Histological Validation of Cardiovascular Magnetic Resonance T1 Mapping for Assessing the Evolution of Myocardial Injury in Myocardial Infarction: An Experimental Study.

OBJECTIVE: To determine whether T1 mapping could monitor the dynamic changes of injury in myocardial infarction (MI) and be histologically validated.
MATERIALS AND METHODS: In 22 pigs, MI was induced by ligating the left anterior descending artery and they underwent serial cardiovascular magnetic resonance examinations with modified Look-Locker inversion T1 mapping and extracellular volume (ECV) computation in acute (within 24 hours, n = 22), subacute (7 days, n = 13), and chronic (3 months, n = 7) phases of MI. Masson's trichrome staining was performed for histological ECV calculation. Myocardial native T1 and ECV were obtained by region of interest measurement in infarcted, peri-infarct, and remote myocardium.
RESULTS: Native T1 and ECV in peri-infarct myocardium differed from remote myocardium in acute (1181 ± 62 ms vs. 1113 ± 64 ms, p = 0.002; 24 ± 4% vs. 19 ± 4%, p = 0.031) and subacute phases (1264 ± 41 ms vs. 1171 ± 56 ms, p < 0.001; 27 ± 4% vs. 22 ± 2%, p = 0.009) but not in chronic phase (1157 ± 57 ms vs. 1120 ± 54 ms, p = 0.934; 23 ± 2% vs. 20 ± 1%, p = 0.109). From acute to chronic MI, infarcted native T1 peaked in subacute phase (1275 ± 63 ms vs. 1637 ± 123 ms vs. 1471 ± 98 ms, p < 0.001), while ECV progressively increased with time (35 ± 7% vs. 46 ± 6% vs. 52 ± 4%, p < 0.001). Native T1 correlated well with histological findings (R² = 0.65 to 0.89, all p < 0.001) so did ECV (R² = 0.73 to 0.94, all p < 0.001).
CONCLUSION: T1 mapping allows the quantitative assessment of injury in MI and the noninvasive monitoring of tissue injury evolution, which correlates well with histological findings.

INTRODUCTION

Although mortality after myocardial infarction (MI) is on the decline, the subsequent complications remain one of the significant causes of death worldwide (12). Myocardial injury in MI is closely associated with long-term recovery and prognosis (345). Therefore, the accurate evaluation of the severity of myocardial damage plays an essential role in cardioprotective therapies and risk stratification. Cardiovascular magnetic resonance (CMR) has currently evolved into a gold standard for tissue characterization and quantification in MI. However, late gadolinium enhancement (LGE) imaging is of limited value to the quantitative assessment of tissue injury in acute MI (67). T1 mapping has been proposed as a promising technique for tissue characterization due to its accurate diagnosis (8910). It provides complementary information to standard LGE imaging with excellent performance in detecting myocardial edema (111213) and fibrosis (14151617). Furthermore, several recent studies demonstrated that native T1 mapping and extracellular volume (ECV) maps could quantify the severity of the injury and predict the left ventricular (LV) function recovery or adverse remodeling (818), adding incremental prognostic value over LGE. However, major efforts related to these works are being applied in clinical practice, while further histological validation is still lacking (18). Accordingly, we sought to histologically validate the feasibility of T1 mapping for quantifying the myocardial injury and monitoring its evolution in the MI pig model.

MATERIALS AND METHODS

Experimental Animal

All experimental protocols in this study were approved by the Institutional Animal Care and Use Committees. Twenty-three male Bama mini pigs with an average weight of 8.8 kg were used in this study. The pigs were premedicated with atropine sulfate (Atropine Sulfate Injection, Jieyang Longyang Animal Pharmaceutical Co., Ltd.) (0.05 mg/kg). 10–15 minutes later, anesthesia was induced by intramuscular injection with Zoletil 50 (Virbac) (10–15 mg/kg). Following intubation with an endotracheal tube, an intravenous channel was established in ear vein to sustain anesthesia with continuous propofol (10 mg/mL, Propofol Injectable Emulsion, XiAn Libang Pharmaceutical Co.,Ltd.). MI was induced by left anterior descending artery ligation for 40 minutes via thoracotomy and was confirmed by the elevation of ST segment on electrocardiogram. The wound and the surgical area were wrapped with povidone-iodine disinfectant after sutured. After drying, the wound was coat with dry penicillin powder and wrapped with sterile gauze, which was changed once a day. To prevent infection, penicillin (100000 IU/kg) was administered intramuscularly twice daily for 3–5 days.

Despite efforts of resuscitation, one pig died before CMR scanning. The surviving pig models underwent serial CMR scans in acute (within 24 hours), subacute (7 days), and chronic (3 months) MI after surgery. Following the examinations, the pigs were sacrificed for histological staining on the same day. The experimental procedure is presented in Figure 1.

Fig. 1

The following chart shows the different steps from coronary artery occlusion to histological staining.

*Two of the 22 pigs were euthanasia for other purposes. CMR = cardiovascular magnetic resonance, MI = myocardial infarction

Cardiovascular Magnetic Resonance Study

The CMR scanning was conducted using a clinical 3T MR scanner (MAGNETOM Skyra, Siemens Healthineers) equipped with an 18-channel receive coil. CMR protocols included cine, T1 mapping, and LGE. The LV function was assessed with a balanced steady-state free precession pulse sequence (echo time [TE] = 1.70 msec; repetition time [TR] = 46.68 msec; flip angle = 34°; slice thickness 5 mm; matrix = 128 × 128 pixels; field of view [FOV] = 130 × 130 mm2). A modified Look-Locker inversion recovery method was used to obtain native T1-maps and post-contrast T1-maps with a 3-3-5 acquisition within one breath-hold's time. Native T1-maps (TE = 1.27 msec; TR = 344.48 msec; flip angle = 35°; slice thickness = 5 mm; matrix = 110 × 192 pixels; FOV = 176 × 150 mm2) were timed consecutively in three standard short-axis levels (basal, mid, and apical), which were acquired from the middle 3 of 5 parallel short-axis slices spaced equally between the mitral annulus and the LV apical tip. Post-contrast T1-maps (TE = 1.27 msec; TR = 344.48 msec; flip angle = 35°; slice thickness = 5 mm; matrix = 110 × 192 pixels; FOV = 136 × 136 mm2) were imaged at approximately 15 minutes after intravenous injection of gadolinium at a dose of 0.15 mmol per kg body weight. Breath-hold LGE-imaging was subsequently acquired at 16–20 minutes by using segmented-turbo-FLASH-PSIR (TE = 1.33 msec; TR = 476.60 msec; flip angle = 40°; slice thickness = 5 mm; matrix = 110 × 192 pixels; FOV = 136 × 136 mm2). Additionally, cine, LGE, and T1 mapping imaging were carried out using a contiguous stack of short-axis slices covering the whole LV for each acquisition. For animals, a breath-hold was acquired during expiration breath-hold mode with a ventilator throughout the CMR examination. At every follow-up stage, comprehensive CMR scans were performed until sacrifice (i.e., animals sacrificed at the chronic stage underwent acute, subacute, and chronic CMR examinations).

Image Analysis

MRI images were analyzed offline using commercially available software (cvi42, Circle Cardiovascular Imaging) (Fig. 2). The basic parameters of LV function were evaluated globally using cine short-axis stack views according to Society for Cardiovascular Magnetic Resonance for reporting of standard post-processing (19). An area of LGE ≥ 5 standard deviations (SD) above remote myocardium (2021) was obtained to examine infarct size, expressed as a percentage of total LV mass. Additionally, microvascular obstruction (MVO) was identified as a low-intensity black area in the middle of the hyperintense zone on LGE images and was included in the measurement of LGE infarct with drawing manually.

Fig. 2

LGE images (left), native T1 maps (middle), ECV maps (right).

The infarcted myocardium was defined using the threshold of 5 standard deviations above the remote myocardial signal intensity. Microvascular obstruction was defined as a hypointense area within the hyperintensity on LGE and included in the measurement of infarct size. The optimal contrast among infarcted, peri-infarcted, and remote myocardium was observed in native T1 and ECV maps in acute and subacute MI but was not significant in the chronic stage. ECV = extracellular volume, LGE = late gadolinium enhancement

For the region of interest (ROI)-based T1 mapping measurement, the ROIs were placed in infarcted myocardium (defined as the matching hyperintense area on LGE), peri-infarcted myocardium (defined as the region adjacent to the infarction), and remote myocardium (defined as the area with no LGE and movement dysfunction). To measure ECV, pre- and post-contrast blood T1 times were obtained in additional ROIs manually delineated in the center of the LV cavity (trabeculations or papillary muscles were excluded). ECV was quantified from pre- and post-contrast T1 times according to the established formula (22): ECV fraction = (1 − hematocrit) × (T1myo post − T1myo pre) / (T1blood post − T1blood pre). The ROIs of various regions on pre-contrast images were transposed on the ECV maps with manual and minor correction if needed. The native T1 and ECV of MVO were also derived by ROI-based measurement. For analysis purposes, the characteristics measured with different techniques were confined to the same location, and the delineation of ROIs was performed by the consensus of two radiologists (with three years of experience and with four years of experience).

Histological Analysis

Following CMR examinationss, all pigs were immediately euthanized using potassium chloride. The hearts of pigs were collected and cleaned in saline, and then were left into plastic containers to cool down for approximately 20 minutes at 20 degrees Celsius below zero until solid. The hearts were cut into short-axis slices with a thickness of 5 mm. After fixed and embedding, slices were cut into sections of 5 µm, followed by hematoxylin and eosin and Masson's trichrome staining. Quantitative analysis was performed using Image-Pro Plus 6.0 software (Media Cybernetics). For each sample, 3 visual fields were taken in the infarcted myocardium, 2 on each side of peri-infarcted myocardium (i.e., with approximately 4–6 mm proximal to the infarct) and 1 in the remote myocardium (6). ECV derived by histology was expressed as (extracellular space / total area) × 100, after removing artifacts and perivascular and endocardial fibrous tissue by manually differentiating the extracellular space from viable myocytes with careful consideration. The ECV for each region was averaged in the defined myocardium of each animal.

Statistical Analysis

All statistical data analyses were conducted using SPSS software (v. 17.0 for Windows, SPSS Inc.). The results are expressed as the mean ± SD or median with range, as appropriate. The Shapiro-Wilk test and Leven's test were performed to check normal distribution and homogeneity of variances, respectively. To analyze the changes in all obtained parameters with time, one-way ANOVA or nonparametric tests were utilized. The linear regression was employed to compare native T1 and ECV by CMR versus ECV by histology, respectively. For all comparisons, two-tailed p values < 0.05 were considered statistically significant.

RESULTS

The Baseline Characteristics of the Experimental Study

The baseline characteristics of the experimental study are described in Table 1. Infarct size was significantly decreased from acute to subacute phase (8.4 ± 1.6% vs. 5.8 ± 2.4%, p = 0.017) but showed no difference between subacute and chronic phase (5.8 ± 2.4% vs. 4.0 ± 1.9%, p = 0.317). From acute to chronic phases, the improvements in LV functional parameters were observed as shown in Table 1. The peak creatine kinase release was obviously higher in the acute phase than in the subacute and chronic phases (p = 0.027).

Table 1

Baseline Characteristics of Myocardial Infarction Pigs

	Acute (n = 22)	Subacute (n = 13)	Chronic (n = 7)	P
LVEF, %	36.3 ± 15.9	41.2 ± 14.3	56.4 ± 7.5*	0.012
LV EDV, mL	21.1 ± 10.0	16.7 ± 6.5	34.3 ± 13.7*†	0.003
LV ESV, mL	12.7 ± 4.1	9.5 ± 2.2	13.9 ± 4.3	0.062
SV, mL	7.9 ± 5.8	7.4 ± 4.9	19.0 ± 9.5*†	0.001
CO, mL	0.7 (0.4–1.5)	0.7 (0.4–1.3)	1.8 ± 0.8	0.463
LV mass, g	12.4 (9.9–30.0)	13.6 (11.5–14.9)	32.2 ± 8.5*†	0.01
Infarct size, %	8.4 ± 1.6	5.8 ± 2.4*	4.0 ± 1.9*	< 0.001
Troponin I, µg/L	0.075 (0.047–0.113)	< 0.012	< 0.012	-
CK, U/L	1019 ± 444–10590	262 ± 163*	403 ± 247	0.027
CK-MB, µg/L	1.3 ± 0.6–2.6	0.8 ± 0.4	0.7 ± 0.3	0.162
Mb, µg/L	302 ± 222	69.4 (19.9–172.8)	127.1 ± 57.1	0.085
Hct, %	37.2 ± 4.4	35.5 ± 3.8	42.5 ± 4.0*†	0.005

Data are presented as the mean ± standard deviation, or median (interquartile range). *p < 0.05 vs. acute, †p < 0.05 vs. subacute. CK = creatine kinase, CK-MB = creatine kinase-MB, CO = cardiac output, EDV = end-diastolic volume, ESV = end-systolic volume, Hct = hematocrit, LV = left ventricular, LVEF = left ventricle ejection fraction, Mb = myoglobin, SV = stroke volume

Time Course of Native T1 and ECV in Different Myocardial Regions

The quantitative values of native T1 and ECV derived from CMR are summarized in Table 2. Native T1 of peri-infarct myocardium in the acute phase was significantly higher than that of remote myocardium (p = 0.002), but less than that of infarcted myocardium (p < 0.001) as was ECV (p = 0.031 and p < 0.001, respectively). This trend persisted in the subacute study (all p < 0.05). However, the difference between peri-infarct and remote myocardium disappeared either in native T1 (p = 0.934) or in ECV (p = 0.109) in the chronic setting. Furthermore, the two parameters of infarcted myocardium were still larger than those of remote myocardium in chronic MI (all p < 0.05).

Table 2

Dynamic Changes in Myocardial Native T1 and ECV

	Native T1 Time, ms				ECV, %
	Infarcted	Peri-Infarct	Remote	P	Infarcted	Peri-Infarct	Remote	P
Acute (n = 22)	1275 ± 63	1181 ± 62*	1113 ± 64*†	< 0.001	35 ± 7	24 ± 4*	19 ± 4*†	< 0.001
Subacute (n = 13)	1637 ± 123	1264 ± 41*	1171 ± 56*†	< 0.001	46 ± 6	27 ± 4*	22 ± 2*†	< 0.001
Chronic (n = 7)	1471 ± 98	1157 ± 57*	1120 ± 54*	< 0.001	52 ± 4	23 ± 2*	20 ± 1*	< 0.001

Data are presented as the mean ± standard deviation. *p < 0.05 vs. acute, †p < 0.05 vs. subacute. ECV = extracellular volume

Interestingly, from acute to chronic phases, native T1 of infarcted myocardium peaked in the subacute phase (acute vs. subacute vs. chronic: 1275 ± 63 ms vs. 1637 ± 123 ms vs. 1471 ± 98 ms, p < 0.001), while a progressive elevation was observed in ECV over 3 months (acute vs. subacute vs. chronic: 35 ± 7% vs. 46 ± 6% vs. 52 ± 4%, p < 0.001). Similar to the trend of native T1 in infarcted myocardium, both native T1 and ECV of peri-infarct myocardium reached peak values in the subacute phase and then decreased in the chronic phase (Fig. 3B, C). Besides, no difference was observed in remote myocardium at all imaging time-points (all p > 0.05).

Fig. 3

Native T1, CMR-derived ECV, and pathological findings varied with time in different severity of myocardium injury.

A. The histological ECV showed a trend in infarcted and peri-infarcted myocardium, similar to the results of CMR-based ECV. B. Infarcted native T1 reached a peak value at the subacute phase. C. Infarcted ECV progressively increased with time.

In the acute phase, 15 out of 22 pigs presented MVO (68.2%). The native T1 of the MVO (1219 ± 49 ms) was higher than that of the remote myocardium (1105 ± 55 ms, p < 0.001) but lower than that of the infarcted myocardium (1282 ± 66 ms, p = 0.03). ECV of the MVO (15 ± 6%) was lower than that of the infarcted myocardium (38 ± 10%, p < 0.001) and had no significant difference compared with the remote myocardium (20 ± 3%, p = 0. 338).

Measurement of ECV Using Histology

In total, 120 microscopy regions from 18 pigs were used in the analysis; two pigs in the chronic phase were excluded because of poor stained. Figure 4 shows the typical histological features of injured myocardium in MI. Damaged tissue following MI evolved from the loss of nuclei, myocardial edema, and inflammatory cell infiltration to the replacement of collagenous fiber. The ECV of peri-infarct myocardium differed from that of remote myocardium in acute (p < 0.001) and subacute MI (p < 0.001) but disappeared in chronic MI (p = 0.668). Furthermore, a longer duration of myocardial ischemia resulted in more serious ECV expansion in the infarcted myocardium (p < 0.001) (Fig. 3A).

Fig. 4

Representative sections from three areas of the myocardium: remote myocardium, peri-infarcted myocardium, and infarcted myocardium (magnification, × 200).

H&E staining showed myocytolysis, inflammatory cell recruitment, and intracellular and extracellular edema in acutely infarcted myocardium. The inflammation and edema of the infarcted myocardium gradually released over time. On the other hand, Masson staining (red = myocardial cells) demonstrated the progressive deposition of collagen and increased ECV in infarcted myocardium with a longer duration of ischemia. H&E = hematoxylin and eosin

Relationship of CMR-Derived Native T1 and ECV with Histological Results

From acute to chronic MI, an excellent correlation was found in the comparison between native T1 and histological ECV (R2 = 0.65 to 0.89, all p < 0.001) (Fig. 5A–C). Similarly, CMR-measured ECV showed a strong association (R2 = 0.73 to 0.94, all p < 0.001) (Fig. 5D–F) and close agreement with histological ECV ([acute: bias, 1.4 ± 5.7], [subacute: bias, 1.4 ± 8.2], [chronic: bias, 0.4 ± 9.4]).

Fig. 5

Correlation between T1 mapping (native T1 and ECV) and histological ECV at three time-points.

Both native T1 and ECV correlated well with histological ECV at each imaging time point.

DISCUSSION

This study performed a longitudinal assessment of the variation of native T1 and ECV at multiple time points in the MI pig model and conducted a corresponding histological validation. The main findings of our study are that 1) the patterns of the myocardium in native T1 and ECV were different, showing that native T1 reached a peak value in the subacute phase while ECV progressively increased over 3 months post-infarction, and 2) native T1 and ECV could quantify the severity of myocardial damage in MI and were correlated well with histology.

Native T1 mapping and ECV measurements have accurate performance in the quantitative evaluation of myocardial injury after MI (818). These findings are explained by the results from our analysis of mean native T1 and ECV in different areas of the myocardium, which in our study were investigated by serial cardiac MR scans of pig models in the acute, subacute and chronic stages. Our data showed that from remote to infarcted myocardium, the progressive elevation occurred in both native T1 and ECV in acute and subacute phases, which is consistent with the data of previous study demonstrating myocardial blood flow was worst in infarcted myocardium (8). In the chronic stage, we found no difference in the two parameters between the peri-infarct and remote myocardium, which indicates the resorption of edema and tissue healing occurring in the adjacent to the infarcted zone. Additionally, our results are agreement with the findings in published literature reporting higher native T1 and ECV in infarcted myocardium compared with remote myocardium (2324).

From acute to chronic study, native T1 of infarcted myocardium showed a wave-like trend with a peak value in the subacute phase. This trend is similar to the new perspective (252627) that myocardial edema presents a bimodal phenomenon post-infarction, although the current study was not performed at earlier imaging times. The peak value of infarcted native T1 might indicate severe myocardial edema after the first week in MI because of the inflammation response and tissue healing-related granulation developing (27). Interestingly, ECV of the infarcted myocardium showed a different trend as progressive elevation. ECV allows the measurement of extracellular space expansion and appears as a quantitative map due to gadolinium being trapped in the significant expanded extracellular space and intra-cell through the raptured membrane and capillaries after acute MI. As myocardial edema resolved after MI, already damaged cells were increasingly replaced by extensive collagenous deposition with time (27), leading to an increase in gadolinium concentration (28). Several researches have demonstrated that ECV could serve as a reliable quantitative marker of myocardial fibrosis (2930). Thus, the observed progressive ECV elevation over time potentially suggests that myocardial fibrosis became increasingly severe in the infarcted myocardium, which is in line with the exacerbating in fibrous deposition observed in our histological study.

The occurrence of MVO has different influences on native T1 and ECV. Native T1 increased due to tissue edema by activation of inflammatory reactions in the area of MVO (31). On the contrary, the blood degradation products, such as deoxyhemoglobin, could prolong T1 relaxation time, resulting in a lower native T1 compared to the infarcted myocardium without MVO. As for the influence of MVO on ECV, the microvasculature surrounding MVO is morphological intact, severe disruption of endothelial cells and microthrombi occur in the area of microvascular injury (32), leading to an inability of LGE penetrating. This mechanism might result in the failure measurement of true ECV value for MVO.

For MI modeling, pigs are more preferable than dogs due to the close similarities of the anatomical distribution of collateral circulation between the hearts of pigs and humans (3334). The islands and peninsulas pattern of necrosis at the edge of infarction yield the lateral extension of salvaged myocardium (6), which is also seen in humans (1213). The infarct size showed an apparent reduction in the first week after MI, consistent with previous reports (635) that LGE imaging might overestimate early infarct size in a pig model due to the resolution of edema at the edge of the infarcted region. The slight reduction of infarct size even seen in the subsequent imaging could be explained by the further infarct “shrinkage” (36). However, the ECV of infarcted myocardium increased progressively. Presumably, the myocardial injury is progressive though the infarct size is reducing after MI.

Histological features in MI are complex and dynamic changes over time (2736). Thus, an appropriate strategy is needed to monitor the evolution of myocardium injury which plays a key role in clinical therapy decision and prognosis prediction. Numerous studies have provided robust evidence that native T1 and ECV are susceptible to varying degrees of myocardial edema (1137) and collagenous fibrosis (38). The different trends between native T1 and ECV and the strong correlation compared with histological findings observed in the study suggest that native T1 mapping is more sensitive to edematous assessment, while ECV measurement is more helpful in the quantification of myocardial fibrosis in MI. Furthermore, these data are a reminder of the importance of appropriate time points for evaluating myocardial injury progression using T1 mapping parameters.

Limitations

The current study has several limitations. First, the native T1 and ECV of different degrees of severity of injured myocardium were measured by manually drawing ROIs, which are subject to the influence of subjective judgment. Despite this drawback, this measurement is widely used in existing clinical practice. Furthermore, the experimental study did not set up the baseline examination of the same pig. However, the pig cohorts were imaged at three-time points for monitoring the dynamic changes of myocardial injury after MI. Additionally, in this study, we did not assess post-T1 values because this index easily affects various factors during CMR examination (3940).

Conclusion

In conclusion, we successfully imaged and quantified the different severities of damaged myocardium using native T1 and ECV maps. These results indicated that T1 mapping could provide a noninvasive and robust means to assess and monitor the myocardial injury evolution post-infarction.