Multiparametric MRI detection of cerebral metabolism: a tool for early differentiation between MELAS and ischemic cerebral infarction

Introduction

Mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS) syndrome is one of the most common mitochondrial diseases caused by mutations in mitochondrial DNA (1). The diagnostic criteria for MELAS are based on clinical findings of stroke-like episodes and imaging examinations, together with evidence of mitochondrial dysfunction (2). Stroke-like episode is a characteristic manifestation of MELAS, and also one of the typical presentations of ischemic cerebral infarction (ICI). However, the increasing incidence of stroke in young people in recent years (3) has posed more challenges for clinical diagnosis. Conventional magnetic resonance imaging (MRI) features of MELAS have broad manifestations which are apt to be confused with ICI (4-6), viral or autoimmune encephalitis, or central nervous system vasculitis (7). Advanced MRI is increasingly being recognized as a valuable tool in the diagnosis and differentiation of MELAS (6,8-11).

Proton magnetic resonance spectroscopy (¹H-MRS) is able to detect the metabolic changes of the brain in vivo, which precede visible structural changes on conventional MRI in patients with MELAS (10). Accumulation of lactate (Lac) and decrease of N-acetyl-aspartate (NAA) are the most often reported alterations (12). However, both changes are not specific for MELAS because accumulation of Lac is also observed in ICI (13) and brain tumors (14). Perfusion-weighted imaging (PWI) can also be useful for the evaluation of MELAS. According to previous research, hyperperfusion in acute encephalomyopathy represents an imaging landmark distinct from ischemia (15-17). Many studies have shown almost consistent results regarding the perfusion changes around the time of a stroke: hyperperfusion in the acute stage and hypoperfusion in the subacute and chronic stages (8,18). Diffusion-weighted imaging (DWI) has been used for decades to diagnose ischemic pathology. This technique has also shown its worth in diagnosing other non-ischemic lesions, including mitochondrial disorders that could mimic cerebral ischemia. The consensus in the literature remains unclear whether DWI shows high signal intensity pertaining to the presence of vasogenic or cytotoxic edema in MELAS of acute phase (19,20).

Although MRS, PWI, and DWI have been applied to MELAS to explore the characteristic changes, they have not been combined to investigate the potential in the differential diagnosis of similar diseases, specifically, ICI. In addition, most studies of MELAS have simply focused on MRI-visible lesions (MVL) and little attention has been paid to normal-appearing regions (NAR) (19-21). A few studies have noted that Lac peak was present in NAR (10,22-24). Therefore, we hypothesized that MELAS, as a metabolic encephalopathy, will affect the whole brain, including both MVL and NAR. Therefore, in this study, we employed multiparametric MRI including MRS, DWI and PWI to explore the multi-dimensional changes of MELAS and its difference from ICI, both in MVL and NAR. We present this article in accordance with the CLEAR reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-2396/rc).

Methods

Patient enrollment

This study was approved by the Ethics Review Board of West China Hospital of Sichuan University (Ethical Approval No. 2020-92) and was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The requirement for informed consent was waived due to the minimal-risk nature of this retrospective analysis.

In this retrospective study, MELAS patients were enrolled between November 2011 and November 2020 in West China Hospital of Sichuan University. According to the diagnostic criteria of Yatsuga et al. (2), the definite diagnosis of MELAS was based on clinical presentation, pathology, plasma lactate, and MRI features. The inclusion criteria were as follows: (I) the diagnosis of MELAS was confirmed by biceps brachii muscle biopsy and pathology; (II) patients underwent conventional MRI and ¹H-MRS, or DWI, or DSC; (III) all patients were followed up for more than 6 months by MRI. The exclusion criteria were as follows: (I) history of other central nervous system diseases such as tumors and leukoencephalopathy, or history of brain surgery; (II) contraindications of MR examination such as cardiac pacemakers and aneurysm clips; (III) having claustrophobia or mental disorders undermining the guarantee of good compliance.

ICI patients were included as a control group. The diagnosis of ICI was confirmed based on their clinical manifestations and medical imaging data (MRI plus angiography). The inclusion criteria were as follows: (I) patients underwent conventional MRI and ¹H-MRS, or DWI, or DSC after onset; (II) patients were followed up for more than 6 months by MRI. The exclusion criteria were consistent with those for MELAS patients.

According to the time interval between clinical onset and MRI scans, MELAS and ICI patients were divided into an acute group (≤14 days after onset) and a chronic group (14–30 days after onset) respectively (25). Patient information and MRI sequences are summarized in Table 1.

MRI, ¹H-MRS, DSC, and DWI acquisition

The MRI data of patients were collected on a 3.0 T scanner (Skyra, Siemens Healthcare, Erlangen, Germany). Patients underwent conventional MRI, including T₁-weighted imaging (T₁WI): [repetition time (TR) =500 ms, echo time (TE) =12 ms, slice thickness =6 mm, slice gaP=1.0 mm, field of view (FOV) =20 cm × 20 cm, matrix =256×256], T₂-weighted imaging (T₂WI): (TR =5,000 ms, TE =102 ms, slice thickness =6 mm, slice gaP =1.0 mm), T₂-fluid-attenuated inversion recovery (T₂-FLAIR): [TR =10,000 ms, TE =104 ms, inversion time (TI) =2,500 ms, slice thickness =6 mm, slice gaP =1.0 mm], and gadolinium (Gd-DTPA)-enhanced T₁WI (contrast agent usage: 0.1 mmol/kg body weight). Multi-voxel ¹H-MRS was performed using chemical shift imaging (CSI) with the following parameters: TR/TE =2,000 ms/135 ms, FOV =16 cm × 16 cm, matrix =16×16, number of excitation (NEX) =1, scanning time of 6 minutes and 34 seconds. DSC was performed using perfusion-weighted echo planar imaging (EPI), TR/TE =1,640/30 ms, flip angle =90°, and matrix =128×128. DWI was performed using diffusion-weighted EPI, axial sections =42, TR/TE =5,300/102 ms, matrix =192×192, and 2 weighted diffusion acquisitions (b=0/1,000 sec/mm²).

Data postprocessing and analysis

In this study, MVL was defined as abnormal signal on conventional MRI, including T₂WI and/or T₂-FLAIR, and NAR was located beyond the MVL at least 1 MRS-voxel away from the MVL to avoid contamination from MVL. For all region of interest (ROI) measurements, cystic degeneration, necrosis, ventricles, or hemorrhage were excluded depending on T₂-weighted, T₁-weighted, and FLAIR images.

Raw spectral data were exported and processed with the Linear Combination Model (LCModel, version 6.3-1H, Provencher SW, http://s-provencher.com/lcmodel.shtml), a widely used software tool for automatic quantification of in vivo proton MR spectra without parameter tuning. Additional inclusion criteria for MRS data comprised signal/noise (S/N) ≥3, full width at half maximum (FWHM) ≤0.096 ppm, and percentage standard deviation (%SD) <20%.

The data of DSC and DWI were transferred to a dedicated workstation (Syngo Workplace, Siemens Healthcare) and analyzed. Cerebral blood volume (CBV) and cerebral blood flow (CBF) values were obtained by placing ROIs (60–80 mm²) on regions with abnormal or normal signal intensity on FLAIR and T₂WI. Furthermore, ROIs contaminated by ischemic penumbra were excluded depending on DWI and PWI (the ischemic penumbra was defined by DWI-PWI mismatch, i.e., abnormal perfusion area was 20% greater than diffusion-restricted area) (26). FLAIR images overlaid on CBV/CBF maps were used to ensure the correct position. For normalization, ROIs of the same size were placed in the contralateral mirror brain area, excluding ventricles. Finally, the relative CBV/CBF (rCBV/rCBF) was calculated by dividing the CBV of MVL or NAR to that of the contralateral mirror brain area. Measurements of ADC were obtained from the darkest areas with an ROI size ranging from 60 mm² to 80 mm², which were similar to perfusion evaluation. After measuring the ROIs 3 times, the average value of the 3 measurements was taken. Additionally, the ROIs of ADC and PWI should be consistent with those of MRS as much as possible. Lac peak, NAA/(creatine + phosphocreatine) (Cr) and Lac/Cr ratios were analyzed within the ROI. The interpretation of MR images and selections of ROIs were performed by X.Y. (more than 10 years of experience in neuroradiology) and supervised by Q.Y. (more than 20 years of experience in neuroradiology).

Statistics

Statistical analyses were performed using R software package (R Foundation for Statistical Computing, Vienna, Austria), and a P value less than 0.05 was accepted as the criterion of significance. Data distribution type and homogeneity of variance were analyzed first using Shapiro-Wilk test. For categorical variables, chi-square statistic was chosen. Paired t-test was used to compare the metabolic difference between MVL and NAR in MELAS or ICI. Mann-Whitney U test or independent sample t-test was used in the comparison of MELAS and ICI, as well as between acute and chronic stages in patients with MELAS or ICI. The area under the curve (AUC) from receiver operating characteristic (ROC) analysis was used to evaluate the differential diagnostic performance of MRI parameters between MELAS and ICI groups. Optimal cut-off value was determined by Youden index.

Results

Demographic data and clinical features

A total of 24 patients with MELAS were enrolled, including 18 males and 6 females, with an median age of 29 years (range: 11–68 years). MRI data was acquired in 2–28 days (median: 12 days) after onset. During their first admission to hospital, 8 cases were misdiagnosed as ICI, 6 were misdiagnosed as virus encephalitis, and 1 was misdiagnosed as low-grade glioma. A total of 25 patients with ICI (aged from 32 to 78 years, median age of 55 years) including 17 males and 8 females were also enrolled. MR examination was performed in 2–30 days (median: 15 days) after the onset of stroke. The average age of ICI group was significantly older than that of MELAS (P<0.001), as summarized in Table 1.

The main clinical features of patients were as follows: (I) MELAS: recurrent stroke-like seizures and epilepsy, nausea and vomiting, dizziness, headache, hemiparesis, and decreased exercise tolerance, cortical blindness, or hearing impairment, and similar presentations. (II) ICI: disturbance of consciousness, convulsion, vomiting, severe headache, dizziness, facial pain, numbness or oblique angle, impaired vision, and weakness or numbness of limbs, and so on. The similar clinical features were epilepsy, nausea, vomiting, dizziness, and headache. However, the clinical manifestations of MELAS are similar to those of ICI during initial attack.

MVL versus NAR: the metabolism, perfusion, and diffusion difference in MELAS or ICI

MELAS: MVL versus NAR

In MELAS patients, the Lac peak could be observed in MVL (23/24 patients, 95.8%) as well as in the NAR (18/22 patients, 81.8%) (Figure 1). The metabolic, perfusion, and ADC profiles of the MVL and NAR in MELAS are shown in Figure 2A. In two MELAS patients, all voxels with desirable spectra were located within the MVL, and NAR was obtained in 22 patients. Compared with MVL, NAR showed significantly lower Lac/Cr (P=0.005) and rADC (P=0.000), as well as significantly higher NAA/Cr (P=0.002). However, there was no significant difference in rCBF (P=0.959) and rCBV (P=0.605).

Figure 1 A 32-year-old male patient was diagnosed with MELAS. (A) Intensity map of Lac/Cr overlaid with T₂WI. The Lac/Cr intensity decreases gradually from the central region of the MRI-visible lesion to its peripheral region (A). Axial T₂WI (B) and DWI (C, b value =1,000) reveal a hyperintense lesion in the left occipital gray matter. The axial perfusion CBF (D) and CBV (E) of the left occipital lobe were slightly higher than those of the right occipital lobe. The yellow ellipse indicates the normal-appearing region, and the red ellipse denotes the MRI-visible lesion. A peak of Lac at 1.33 ppm appears both in the normal-appearing region (H, Lac is indicated by the yellow arrow) [location indicated by the voxel of yellow dotted line in (A), percentage standard deviation (%SD) =6%] and in the MRI-visible lesion (G, lactate is indicated by the red arrow) [location indicated by the voxel of red dotted line in (A), %SD =5%]. Relatively normal region dose not identify abnormal Lac peak [F and I, location indicated by the voxel of white and black dotted line in (A), respectively]. CBF, cerebral blood flow; CBV, cerebral blood volume; DWI, diffusion-weighted images; Lac, lactate; Lac/Cr, lactate/creatine; MELAS, mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; MRI, magnetic resonance imaging; NAA, N-acetylaspartate; T₂WI, T₂-weighted imaging.

Figure 2 Differences between MVL and NAR in metabolism (Lac/Cr, NAA/Cr), perfusion (CBF and CBV) and diffusion (ADC) changes of MELAS and ICI, including MVL versus NAR in MELAS (A) and ICI (B). **, P<0.01; ***, P<0.001. ADC, apparent diffusion coefficient; CBF, cerebral blood flow; CBV, cerebral blood volume; ICI, ischemic cerebral infarction; Lac/Cr, lactate/creatine; MELAS, mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; MRI, magnetic resonance imaging; MVL, MRI-visible lesions; NAA/Cr, N-acetylaspartate/Cr; NAR, normal-appearing regions.

ICI: MVL versus NAR

The metabolic, perfusion, and ADC profiles of the MVL and NAR in ICI are shown in Figure 2B. Lac peaks (14/25 patients, 56%) were observed in MVL, whereas no abnormal Lac peaks were observed in NAR (see Figure 3). Compared with MVL, NAA/Cr ratio and rCBF and rCBV of NAR were significantly higher (P<0.001, P=0.004 and P=0.003, respectively). The rADC values of NAR were slightly lower than those of MVL, but this difference was not statistically significant (P=0.687).

Figure 3 A 55-year-old male patient was diagnosed with ICI. (A) Intensity map of Lac/Cr overlaid on T₂-FLAIR. Axial FLAIR (B) and DWI (C, b value =1,000) reveal a hyperintense lesion in the right parietal lobe. Lac [percentage standard deviation (%SD%) =9%] appears in the MRI-visible lesion (D, the voxel is indicated by the red dotted line in A), but not in the normal-appearing region of the right frontal lobe (E, indicated by the white dotted line in A). DWI, diffusion-weighted images; ICI, ischemic cerebral infarction; Lac/Cr, lactate/creatine; MRI, magnetic resonance imaging; T₂-FLAIR, T₂-fluid attenuated inversion recovery.

MELAS versus ICI: the metabolism, perfusion, and diffusion difference in MVL or NAR

MVL: MELAS versus ICI

In the MVL of MELAS and ICI, the MRI parameters of MELAS and ICI were different, as shown in Figure 4A. Specifically, Lac/Cr, rCBF, and rCBV in MELAS were significantly higher than those in ICI (P=0.022, P=0.003, and P=0.001, respectively), with the Lac peak being more frequently observed (23/24 vs. 14/25, P=0.000) in MELAS. Meanwhile, NAA/Cr was not significantly different (P=0.443). rADC value was slightly higher in MELAS than in ICI, but the difference was not significant level (P=0.100).

Figure 4 Differences between MVL and NAR in metabolism (Lac/Cr, NAA/Cr), perfusion (CBF and CBV) and diffusion (ADC) changes of MELAS and ICI, including MELAS versus ICI in MVL (A) and NAR (B). *, P<0.05; **, P<0.01; ***, P<0.001. ADC, apparent diffusion coefficient; CBF, cerebral blood flow; CBV, cerebral blood volume; ICI, ischemic cerebral infarction; Lac/Cr, lactate/creatine; MELAS, mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; MRI, magnetic resonance imaging; MVL, MRI-visible lesions; NAA/Cr, N-acetylaspartate/Cr; NAR, normal-appearing regions.

NAR: MELAS versus ICI

Lac peaks were observed in the NAR of MELAS (18/22 patients, 81.8%), but not in any NAR of ICI.

NAA/Cr, rCBF, rCBV, and rADC showed no statistically significant difference between the 2 groups of patients (P>0.05 for each item, see Figure 4B).

Differential diagnostic performance of MRI parameters

As shown in Table S1, in the MVL, the Lac/Cr, rCBV, and rCBF had the best discriminating performance (AUC =0.829, 0.871, 0.886, respectively, see Figure 5A), but the recognition performances of NAA/Cr and rADC were poor. In the NAR region, the presence of Lac was highly indicative of MELAS, but NAA/Cr, rCBF, rCBV, and rADC had poor recognition performance (see Figure 5B).

Figure 5 Receiver operating characteristic curves of Lac/Cr, NAA/Cr, rCBF and rCBV, rADC of MVL (A) and NAR (B) in differentiating MELAS from ICI. AUC, area under the curve; ICI, ischemic cerebral infarction; Lac/Cr, lactate/creatine; MELAS, mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; MRI, magnetic resonance imaging; MVL, MRI-visible lesions; NAA/Cr, N-acetylaspartate/Cr; NAR, normal-appearing regions; rADC, relative apparent diffusion coefficient; rCBF, relative cerebral blood flow; rCBV, relative cerebral blood volume.

Acute versus chronic phase and MELAS versus ICI: the metabolism, perfusion, and diffusion difference

Acute versus chronic phase in MELAS or ICI

The difference of MRI parameters of MELAS and ICI in acute stage and chronic stage is shown in Figure 6. Compared with chronic MELAS (n=7), the acute-phase group (n=17) showed significantly higher values of rCBF (P=0.045) and rCBV (P=0.025) in the NAR. Meanwhile, there was no difference of Lac/Cr, NAA/Cr, or rADC in MVL or NAR (P>0.05, for each item). Compared with chronic ICI, the MVL of the acute group (n=10) showed lower rADC compared to the chronic group (n=15) (P=0.010). Meanwhile, there was no difference of Lac/Cr, NAA/Cr, rCBF, or rCBV in MVL or NAR (P>0.05, for each item).

Figure 6 Differences between acute stage and chronic stage MELAS (A) or ICI (B), and differences between MELAS and ICI in the acute stage (C), or chronic stage (D). *, P<0.05; **, P<0.01. ICI, ischemic cerebral infarction; Lac/Cr, lactate/creatine; MELAS, mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes; MRI, magnetic resonance imaging; MVL, MRI-visible lesions; NAA/Cr, N-acetylaspartate/Cr; NAR, normalappearing regions; rADC, relative apparent diffusion coefficient; rCBF, relative cerebral blood flow; rCBV, relative cerebral blood volume.

MELAS versus ICI in acute or chronic phases

In the acute stage, rCBF (P=0.036) and rADC (P=0.005) of MVL in MELAS were significantly higher than those in ICI (Figure 6C). In the chronic stage, rCBF (P=0.029) and rCBV (P=0.008) of MVL in MELAS were significantly higher than those in ICI (Figure 6D).

Discussion

We have shown that the onset age of MELAS patients was younger than that in ICI patients. Moreover, we used multiparametric MRI to investigate the metabolism, perfusion, and diffusion changes of MELAS and ICI, and their potential value in the differential diagnosis. For MELAS, this study revealed the following: (I) MRS showed that the Lac peak was observed not only in the MVL, but also in NAR. The Lac/Cr ratio of the MVL was significantly higher than that of NAR, whereas the NAA/Cr ratio was significantly lower; (II) the ADC value of MVL was significantly higher than that of NAR; (III) there was no difference in DSC parameters between MVL and NAR. Those characteristic changes of MELAS were different from ICI. Compared with ICI, MELAS demonstrated higher levels of Lac/Cr ratio, rCBF, and rCBV in the MVL, and unique presence of Lac in NAR. The presence of Lac in NAR, rCBV, and rCBF of MVL was also effective in discriminating MELAS from ICI (AUC of 0.929, 0.886, and 0.871, respectively).

The onset age of MELAS is broad. In this study, the age of onset for MELAS ranged from 11 to 68 years old, exhibiting an overlapping age distribution with the onset of ICI. Previous research demonstrated that MELAS can occur at any age, especially before the age of 40 years (2). A recent study showed that the onset of ICI was showing a trend towards younger ages (3), which may pose challenges for the differential diagnosis of MELAS. Our research revealed that the median age of MELAS was younger than that of ICI patients, which may provide some assistance in distinguishing MELAS from ICI. However, for distinguishing between relatively older MELAS patients and relatively younger ICI patients, our study demonstrated that multiparametric MRI may provide additional valuable information.

The simultaneous presence of Lac in both MVL and NAR verified our hypothesis that MELAS, as a metabolic encephalopathy, may involve the whole brain. The mitochondrial cytopathy caused by an oxidative phosphorylation defect in neurons and/or glial cells should account for the accumulation of Lac. However, only MVL demonstrated signal changes on conventional MRI. This may be explained by the hypothesis that MELAS may develop in two phases: a compensatory phase and a decompensated phase (27). During the compensatory phase, anaerobic metabolism occurs in NAR to make up for oxidative phosphorylation deficit. During this phase, elevated Lac peaks can be detected by MRS, but no abnormal signal can be observed on conventional MRI and DWI, because the damage to neurons may be mild. Throughout the progression of disease, anaerobic metabolism is insufficient to meet the physiological needs, and the accumulation of lactic acid becomes more severe and damages neurons and glial cells. These factors combined result in decompensation and cause irreversible cell impairment or even death. Therefore, during the second stage, the decrease of NAA peak, a valuable marker of neurons, may be observed by MRS (28), and the abnormal signal can be detected by conventional MRI and DWI. This speculation can explain our finding that Lac/Cr is higher in MVL than in NAR, whereas NAA/Cr is lower in MVL than in NAR. Likewise, the findings revealed by Abe et al. (23) were consistent with ours: Lac peak was observed in an area exhibiting normal signal intensity on DWI, and after 2 weeks abnormal signal emerged in the same area on DWI. These findings are of great clinical importance since in the compensatory phase, early clinical intervention may bring benefits to patients.

We did not find a significant difference in perfusion between MVL and NAR in MELAS. However, when disease stages were taken into consideration, we found that both the rCBF and CBV of NAR in the acute phase were significantly higher than those in the chronic phase. Previous reports about the perfusion changes of MELAS have been controversial in that hyperperfusion (8,18,29,30), hypoperfusion (18), or negative results (31) have all been reported. Based on our findings, we speculate that such discrepancy may derive from the different stages of disease. It is well known that lactic acidosis may decrease pH in the smooth muscle cells and endothelial cells of the vessels, leading to vasodilatation and hyperemia in acute stage MELAS (8,32). One study reported that lactate level in the acute stage of MELAS was positively correlated with focal rCBF in MELAS patients (8). However, when the disorder progresses to the chronic stage, substantial neural apoptosis in lesions may occur after recurrent episodes, and the need for energy may drop, and thus perfusion is expected to gradually decrease and approach normal levels or even drop below normal level. Wang et al. (8) further investigated the time-dependent changes in acute MELAS patients but did not find a significant difference in perfusion in MVL.

Our study revealed that the ADC value of MVL was higher than that of NAR in MELAS patients. This could be attributed to the relative period (median time 12 days) when MRI is performed since the onset of stroke-like episode. Studies have shown that MVL is a variable mixture of cytotoxic and vasogenic edema (15); therefore, DWI of MELAS may be variable depending on the phase. In the acute phase, DWI shows iso-intense or hyperintense signals, with unchanged or decreased ADC. During the chronic phase or after treatment, ADC may be increased and may return to normal. Recent findings suggest that the initial energy insufficiency causes cytotoxic edema and reduced ADC signal within the first 24 hours of a stroke-like episode. Subsequently, the development of vasogenic edema occurs and ADC increases within days to weeks (33).

Compared with ICI, MELAS demonstrated a higher level of Lac/Cr ratio, rCBF, and rCBV in the MVL, and the presence of lactate in NAR was only observed in MELAS. Similar results have also been reported in previous studies (6,16,30), but none of them had employed MRS, PWI, and DWI in a single study. The Lac peak is usually not observed in the normal population, especially among young individuals. Lac peaks can occasionally be observed in healthy elderly individuals (60–90 years), and the presence of these metabolites is limited to very focal areas (34). As noted above, Lac accumulation can lead to vasodilatation and hyperemia in MELAS. The abnormal metabolic change predominantly affects gray matter and spans arterial territories (35), dynamically progresses or regresses during the course of the episodes (36), and appears hyperperfused, which is a distinguishing difference from ischemia (16). Lac accumulation and hypoperfusion were also present in ICI patients in the subacute and/or chronic stage, but only observed in MVL. ICI could be ascribed to stenosis or arterial occlusion, which leads to perfusion deficiency of the responsible vascular brain region and subsequent cell dysfunction and death. Although Lac accumulation was frequently observed in ICI lesions, brain area outside the vascular territory was not supposed to exhibit any metabolic changes and abnormal vascular-related perfusion. The differences of Lac, rCBV, and rCBF may serve as useful markers for the differential diagnosis of MELAS and ICI since they demonstrate a good discriminative power. In addition, although we did not find difference in rADC between MELAS and ICI when disease stage was not considered, we did find that rADC of MELAS was significantly higher than that of ICI in the acute phase. The results of this study showed that multiparametric MRI was helpful for the diagnosis and differential diagnosis of MELAS and ICI. In recent years, there has been an increasing interest in developing methods for quantitative MRI, which provides information about structural differences in brain tissue and has been applied in multiple sclerosis (MS) (37). This research may open up new avenues for the study of MELAS. With the application of multiparametric MRI and quantitative MRI, patients with MELAS are expected to benefit significantly in terms of diagnosis, differential diagnosis, and treatment.

The major limitation of this study is that we only conducted a retrospective clinical analysis with a small sample size. This small size is due to a limited number of MELAS patients undergoing multiparametric MRI, which is not yet used as the routine clinical imaging procedure in patients suspected of ICI or MELAS. The findings of our study justify the application of multiparametric MRI when a young patient presents with stroke-like episodes. Another limitation of our study is that the patients were mainly in the late-acute (median examination time of MELAS: 12 days after onset) or early-chronic stage (median examination time of ICI: 15 days after onset), which may not accurately reflect the changes of each stage. Thus, a prospective study using multiparametric MRI is still needed to establish a full-scope view of MELAS and ICI, either at the acute or chronic stage.

Conclusions

In summary, the present study demonstrated that MELAS involves the whole brain; as a result, abnormal changes may occur not only in MVL but also in ‘normal-appearing’ brain regions. In contrast, ICI only affects the responsible vascular territory. The accumulation of Lac in the NAR of MELAS and the abnormal perfusion in the MVL may serve as useful biomarkers for the differential diagnosis. In the clinical setting, if a young person presents with stroke-like episode, multiparametric MRI including MRS, PWI, and DWI is highly recommended. If Lac is present in NAR with relatively high or intact perfusion, MELAS should be considered as the possible diagnosis.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the CLEAR reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-2396/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-24-2396/dss

Funding: This work was supported by the National Natural Science Foundation of China (grant No. 82471961), the fellowship of China Postdoctoral Science Foundation (grant No. 2022M712257), and the Sichuan Provincial Foundation of Science and Technology (grant Nos. 2022YFS0073 and 2023NSFSC0626).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-2396/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was approved by the Ethics Review Board of West China Hospital of Sichuan University (Ethical Approval No. 2020-92) and was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Informed consent was waived due to the minimal-risk nature of this retrospective analysis.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Yang X, Sun A, Ji K, Wang X, Yang X, Zhao X. Clinical features of epileptic seizures in patients with mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes. Seizure 2023;106:110-6. [Crossref] [PubMed]
2. Yatsuga S, Povalko N, Nishioka J, Katayama K, Kakimoto N, Matsuishi T, Kakuma T, Koga Y. MELAS: a nationwide prospective cohort study of 96 patients in Japan. Biochim Biophys Acta 2012;1820:619-24. [Crossref] [PubMed]
3. Béjot Y, Duloquin G, Thomas Q, Mohr S, Garnier L, Graber M, Giroud M. Temporal Trends in the Incidence of Ischemic Stroke in Young Adults: Dijon Stroke Registry. Neuroepidemiology 2021;55:239-44. [Crossref] [PubMed]
4. Kolb SJ, Costello F, Lee AG, White M, Wong S, Schwartz ED, Messé SR, Ellenbogen J, Kasner SE, Galetta SL. Distinguishing ischemic stroke from the stroke-like lesions of MELAS using apparent diffusion coefficient mapping. J Neurol Sci 2003;216:11-5. [Crossref] [PubMed]
5. Khandwala K, Ahmed A, Sheikh T. MELAS: A Complex and Challenging Diagnosis. J Coll Physicians Surg Pak 2018;28:S46-8. [Crossref] [PubMed]
6. Li X, Wang Y, Wang Z, Lu J, Xu Y, Ye J, Kuang Z, Li B, Pan M, Chen W, Lu S, Zhou Z, Cheng L, Wang H. Comparison of magnetic resonance spectroscopy (MRS) with arterial spin labeling (ASL) in the differentiation between mitochondrial encephalomyopathy, lactic Acidosis, plus stroke-like episodes (MELAS) and acute ischemic stroke (AIS). J Clin Neurosci 2018;55:65-70. [Crossref] [PubMed]
7. Pizzamiglio C, Bugiardini E, Macken WL, Woodward CE, Hanna MG, Pitceathly RDS. Mitochondrial Strokes: Diagnostic Challenges and Chameleons. Genes (Basel) 2021.
8. Wang R, Hu B, Sun C, Geng D, Lin J, Li Y. Metabolic abnormality in acute stroke-like lesion and its relationship with focal cerebral blood flow in patients with MELAS: Evidence from proton MR spectroscopy and arterial spin labeling. Mitochondrion 2021;59:276-82. [Crossref] [PubMed]
9. Niu FN, Meng HL, Chang LL, Wu HY, Li WP, Liu RY, Wang HT, Zhang B, Xu Y. Mitochondrial dysfunction and cerebral metabolic abnormalities in patients with mitochondrial encephalomyopathy subtypes: Evidence from proton MR spectroscopy and muscle biopsy. CNS Neurosci Ther 2017;23:686-97. [Crossref] [PubMed]
10. Mitani T, Aida N, Tomiyasu M, Wada T, Osaka H. Transient ischemic attack-like episodes without stroke-like lesions in MELAS. Pediatr Radiol 2013;43:1400-3. [Crossref] [PubMed]
11. Li Y, Lin J, Sun C, Zhao C, Li H. Increased cerebral blood flow as a predictor of episodes in MELAS using multimodal MRI. J Magn Reson Imaging 2017;46:915-8. [Crossref] [PubMed]
12. Malhotra K, Liebeskind DS. Imaging of MELAS. Curr Pain Headache Rep 2016;20:54. [Crossref] [PubMed]
13. Kamada K, Takeuchi F, Houkin K, Kitagawa M, Kuriki S, Ogata A, Tashiro K, Koyanagi I, Mitsumori K, Iwasaki Y. Reversible brain dysfunction in MELAS: MEG, and (1)H MRS analysis. J Neurol Neurosurg Psychiatry 2001;70:675-8. [Crossref] [PubMed]
14. Crain ID, Elias PS, Chapple K, Scheck AC, Karis JP, Preul MC. Improving the utility of (1)H-MRS for the differentiation of glioma recurrence from radiation necrosis. J Neurooncol 2017;133:97-105. [Crossref] [PubMed]
15. Casimiro C, Martins J, Nunes C, Parreira T, Batista S, Cordeiro M, Matias F, Rebelo O, Freitas P. Conventional and diffusion-weighted magnetic resonance imaging and proton spectroscopy in MELAS. Acta Med Port 2012;25:59-64.
16. Li R, Xiao HF, Lyu JH, Wang DJJ, Ma L, Lou X. Differential diagnosis of mitochondrial encephalopathy with lactic acidosis and stroke-like episodes (MELAS) and ischemic stroke using 3D pseudocontinuous arterial spin labeling. J Magn Reson Imaging 2017;45:199-206. [Crossref] [PubMed]
17. Yeh HL, Chen YK, Chen WH, Wang HC, Chiu HC, Lien LM, Wei YH. Perfusion status of the stroke-like lesion at the hyperacute stage in MELAS. Brain Dev 2013;35:158-64. [Crossref] [PubMed]
18. Wang Z, Xiao J, Xie S, Zhao D, Liu X, Zhang J, Yuan Y, Huang Y. MR evaluation of cerebral oxygen metabolism and blood flow in stroke-like episodes of MELAS. J Neurol Sci 2012;323:173-7. [Crossref] [PubMed]
19. Yoneda M, Maeda M, Kimura H, Fujii A, Katayama K, Kuriyama M. Vasogenic edema on MELAS: a serial study with diffusion-weighted MR imaging. Neurology 1999;53:2182-4. [Crossref] [PubMed]
20. Yonemura K, Hasegawa Y, Kimura K, Minematsu K, Yamaguchi T. Diffusion-weighted MR imaging in a case of mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes. AJNR Am J Neuroradiol 2001;22:269-72.
21. Yang X, Xu Y, He D, Yang J, Yue Q, Gong Q. MR imaging and MR spectroscopy features of MELAS and their clinical values. Radiologic Practice 2014;29:149-54.
22. Cai SS, von Coelln R, Kouo TJ. Migratory stroke-like lesions in a case of adult-onset mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes (MELAS) syndrome and a review of imaging findings. Radiol Case Rep 2016;11:425-9. [Crossref] [PubMed]
23. Abe K, Yoshimura H, Tanaka H, Fujita N, Hikita T, Sakoda S. Comparison of conventional and diffusion-weighted MRI and proton MR spectroscopy in patients with mitochondrial encephalomyopathy, lactic acidosis, and stroke-like events. Neuroradiology 2004;46:113-7. [Crossref] [PubMed]
24. Ikawa M, Yoneda M, Muramatsu T, Matsunaga A, Tsujikawa T, Yamamoto T, Kosaka N, Kinoshita K, Yamamura O, Hamano T, Nakamoto Y, Kimura H. Detection of preclinically latent hyperperfusion due to stroke-like episodes by arterial spin-labeling perfusion MRI in MELAS patients. Mitochondrion 2013;13:676-80. [Crossref] [PubMed]
25. Sun EX, Deng F, Mishra S, Huang RY. Neuroimaging: Brain. In: Sun EX, Shi J, Mandell JC eds. Core Radiology: A Visual Approach to Diagnostic Imaging. 2 ed. Cambridge: Cambridge University Press; 2021:650-752.
26. Ben Alaya I, Limam H, Kraiem T. Applications of artificial intelligence for DWI and PWI data processing in acute ischemic stroke: Current practices and future directions. Clin Imaging 2022;81:79-86. [Crossref] [PubMed]
27. Iizuka T, Sakai F. Pathogenesis of stroke-like episodes in MELAS: analysis of neurovascular cellular mechanisms. Curr Neurovasc Res 2005;2:29-45. [Crossref] [PubMed]
28. Petroff OA, Errante LD, Kim JH, Spencer DD. N-acetyl-aspartate, total creatine, and myo-inositol in the epileptogenic human hippocampus. Neurology 2003;60:1646-51. [Crossref] [PubMed]
29. Zhao M, Zuo C, Hao H, Xing X, Zhao L, Li N. A patient with MELAS syndrome combined with autoimmune abnormalities: a case report. Front Neurol 2023;14:1239664. [Crossref] [PubMed]
30. Li Y, Xu W, Sun C, Lin J, Qu J, Cao J, Li H, Yang L. Reversible Dilation of Cerebral Macrovascular Changes in MELAS Episodes. Clin Neuroradiol 2019;29:321-9. [Crossref] [PubMed]
31. Tsujikawa T, Yoneda M, Shimizu Y, Uematsu H, Toyooka M, Ikawa M, Kudo T, Okazawa H, Kuriyama M, Kimura H. Pathophysiologic evaluation of MELAS strokes by serially quantified MRS and CASL perfusion images. Brain Dev 2010;32:143-9. [Crossref] [PubMed]
32. Iizuka T, Sakai F, Suzuki N, Hata T, Tsukahara S, Fukuda M, Takiyama Y. Neuronal hyperexcitability in stroke-like episodes of MELAS syndrome. Neurology 2002;59:816-24. [Crossref] [PubMed]
33. Tzoulis C, Bindoff LA. Serial diffusion imaging in a case of mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes. Stroke 2009;40:e15-7. [Crossref] [PubMed]
34. Sijens PE, den Heijer T, de Leeuw FE, de Groot JC, Achten E, Heijboer RJ, Hofman A, Breteler MM, Oudkerk M. MR spectroscopy detection of lactate and lipid signals in the brains of healthy elderly people. Eur Radiol 2001;11:1495-501. [Crossref] [PubMed]
35. El-Hattab AW, Adesina AM, Jones J, Scaglia F. MELAS syndrome: Clinical manifestations, pathogenesis, and treatment options. Mol Genet Metab 2015;116:4-12. [Crossref] [PubMed]
36. Tzoulis C, Bindoff LA. Acute mitochondrial encephalopathy reflects neuronal energy failure irrespective of which genome the genetic defect affects. Brain 2012;135:3627-34. [Crossref] [PubMed]
37. Engström M, Warntjes JB, Tisell A, Landtblom AM, Lundberg P. Multi-parametric representation of voxel-based quantitative magnetic resonance imaging. PLoS One 2014;9:e111688. [Crossref] [PubMed]