Native T₁ value in the remote myocardium is independently associated with left ventricular dysfunction in patients with prior myocardial infarction

Abstract

Purpose

To compare remote myocardium native T₁ in patients with chronic myocardial infarction (MI) and controls without MI and to elucidate the relationship of infarct size and native T₁ in the remote myocardium for the prediction of left ventricular (LV) systolic dysfunction after MI.

Materials and Methods

A total of 41 chronic MI (18 anterior MI) patients and 15 age-matched volunteers with normal LV systolic function and no history of MI underwent cardiac MR at 1.5 T. Native T₁ map was performed using a slice interleaved T₁ mapping and late gadolinium enhancement (LGE) imaging. Cine MR was acquired to assess LV function and mass.

Results

The remote myocardium native T₁ time was significantly elevated in patients with prior MI, compared to controls, for both anterior MI and non-anterior MI (anterior MI:1099 ± 30, non-anterior MI:1097 ± 39, controls:1068 ± 25 msec, P <0.05). Remote myocardium native T₁ moderately correlated with LV volume, mass index and ejection fraction (r=0.38, 0.50, −0.49, respectively, all p <0.05). LGE infarct size had a moderate correlation with reduced LV ejection fraction (r=−0.33, p<0.05), but there was no significant association between native T₁ and infarct size. Native T₁ time in the remote myocardium was independently associated with reduced LV ejection fraction, after adjusting for age, gender, infarct size and comorbidity (β=−0.34, p=0.03).

Conclusion

In chronic MI, the severity of LV systolic dysfunction after MI is independently associated with native T₁ in the remote myocardium. Diffuse myocardial fibrosis in the remote myocardium may play an important pathophysiological role of post-MI LV dysfunction.

Introduction

Left ventricular (LV) systolic dysfunction after myocardial infarction (MI) is a major cause of symptomatic heart failure and thought to be an important aspect of mortality^1,2. The prophylactic use of primary prevention implantable cardioverter-defibrillator for reducing mortality due to ventricular arrhythmias is widely accepted in patients with heart failure and a low left ventricular ejection fraction (LVEF)^3,4. However, the importance of left ventricular dysfunction as a pathogenic mechanism after MI is incompletely understood.

Following MI, cardiac magnetic resonance (MR) studies have focused on the assessment of the infarcted tissue^5–8. Native T₁ mapping is a noninvasive cardiac MR biomarker for the assessment of diffuse myocardial fibrosis, and area where late gadolinium enhancement (LGE) cardiac MR has limited accuracy^9,10 and may detect reactive fibrosis in the non-infarcted myocardium. In human as well as animal models of post-MI remodeling, fibroblast stimulation has been shown to cause extracellular volume expansion of both the infarcted and the non-infarcted myocardium^11,12 and studies have reported on native T₁ during the acute phase of MI^13,14. However, heart failure is more common in chronic MI¹⁵ and the association between native T₁ and LV function/geometry in chronic MI remains to be fully clarified. We hypothesized that diffuse myocardial fibrosis of the remote non-infarcted area would be associated with progressive deterioration of LV systolic function.

Methods

We prospectively recruited 41 consecutive patients (30 male, mean age; 63 yrs) with prior MI (18 anterior MI) and 15 age-matched control subjects with normal LV systolic function, no symptoms of heart failure or MI and no other cardiovascular diseases. The severity of heart failure was assessed based on the New York Heart Association (NYHA) Class guidelines¹⁶. All participants were in sinus rhythm at the time of scan. The study protocol was approved by our institutional review board. Written informed consent was obtained from all subjects. Imaging was performed on a 1.5T MRI scanner (Achieva, Philips Medical Systems, Netherlands) equipped with a 32-element cardiac-surface coil. The protocol included cine cardiac MR and native T₁ map, followed by LGE. To assess LV/right ventricular myocardial function, geometry and mass, 10 to 12 short-axis stack images and 4-chamber cines were acquired using a breath-hold, ECG-gated steady state free precession sequence (slice thickness, 8-mm; gap, 2-mm, in-plane spatial resolution 2×2mm, 30 msec temporal resolution). Native T₁ map was acquired using a ECG-triggered, free-breathing slice-interleaved T₁ (STONE) sequence which enables acquisition of 5 slices in the short-axis plane within 90 sec (TR/TE=2.8/1.4 msec, flip angle=70°, FOV=360×351 mm, voxel size=2.1×2.1 mm, slice thickness=8 mm, TFE factor=86, SENSE factor=2)¹⁷. Then, 10 to 20 min after injected 0.1–0.2 mmol/kg of Gd-DTPA (Magnevist; Bayer Schering, Berlin, Germany) or Gd-BOPTA (MultiHance; Bracco Imaging SpA, Milan, Italy), short- and long-axis inversion recovery LGE images were acquired with a 3-dimensional phase sensitive inversion recovery sequence (PSIR) (5-mm slice thickness, TR/TE=5.3/2.1 msec, flip angle=70°, FOV=320×320×125 mm, acquisition matrix=224×224×23 and spatial resolution=1.4×1.4×1.5 mm). Cardiac MR images were analyzed using a commercial workstation (Extend MR WorkSpace, version 2.3.6.3, Philips Healthcare). At end-diastole, epi- and endocardial LV borders were manually traced in contiguous short-axis cine images covering the apex to mitral valve plane to calculate LV mass and end-diastolic volume (EDV) and endo of end-systole to end-systolic volume (ESV), stroke volume, and ejection fraction (EF). LV mass was calculated as the sum of the myocardial volume multiplied by the specific gravity (1.05g/mL) of myocardial tissue¹⁸. Sphericity index was calculated as the ratio of the LV diastolic volume to the volume of a sphere with the diameter of the long axis of the LV in diastole obtained from a 4-chamber cine image (LV volume/[LV long axis length³×π/6]), by using a commercial workstation (OsiriX environment, Pixmeo, Geneva, Switzerland)^19,20. Short-axis slices of native T₁ mapping images were analyzed using custom software (MediaCare, Boston, MA). T₁ map of each scan was estimated by voxel-wise curve fitting using a 2-paramter fit model. Motion correction was performed using the adaptive registration of varying contrast-weighted images for improved tissue characterization (ARCTIC)²¹. In this study, the remote myocardium measurements relied on visually identifying normal area on the corresponding LGE image. For each short-axis cross section, after the endocardial and epicardial borders were traced, a region of interest (ROI) was placed on myocardium without enhancement on LGE image and standardized to be of similar size and shape with relatively large size of 150 pixels or greater, by avoiding contamination with signal from the blood pool and artifact due to a misregistration error. The native T₁ value of the non-infarct remote myocardium in patients with prior MI were calculated as an average ROI value on the three mid-ventricular slices. Native myocardial T₁ in control subjects were measured over the three mid-ventricular slices by manually drawing epicardial and endocardial contours. To evaluate inter-observer reproducibility, measurements of LV native T₁ from 10 random patients with prior MI were independently assessed by two observers (S.N., with 7 years of experience and J.A., with 2 years of experience). One of the two observers measured LV native T₁ twice on two separate days with a washout period of at least 2 weeks to assess intra-observer reproducibility. In 30 MI patients with hematocrit assessment on the day of scanning, T₁ values of remote myocardium and LV blood pool were similarly determined, before and after contrast injection. Then extracellular volume fraction was calculated according to the formula as:

$$\text{Extracellular volume fraction}=(1-\text{hematocrit})\times \{[(1/{\mathrm{T}}_1\text{myocardium.post})-(1/{\mathrm{T}}_1\text{myocardium.pre})]/[(1/{\mathrm{T}}_1\text{blood.post})-(1/{\mathrm{T}}_1\text{blood.pre})]\}\times 100(\%)$$

On LGE images, the presence or absence of LGE was visually assessed by two experienced cardiologists (S.N., with 7 years of experience and S.R.M., with 3 years of experience). If LGE was present, the quantitative extent of hyperenhancement was manually traced and infarct size was expressed as a percent of total LV mass and as a total volume. Infarct regions with evidence of microvascular obstruction were included within the infarct area.

Statistical analyses were performed using SPSS version 19 software (IBM Inc, Chicago, IL) and MedCalc for Windows (version 14.8.1, MedCalc Software, Ostend, Belgium). Continuous variables are expressed as mean ± standard deviation (SD) or median [quartiles] if not normally distributed, and compared using an unpaired Student`s t-test or Mann-Whitney nonparametric test as appropriate. Significance of difference of native T₁ time among the 3 groups were evaluated by one-way ANOVA with Bonferroni’s post-hoc test. Categorical variables were reported as counts and percentages, and compared using a chi-square test. All tests were 2 sided and p value <0.05 was considered statistically significant. Depending on data distribution, either a Pearson or Spearman correlation coefficient was calculated to investigate possible associations of continuous outcome measures. Skewed distributions were logarithmically transformed before regression analysis. Multivariate stepwise regression analyses with several potentially confounding factors were performed with LVEF as the dependent variable. Intra and inter-observer of T₁ times in the remote myocardium were assessed with Bland-Altman methods and intraclass correlation coefficient.

Results

Baseline clinical characteristics of the controls and chronic MI groups are summarized in Table 1. The mean age of all subjects was 63 years (range 40 to 81). Chronic MI patients were more frequently male, obese and had a higher body mass index, while both groups had similar history of hypertension, diabetes mellitus, dyslipidemia and current smoking. Two-thirds of chronic MI patients reported exertional dyspnea, predominantly New York Heart Association Functional class II. Table 2 summarizes cardiac MR findings of the two groups. Chronic MI patients had significantly bigger LV mass index, higher LV ED and ES volumes, and lower LVEF (all p<0.001). LGE hyperenhancement was observed in all chronic MI patients; transmural LGE in 13 patients, and subendocardial LGE in 28 patients. The mean infarct size was 13.2 ± 7.6 g and 12% of total LV myocardium, indicating relatively small infarctions. The mean remote myocardium native T₁ was significantly higher in patients with prior MI compared with that in controls (p=0.003). Even after subclassified into anterior MI (n=18) and non-anterior MI (n=23), the mean remote myocardium native T₁ was significantly higher in both MI groups than that in control subjects (anterior MI: 1099 ± 30, non-anterior MI: 1097 ± 39, and controls: 1068 ± 25 msec, P<0.05 after Bonferroni correction, respectively). There was no significant difference in the native T₁ value of remote myocardium between anterior MI vs. non-anterior MI groups (Figure 1). Figure 2 shows the representative cases from chronic anterior MI with and without increased remote myocardium native T₁. Compared to the case without increased remote native T₁ (Figure 2a), the case with increased native remote T₁ (Figure 2b) demonstrated severe LV systolic dysfunction with mild hypertrophy in the remote myocardium despite similar transmural infarct and smaller infarct size. Figure 3 shows the relationship between remote myocardium native T₁ and LV volumes, mass, and EF. There was a moderate positive correlation with LV mass index (r=0.50, p<0.001), LV EDV index (r=0.38, p=0.004), LV ESV index (r=0.45, p=0.001), and negative correlation with LVEF (r=−0.49, p<0.001). Remote myocardium native T₁ value also had univariate association with the presence of any MI (β=0.39, p=0.003) as well as LV dysfunction. There were no significant correlations between remote myocardium native T₁ and LV spherical geometry or infarct size (Table 3). Table 4 shows the univariate coefficients between reduced LVEF, patient characteristics and cardiac MR parameters and multiple stepwise regression analysis investigating the relationship between reduced LVEF with clinical characteristics and cardiac MR findings in chronic MI subjects. On multi-variable analysis that included age, gender and variables with p value <0.15 in the univariate analysis, the native T₁ value in the remote myocardium were independent predictors of reduced LVEF (β=−0.34, p=0.03). Although infarct size showed significant univariate association with reduced LVEF, this relationship was no longer significant with multi-variable analysis. In 30 of 41 MI patients with extracellular volume fraction data, remote extracellular volume fraction appeared more likely to be similarly and moderately associated with reduced LVEF (r=−0.32, p=0.085). The intraclass coefficients (ICC) for interobserver and intraobserver measurements of native remote myocardial T₁ were 0.88 (95% confidence interval [CI]: 0.72 to 0.94) and 0.91 (95% CI: 0.82 to 0.96), respectively. Bland-Altman analysis revealed acceptable agreement with a mean difference of 0.8% (95% CI; 0.2% to 1.5%) and −0.6% (95% CI; −1.1% to 0%) for inter- and intra-observer measurements, respectively.

Table 1.

Clinical characteristics of the study population

Characteristics	Control subjects (n=15)	Chronic MI patients (n=41)	p Value
Age (years)	62 ± 8	63 ± 9	0.57
Male (%)	8 (53)	30 (73)	0.16
BMI (kg/m2)	24.0 ± 3.2	27.4 ± 4.9	0.017
Hypertension (%)	3 (20)	15 (37)	0.24
Systolic blood pressure (mmHg)	123 ± 19	120 ± 18	0.51
Diastolic blood pressure (mmHg)	72 ± 13	70 ± 12	0.57
Heart rate (beats/min)	67 ± 10	68 ± 14	0.74
Diabetes mellitus (%)	3 (20)	13 (32)	0.39
Hypercholesteloremia (%)	5 (33)	14 (34)	0.96
Current smoking (%)	1 (7)	8 (20)	0.25
BMI ≥30 (%)	1 (7)	10 (24)	0.14
Medication
Aspirin	0 (0)	41 (100)	<0.001
ACEi/ARB	0 (0)	35 (85)	<0.001
Beta-blocker	3 (20)	38 (93)	<0.001
Aldosterone-receptor antagonists	0 (0)	12 (29)	0.018
Statin	5 (33)	36 (88)	<0.001
NYHA functional class (I/II/III/IV)	–	12/27/2/0	–
Anterior MI (%)	–	18 (44)	–
Primary PCI (%)	–	22 (54)	–
Age of infarct (years)	–	5 (1–15)	–

Variables given are mean ± SD or N (%) or median (interquartile range).

ACEi=angiotensin-converting enzyme inhibitors; ARB=angiotensin-receptor blockers; BMI=body mass index; MI=myocardial infarction NYHA=New York Heart Association; PCI=percutaneous coronary intervention

Table 2.

Clinical characteristics of the study population

Characteristics	Control subjects (n=15)	Chronic MI patients (n=41)	p Value
LV EDV (ml)	114.9 ± 23.1	210.8 ± 63.5	<0.001
LV EDV index (ml/m2)	66.8 ± 10.1	107.9 ± 29.8	<0.001
LV ESV (ml)	44.6 ± 10.4	132.5 ± 55.8	<0.001
LV ESV index (ml/m2)	25.9 ± 5.1	68.4 ± 29.0	<0.001
LV stroke volume (ml)	70.3 ± 13.9	78.2 ± 24.3	0.24
LV ejection fraction (%)	61.3 ± 3.4	38.4 ± 11.0	<0.001
LV mass (g)	65.9 ± 14.6	118.1 ± 31.0	<0.001
LV mass index (g/m2)	38.3 ± 6.4	60.6 ± 15.3	<0.001
LV mass/LV EDV (g/ml)	0.58 ± 0.10	0.59 ± 0.18	0.92
Sphericity index	0.49 ± 0.12	0.55 ± 0.16	0.21
LVDd (short axis) (mm)	47.7 ± 4.0	57.3 ± 7.5	<0.001
LV length (4 chamber) (mm)	76.9 ± 6.6	90.7 ± 9.6	<0.001
RV EDV (ml)	119.3 ± 24.0	133.2 ± 41.1	0.22
RV EDV index (ml/m2)	69.3 ± 10.5	67.7 ± 17.4	0.74
RV ESV (ml)	48.6 ± 13.3	59.1 ± 27.7	0.06
RV stroke volume (ml)	70.7 ± 14.9	74.1 ± 21.9	0.58
RV ejection fraction (%)	59.5 ± 6.0	56.7 ± 10.5	0.33
LGE, n (%)	0 (0%)	41 (100%)	<0.001
Infarct size (g)	–	13.2 ± 7.6	–
Infarct size/LV mass (%)	–	11.6 ± 6.6	–
Average remote T1 (msec)	1068 ± 25	1098 ± 35	0.003
Remote T1 in apical slice (msec)	1079 ± 28	1107 ± 40	0.02
Remote T1 in mid slice (msec)	1056 ± 31	1096 ± 37	<0.001
Remote T1 in basal slice (msec)	1075 ± 32	1094 ± 42	0.11

Variables given are mean ± SD or N (%)

LV=left ventricular; EDV=end-diastolic volume; ESV=end-systolic volume; RV=right ventricular; LGE=late gadolinium enhancement; Dd=diastolic diameter.

Figure 1. Comparison of native remote myocardium T₁.

Native T₁ time was significantly elevated in patients with chronic myocardial infarction in comparison to healthy control subjects. There was no significant difference in native T₁ time between anterior MI vs. non-anterior MI.

*P<0.05 after ANOVA with Bonferroni’s multiple comparison Non anterior MI vs. control

^†P<0.05 after ANOVA with Bonferroni’s multiple comparison Anterior MI vs. control

Figure 2. Representative cases with inferior myocardial infarction.

Example cases (a) 61 years old, 3 yrs post anteroseptal myocardial infarction and remote T₁ of 1096 msec. Moderate LV systolic dysfunction was documented with mildly dilated LV volume and thinning of the mid and distal anteroseptal wall and apex. LGE image by cardiac MR showed transmural enhancement in the mid and distal septum, anterior wall and apex. (b) 68 years old, 2.5 yrs post similar anteroseptal myocardial infarction and high remote T₁ of 1164 msec. LGE infarct was transmural, but infarct size was relatively smaller than that of case A on LGE images. There was severe LV systolic dysfunction with mild hypertrophy in the remote myocardium.

Figure 3. Correlation between native remote T₁ value and LV volumes, mass and ejection fraction.

Remote native T₁ time was moderately correlated with LV a) ejection fraction, b) end-diastolic volume index, c) end-systolic volume index, and d) LV mass index.

Table 3.

Univariate regression analysis between native T₁ in remote myocardium in all subjects (n=56)

	beta (95% CI for the coefficient)	p-value
Patient characteristics
Age	0.01 (−1.01, 1.09)	0.94
Gender	0.19 (−5.29, 33.21)	0.15
Body mass index	−0.05 (−2.41, 1.66)	0.71
Systolic blood pressure	−0.28 (−1.07, −0.03)	0.039
Heart rate	0.18 (−0.24, 1.18)	0.19
Hypertension	0.16 (−7.97, 32.18)	0.23
Diabetes mellitus	0.07 (−15.52, 26.43)	0.60
Any MI	0.39 (10.61, 50.18)	0.003
Primary PCI	−0.20 (−35.80, 8.10)	0.21
Infarct age	−0.19 (−2.35, 0.85)	0.34
CMR findings
LV ejection fraction	−0.49 (−1.83, −0.64)	<0.001
LV end-diastolic volume index	0.38 (0.14, 0.70)	0.004
LV end-systolic volume index	0.45 (0.23, 0.78)	0.001
LV mass index	0.50 (0.56, 1.55)	<0.001
LV mass/LV end-diastolic volume	0.18 (−20.24, 98.60)	0.19
Sphericity index	−0.05 (−74.23, 53.32)	0.74
RV ejection fraction	−0.17 (−1.61, 0.38)	0.22
RV end-diastolic volume index	−0.13 (−0.90, 0.31)	0.34
RV end-systolic volume	0.06 (−0.30, 0.46)	0.67
Infarct size, g	0.19 (−0.57, 2.37)	0.22
Infarct size, % LV mass	−0.01 (−1.75, 1.68)	0.97

Abbreviations as in Table 1 and 2.

Table 4.

Univariate regression analysis between LVEF in chronic MI patients (n=41)

	beta (95% CI for the coefficient)	p-value
Patient characteristics
Age	0.07 (−0.30, 0.45)	0.68
Gender	0.21 (−2.59, 12.96)	0.19
Systolic blood pressure	0.32 (0.01, 0.40)	0.043
Hypertension	−0.24 (−12.56, 1.65)	0.13
Diabetes mellitus	−0.15 (−11.05, 3.92)	0.34
Anterior MI	0.03 (−6.37, 7.83)	0.84
Primary PCI	0.08 (−5.28, 8.81)	0.62
Medications
ACEi/ARB	−0.18 (−15.19, 4.45)	0.28
Beta blocker	0.20 (−4.85, 21.67)	0.21
Aldosterone-receptor antagonist	−0.14 (−10.99, 4.36)	0.39
Statin	0.16 (−5.19, 16.07)	0.31
CMR findings
Native T1 in remote myocardium	−0.35 (−0.20, −0.02)	0.023
Infarct size, g	−0.36 (−1.01, −0.04)	0.034
Infarct size, % LV mass	−0.31 (−1.04, −0.01)	0.046

Abbreviations as in Table 1 and 2.

Discussions

The present study demonstrated that 1) native remote myocardium T₁ is elevated in patients with prior MI compared with age-matched controls, 2) native remote myocardium T₁ and infarct size have similarly moderate correlation with reduced LVEF, and 3) in multivariate analysis, native T₁ independently correlates well with reduced LVEF beyond infarct. To the best of our knowledge, this is the first study to assess native remote myocardium T₁ in chronic MI patients using a slice interleaved T₁ mapping sequence (STONE) and to compare native remote myocardium T₁ with infarct size against LV dysfunction in the chronic phase after MI.

A diffuse myocardial fibrosis is a marker for subclinical disease, a fundamental feature of myocardial remodeling²², an independent predictor of adverse cardiovascular events^23,24, and a potential target for medical interventions. Despite the importance of this measure of diffuse myocardial fibrosis, previous reports of post-MI remodeling were based on studies in infarct size and clinical findings^25,26. Several investigators recently attempted to elucidate myocardial tissue characterization in the remote myocardium by using noninvasive T₁ mapping. In a study by Carrick et al., native remote myocardium T₁ predicted LV adverse remodeling in patients with acute MI¹⁴. However, T₁ mapping was acquired in the acute MI phase, without comparison of its diagnostic utility in the chronic phase with LV dysfunction. Native remote myocardium T₁ in the acute phase theoretically mirror the extent of myocardial edema as well as extracellular volume expansion and does not necessarily have a strong correlation with histological fibrosis in the chronic phase. Therefore, increased native remote myocardium T₁ value shortly after MI may represent the initiation of diffuse fibrosis as suggested by prior animal studies. In addition, our observations are similar to Chan et al. that post-contrast remote myocardium T₁ is shorter in both patients with subacute and chronic MI¹³.

In contrast to the studies of Anversa et al²⁷. and Rumberger et al.²⁶, which found that the magnitude of post-MI LV systolic dysfunction is related to infarct size, we found that native remote T₁ value was associated with LV dysfunction beyond infarct size. Moreover, it should be noted that the elevated native remote myocardium T₁ was a powerful, independent predictor for reduced LVEF, even after taking potential confounders into consideration. These findings are consistent with previous animal and human studies^11,12,28,29 and suggest that post-MI LV dysfunction is related to diffuse myocardial fibrosis in the remote myocardium in patients with relatively small MI, which plays a pivotal role in the development, progression, and clinical manifestations of heart failure.

In a recent animal study of Kali et al³⁰, native T1 at 3T provided the similar diagnostic accuracy for detecting infarct location, size, and transmurality of chronic MI to LGE image. The present study demonstrated that native T1 method can be applicable for assessment of non-infarct remote myocardium mostly associated with post-MI LV remodeling. Considering this information not given by LGE image, shorter scan time and non-contrast material requirement, the comprehensive native T1 map approach has the potential for widespread clinical application and might improve outcomes in MI patient with chronic kidney disease.

Our study has several limitations. First, the present study is a single-center study of relatively small sample size. Second, traditional coronary risk factors potentially may contribute to the elevation of native remote myocardium T₁ value. However, given that fibroblast stimulation increases collagen synthesis and causes fibrosis of both the infarcted and noninfarcted myocardium after MI, the elevated native remote myocardium T₁ following MI could contribute to adverse LV remodeling.

In conclusion, LV systolic dysfunction in chronic MI is independently associated with diffuse myocardial fibrosis in the remote myocardium. The current results should alert the clinician to the potential coexistence of diffuse myocardial fibrosis in the remote myocardium. Larger, multicenter studies are needed to further confirm whether these results represent a potential therapeutic target.

Multivariate Stepwise Analysis of Chronic MI Subjects with reduced LVEF as the Dependent Variables (n=41)

	β	SE	p-value
Native remote myocardium T1, msec	−0.34	0.047	0.029

The clinical characteristics that were univariate predictors of LVEF that were also included in the multivariate model were age (p=0.65), male (p=0.051), systolic blood pressure (p=0.17), hypertension (p=0.26) and infarct size (g) (p=0.079).