Synopsis

Keywords: Myocardium, Cardiomyopathy

This study evaluated a cardiac T1ρdispersion mapping technique, termed as myocardial fibrosis index (mFI), to distinguish cardiac amyloidosis (CA) from hypertrophic cardiomyopathy (HCM). Our data showed that mFI is significantly superior to native T1 and contrasts medium-derived extracellular volume in the differential diagnosis of CA and HCM.

Introduction

There are challenges in diagnosing CA because of varying clinical and imaging manifestations of the disease. Similar imaging appearances contribute to CA patients being misdiagnosed early as having HCM. Moreover, there are significant differences in the treatment and prognosis of HCM and CA [1]. Therefore, early and timely diagnosis is essential for the CA clinical treatment and prognosis evaluation. Using new CMR imaging biomarkers might provide input from different pathophysiological pathways that could add essential data for CA diagnosis. We hypothesized that contrast-free myocardial T1ρ dispersion (termed myocardial fibrosis index [mFI] maps) [2] might allow for the differential diagnosis of the disease. This study thus aimed to evaluate the value of the T1ρ dispersion mapping technique for detecting CA from HCM.

Methods

We retrospectively studied 30 subjects with CA and 30 with HCM (Table 1). All subjects underwent cine, T1ρ-mapping (B0 0Hz and B1 298Hz), pre- and post-contrast T1-mapping, and late gadolinium enhancement imaging on a 3T magnetic resonance system (MAGNETOM Skyra, Siemens Healthcare, Erlangen, Germany). Myocardial extracellular volume fraction (ECV) maps were calculated using pre- and post-contrast T1 maps. Myocardial T1ρ and T1ρ dispersion maps, called myocardial fibrosis index (mFI) maps, defined the difference in the T1ρ maps to between 0 and 298 Hz. In addition, all quantitative parameters of the mid-short-axis slice were compared.

Results

Both native T1 and ECV values were significantly higher in patients with CA compared with the HCM group (native T1, HCM: 1274.7±54.3 ms vs. CA: 1428.2±71.9 ms; P < 0.0001), (ECV, HCM: 32.6±5.8% vs. CA: 46±8.9%; P < 0.0001) (Table 2). In contrast, mFI values were significantly lower (Fig.1) in the CA group compared with the HCM group (HCM: 8.8±2.8 ms vs. CA: 3.4±2.1 ms; P < 0.0001) (Table 2). In addition, there was a negative correlation between the ECV and mFI (r = -0.314, P < 0.05) (Table 3). The diagnostic accuracy of native T1, ECV, and mFI to diagnose CA from patients with HCM was assessed on an ROC curve. The mFI index showed the highest area under the curve (AUC) (0.957; 95% CI, 0.937-0.978; P < 0.0001) (Fig. 2a) compared to ECV (0.911; 95% CI, 0881-0.940; P < 0.0001) and native T1 (0.827; 95% CI, 0.777-0.876; P < 0.0001) (Table 4) (Fig.2b).

Discussion & Conclusion

T1 mapping and ECV could be used to measure the intrinsic signal of the myocardium in CA. Although these markers have pathognomonic features, they are not specific to CA and may be elevated in other cardiovascular diseases. In addition, MFI was low and negatively correlated with ECV in patients with CA, a finding that might be specific and could be used as a marker for the diagnosis of CA.

Table 1. Clinical demographics of the participants in the two groups.

	HCM (n=30)	CA (n=30)	P-value
Sex (male/female)	21/9	18/12
Age (years)	53.6 ± 11.3	49.7±12.9	0.513
Body surface area (m2)	1.8±0.3	2±0.6	0.343
Heart rate (beats/min)	68.7 ± 9.3	69.5 ± 8.7	0.522
Troponin	0.02 (0.013,0.122)	0.1(0.07,0.756)	< 0.0001
Brain natriuretic peptide	1000 (378,2423)	7926(6987,15863)	< 0.0001
NYHA class, I/II/III/IV	5/18/6/1	0/2/3/25	0.001
Hematocrit (%)	39 (36,44.1)	34.3(32.1,35.9)	0.001
Arrhythmia	5/30	18/30	< 0.0001

NYHA: New York Heart Association; LV: left ventricle; LVEDV: LV end-diastolic volume; LVESV: LV end-systolic volume; LVSV: LV stroke volume; LVEF: LV ejection fraction; LGE: late gadolinium enhancement.

Table 2. CMR characteristics of the participants in the two groups.

Parameters	HCM (n=30)	CA (n=30)	P
LVEDV (ml)	119.0 ± 24.0	153.5 ± 67.2	0.012
LVESV (ml)	34.6 ± 10.2	96 ± 7.3	< 0.0001
LVSV (ml)	84.4 ±18.6	57.4 ± 21.4	0.001
LVEF (%)	70.9 ± 6.0	58.1 ± 12.6	< 0.0001
Mass (g)	159.9 ± 46.2	166.6 ± 46.3	0.731
Global diastolic thickness (mm)	11.5±3.7	11.2 ± 3.2	0.278
No LGE/patchy/transmural	3/25/2	2/3/25	< 0.0001
Native T1 (ms)	1274.7 ± 54.3	1428.2 ± 71.9	< 0.0001
Post T1 (ms)	637.5 ± 89.5	671.9 ± 80.6	0.001
ECV (%)	32.6 ± 5.8	46 ± 8.9	< 0.0001
mFI (ms)	8.8 ± 2.8	3.4 ± 2.1	< 0.0001

LVEDV: LV end-diastolic volume; LVESV: LV end-systolic volume; LVSV: LV stroke volume; LVEF: LV ejection fraction; LGE: late gadolinium enhancement.
ECV: extracellular volume fraction; mFI: myocardial fibrosis index.

Table 3. Correlation of CMR indexes.

		HCM, mFI			CA, mFI
		r	P		r	P
Native T1		0.301	< 0.0001		0.192	0.028
Post T1		-0.48	< 0.0001		0.088	0.316
ECV		0.876	< 0.0001		-0.314	< 0.0001

ECV: extracellular volume fraction

Table 4. Diagnostic accuracy of native T1, ECV, and mFI to diagnose CA in patients with HCM using the ROC curve.

Parameters	Area	95% CI low	95% CI up	Cut-off	Sensitivity	Specificity
Native T1 (ms)	0.827	0.777	0.876	1386.2	68.7%	92.8%
ECV (%)	0.911	0.881	0.940	39.9	82.4%	90.5%
mFI (ms)	0.957	0.937	0.978	5.0	97.5%	85.5%

ECV: extracellular volume fraction; mFI: myocardial fibrosis index.

Acknowledgements

No acknowledgement found.