comparison of Gadopentetate dimeglumine and Gadobenate dimeglumine

Abstract

Purpose

To evaluate the influence of contrast agents with different relaxivity on the partition coefficient (λ) and timing of equilibration by using a Modified Look-Locker Inversion Recovery (MOLLI) sequence in cardiac MRI.

Materials and Methods

MOLLI was acquired in 20 healthy subjects (1.5T) at mid-ventricular short axis pre contrast and 5, 10, 20, 25, and 30 min after administration of a bolus of 0.15mmol/kg Gadobenate dimeglumine (Gd-BOPTA) (n=10) or Gadopentetate dimeglumine (Gd-DTPA) (n=10). T1 times were measured in myocardium and blood pool. λ was approximated by ΔR1_myocardium /ΔR1_blood. Values for Gd-BOPTA and Gd-DTPA were compared. Inter-observer agreement was evaluated (intraclass correlation coefficient [ICC]).

Results

T1 times of myocardium and blood pool (p<0.001) and λ (0.42±0.03 and 0.47±0.04, respectively, p<0.001; excluding 5 min for Gd-BOPTA) were significantly lower for Gd-BOPTA than Gd-DTPA. λ_(Gd-DTPA) showed no significant variation between 5 and 30min. λ_(Gd-BOPTA) values were significantly lower at 5min compared to other times (0.38 vs. 0.42; p<0.05). Inter-observer agreement for λ values was excellent with Gd-BOPTA (ICC=0.818) and good for Gd-DTPA (ICC=0.631).

Conclusion

λ_(Gd-BOPTA) values are significantly lower compared to λ_(Gd-DTPA) at the same administered dose. Using Gd-BOPTA, the equilibrium between myocardium and blood pool is not achieved at 5min post contrast.

INTRODUCTION

Cardiomyopathy is frequently associated with diffuse myocardial fibrosis (1) resulting in decreased myocardial compliance that may lead to heart failure.(2) Identification of myocardial fibrosis may lead to a potential therapeutic target.(3) Ideally, myocardial fibrosis would be detected and treated at an early, subclinical stage before symptoms occur when the condition may be reversible.(1)

Currently there is no standardized non-invasive method to quantify diffuse myocardial fibrosis. Myocardial biopsy of the left ventricle (LV) is the standard of reference, but is associated with risk of complications such as hemopericardium and stroke.(4) Furthermore, myocardial biopsy might give false-negative results.(5)

Recently, myocardial longitudinal relaxation times (T1) measured by cardiac magnetic resonance imaging (CMRI) have been demonstrated to correlate with histologically proven fibrosis.(6) Modified Look-Locker Inversion Recovery (MOLLI) is a new technique to measure myocardial T1 values and therefore quantify diffuse myocardial fibrosis.(7) Relaxation rates (R1=1/T1) are frequently used to report tissue relaxation. The partition coefficient (Lambda, λ) assumes a two compartment model: the gadolinium based contrast agent distributes in the intravascular and the interstitial compartment and reaches a steady state that is supposed to be maintained over a certain period after contrast administration. The partition coefficient can be approximated from the change in relaxation rate of myocardium and blood reflected as (1/T1_{Myocardium pre contrast} – 1/T1_{Myocardium post contrast}) / (1/T1_{Blood pool pre contrast} – 1/T1_{Blood pool post contrast}). (8–10)

The purpose of this study was to evaluate the influence of contrast agents with different relaxivity on the partition coefficient and timing of equilibration by using MOLLI in CMRI.

MATERIALS AND METHODS

StudyPopulation

All study participants signed informed consent in this institution review board approved study. Study participants were imaged on 1.5T scanners (Avanto/Espree, Siemens Medical Solutions, Erlangen, Germany). The MOLLI sequence was acquired at a mid-ventricular short axis pre contrast and 5, 10, 20, 25, and 30 min after intravenous administration of a bolus of either 0.15mmol/kg Gadobenate dimeglumine (Gd-BOPTA) (MultiHance, Bracco Imaging, Milan, Italy) (group1: n=10; 5 women; mean age±SD 29.8±8.2; age range 21–44) or 0.15mmol/kg Gadopentetate dimeglumine (Gd-DTPA) (Magnevist, Schering, Berlin, Germany) (group 2: n=10; 6 women; mean age±SD 29.3±7.3; age range 19–39) injected at 2ml/s using a power injector. A dose of 0.15mmol/kg was studied previously (7) and reflects the current practices at our institution. The MOLLI sequence acquired 11 images at different inversion times using the following scan parameters: TE/TR 1.1/2.5–2.7ms; flip angle 35°; bandwidth 977–1002Hz/pixel; slice thickness 8mm; field of view 256 × 144–340, matrix 192–270 × 138–192; TI initial 100–150ms; TI increment 80ms.

Image Analysis

T1 maps were generated using MRmap (11) and transferred to QMass V.7.2 (Medis Medical Imaging Systems, Netherlands). LV endocardial and epicardial contours were drawn manually while segments were defined automatically (Fig. 1). T1 times at the mid short axis slice of the LV (American Heart Association segments 7–12) (12) were determined for each segment and also calculated for the entire slice. T1 time of the blood pool was measured by manually drawing a region of interest in the blood pool of the LV cavity. All measurements were performed by two independent readers. The partition coefficient was approximated by ΔR1_myocardium/ΔR1_blood, where R1=1/T1.

Figure 1.

T1 map pre- (a) and post-contrast (b) with measurements.

Statistical Analysis

Statistical analysis was performed using SPSS statistical software (version 17). The continuous variables, T1 time and λ, are expressed as mean ± standard deviation and were compared using Student’s t-test. A p-value <0.0083 was considered statistically significant after Bonferroni correction for multiple tests.

Inter-observer agreement was evaluated for T1 of the whole myocardium, the blood pool, and λ for both groups separately according to the Bland-Altman method and by calculating the two-way random model of the intraclass correlation coefficient (ICC<0.40 = poor; ICC≥0.40 to 0.75 = fair to good; ICC>0.75 = excellent reproducibility).

RESULTS

Of 720 evaluated segments (6 segments per slice in 20 subjects at 6 time points), 706 were available for analysis. Of the 14 segments excluded (4 in group1, 10 in group 2), segment 10 (inferior) was affected most often (6/14) due to artifacts from adjacent stomach / bowel air below the diaphragm.

Before contrast administration, there was no statistical significant difference between the mean T1 values of the myocardium and the blood pool (p>0.05) (Table 1). As expected, post contrast injection, T1 values for Gd-BOPTA were significantly lower than Gd-DTPA in the myocardium and blood pool (Fig. 2a and 2b, respectively) at all time points (Table 1). λ_(Gd-BOPTA) at 5 min post contrast injection was significantly lower compared to the mean λ_(Gd-BOPTA)of all other times (0.38 vs.0.42; p<0.05). The mean λ_(Gd-BOPTA) averaged over all time points (excluding the 5 min point) was 0.42 ± 0.03 compared to the mean λ_(Gd-DTPA) of 0.47 ± 0.04 (p<0.001).

Table 1.

Comparison of T1 times and the partition coefficient of both contrast agents

Time	Measurement	Contrast agent		p-value
		Gd-BOPTA	Gd-DTPA
Pre- contrast	T1 myocardium (mean ± SD in ms)	1008 ± 31	1011 ± 41	0.83
	T1 blood pool (mean ± SD in ms)	1527 ± 62	1596 ± 91	0.06
	T1 myocardium (mean ± SD in ms)	305 ± 15	388 ± 26	<0.001*
5 min p.i.	T1 blood pool (mean ± SD in ms)	151 ± 17	247 ± 19	<0.001*
	Partition coefficient	0.38 ± 0.04	0.47 ± 0.06	0.002*
	T1 myocardium (mean ± SD in ms)	356 ± 20	449 ± 37	<0.001*
10 min p.i.	T1 blood pool (mean ± SD in ms)	197 ± 26	302 ± 27	<0.001*
	Partition coefficient	0.41 ± 0.04	0.46 ± 0.03	0.008*
	T1 myocardium (mean ± SD in ms)	414 ± 27	520 ± 29	<0.001*
20 min p.i.	T1 blood pool (mean ± SD in ms)	246 ± 29	378 ± 26	<0.001*
	Partition coefficient	0.42 ± 0.03	0.46 ± 0.03	0.002*
	T1 myocardium (mean ± SD in ms)	436 ± 27	541 ± 27	<0.001*
25 min p.i.	T1 blood pool (mean ± SD in ms)	265 ± 30	406 ± 25	<0.001*
	Partition coefficient	0.42 ± 0.03	0.47 ± 0.04	0.003*
	T1 myocardium (mean ± SD in ms)	452 ± 21	559 ± 27	<0.001*
30 min p.i.	T1 blood pool (mean ± SD in ms)	283 ± 31	431 ± 28	<0.001*
	Partition coefficient	0.42 ± 0.03	0.47 ± 0.03	0.002*

^*p-value statistically significant; p.i.=post injection; SD=standard deviation

Figure 2.

T1 times post contrast for Gd-BOPTA and Gd-DTPA of myocardium (a), blood pool (b), and λ (c).

Inter-observer Agreement

The inter-observer agreement for T1 values of myocardium and blood pool was excellent for both contrast agents (Table 2). For λ, exams acquired with Gd-BOPTA showed excellent inter-observer agreement (ICC = 0.818) but fair to good for Gd-DTPA (ICC = 0.631) (Table2). Limits of agreement for λ were similar for both Gd-DTPA (Fig. 3a) and Gd-BOPTA (Fig. 3b).

Table 2.

Inter-observer agreement by intraclass correlation coefficient (ICC)

	ICC (Gd-BOPTA)	ICC (Gd-DTPA)
Myocardium	0.996	0.998
Blood pool	1.0	1.0
Partition coefficient	0.818	0.631

Figure 3.

Bland-Altman analysis for the inter-observer agreement of λ_(Gd-BOPTA) (a) and λ_(Gd-DTPA) (b) that were measured by the two readers.

DISCUSSION

The purpose of the study was to evaluate the influence of contrast agents with different relaxivity on the partition coefficient. At equivalent equimolar administration, Gd-BOPTA is known to have higher relaxivity compared to Gd-DTPA due to weak protein binding effects. (13) In this study, we demonstrated that Gd-BOPTA also has a significantly higher partition coefficient compared to Gd-DTPA in the normal myocardium. Further, equilibrium distribution between myocardium and blood pool may take longer to achieve with Gd-BOPTA compared to Gd-DTPA.

As expected, in the current study T1 times of myocardium and blood pool were significantly lower with Gd-BOPTA compared to Gd-DTPA. Compared with a study published by Messroghli et al. normal myocardial T1 values for Gd-DTPA derived from our study were about 5% lower at 10min post-contrast and about 2% higher at 20min.(7) In both studies images were acquired using the MOLLI sequence at a field strength of 1.5T and the same contrast agent was used at the same dose (Magnevist, 0.15mmol/kg). Differences in T1 values may be due to administration of two separate bolus injections in (7) while a single gadolinium bolus was applied in the current study.

Schlosser et al. also demonstrated lower myocardial T1 values for Gd-BOPTA compared to Gd-DTPA in normal and infarcted myocardium. T1 values for both contrast agents reported in their study are lower when compared to T1 values of the current study due to a higher contrast dose of 0.2mmol/kg.(14)

The contrast agent distribution (λ) in contrast to T1 values is considered to be constant over time, and independent of most MR properties as well as of gadolinium dose.(9) In several recent publications λ values were used as an index for measurement of diffuse myocardial fibrosis.(9,15) Sharma et al reported λ values for a single dose of Gd-DTPA-BMA (gadodiamide) of 0.49±0.05 and for a double dose of 0.44±0.06. In the current study λ_(Gd-DTPA) was 0.47±0.04 averaged over all time points. In an animal study using contrast doses of 0.05, 0.1, and 0.2 mmol/kg, the value reported for λ_(Gd-DTPA) was approximately 0.4 regardless of the dose while it varied with dose for Gd-BOPTA (0.43±0.03 at 0.05mmol/kg; 0.44±0.02 at 0.1mmol/kg and 0.32±0.02 at 0.2mmol/kg). (10) A comparison of the mean values of λ_(Gd-DTPA) and λ_(Gd-BOPTA) was not calculated. Our results show that λ_(Gd-BOPTA) is also significantly lower, compared to λ_(Gd-DTPA). This is likely due to differences in compartmental distribution of Gd-BOPTA versus Gd-DTPA, probably related to the protein binding capacity of Gd-BOPTA. The exact mechanism of this effect, however, is unknown.

In a recent study, the myocardial extravascular extracellular volume fraction was measured reliably 12–50 min after injection of a single bolus of 0.2mmol/kg gadoteridol since an equilibrium establishes between plasma and interstitial gadolinium contrast.(16) The current study revealed that equilibrium between myocardium and blood pool using Gd-BOPTA is not achieved at 5min post contrast administration. In contrast for an animal model (rat), Wendland et al. did not find a significant difference of ΔR1_myocardium / ΔR1_blood between 4 to 29 minutes after administration of Gd-BOPTA at three different doses. (10) This discrepancy may be due to interspecies differences. In addition, the study by Wendland was a smaller sample size (six rats per contrast dose) and a different type of T1 sequence (blipped echo planar imaging) scanned on a 2T system.

A limitation of the current study is lack of partition coefficient correction for hematocrit. Since all volunteers were healthy individuals the hematocrit was considered to be within the limits of normal for both groups, so between group differences were likely small. The assumption of normality allowed group comparison for the purposes of this study. Based on other studies of our institution, mean hematocrit of healthy volunteers in the same age range was 0.40. This hematocrit value yields an extracellular volume fraction (ECV) of 0.249 for Gd-DTPA and a value of 0.252 for Gd-BOPTA.

In the current study a bolus infusion technique was used for contrast administration instead of the slow infusion technique proposed by Flett et al. (17) Thornhill et al. evaluated the partition coefficient after contrast infusion and the bolus technique in the same group of patients. Lambda was independent of the contrast administration scheme.(18) Schelbert et al. also compared the slow infusion versus bolus technique and did not find a statistically significant difference over a period of 12–50 minutes including older subjects with significant comorbidity. (16).

In conclusion, partition coefficients are significantly lower for Gd-BOPTA compared to Gd-DTPA at the same administered dose. Using Gd-BOPTA, the equilibrium between myocardium and blood pool is not achieved at 5min post contrast. We conclude that these contrast agents are not interchangeable regarding their potential utility in identifying myocardial fibrosis.