Abstract
Objective: To evaluate the anatomic and clinical relationship between the lenticulostriate artery (LSA) territory and the corticospinal tract (CST) in patients with acute infarcts in this territory using MR tractography.
Methods: Thirteen consecutive patients who presented with acute infarcts in the LSA territory and who also had undergone an MRI study within 3 days after symptom onset were studied. Visualization of the CST was achieved by postprocessing the acquired diffusion tensor imaging data. To classify lesion location, the LSA territory was divided into four subsegments, the boundaries of which were drawn by axial and coronal planes crossing through the foramen of Monro. Infarct volume and extent of CST involvement were measured and compared with neurologic findings.
Results: All of the infarcts were located in the posterior segment. All of the depicted CSTs crossed the LSA territory only at the posterosuperior quadrant. The extent of CST involvement within the infarcts was correlated with the severity of the patient’s motor deficit (p < 0.01) and with the clinical outcome (p < 0.05).
Conclusions: The corticospinal tracts (CSTs) crossed the lenticulostriate artery territory exclusively at the posterosuperior quadrant, and the degree of CST involvement within the infarcts was directly related to stroke severity and functional recovery.
Patients with infarctions involving the area supplied by the lenticulostriate arteries (LSAs) often present with severe motor disturbances, indicating that the motor tract is in close proximity to infarcts located in the LSA territory.1–3⇓⇓ The LSAs are perforating branches that arise from the horizontal portion of the middle cerebral artery (MCA). Studies have indicated that the areas supplied by the LSAs include the body of the caudate nucleus as well as the superior part of its head, the putamen, the lateral segment of the globus pallidus, and the superior part of the anterior and posterior limb of the internal capsule.4–6⇓⇓ The anatomic location of the corticospinal tract (CST) has also been examined in detail using autopsy specimens as well as conventional T2-weighted MRI.7,8⇓ The majority of the fibers contained in the CST arise from the precentral cortex, descend in the corona radiata, and pass through the posterior limb of the internal capsule, cerebral peduncle, basilar pons, and medullary pyramid.
Recent advances in fast MRI techniques have enabled the observation of the functional aspects of brain tissue. Diffusion tensor MRI (DTI), a result of recent advances in MRI,9,10⇓ allows the visualization of the anisotropy of the water movement caused by the presence of axons, axonal sheaths, glial cells, and vasculature.11 Using such data acquired through DTI, we are now able to reconstruct a three-dimensional macroscopic orientation of the brain fibers. This technique is known as tractography, or fiber tracking, and it is currently the only way to observe the neuronal pathways in the living human brain.
In this study, we hypothesized that the three-dimensional relationship between the CST and the infarcts involving the LSA territory could be depicted using tractography and that this information might enable further correlation between the lesion location and the degree of motor symptoms.
Materials and methods.
Patient population.
Patients seen between November 2002 and February 2004 were entered into the study based on the following inclusion criteria: admission to our university hospital with acute infarcts involving the LSA territory, no involvement of any other brain areas, and evaluation by MRI within 3 days of symptom onset. There were 11 men and 2 women who met these inclusion criteria; their age ranged from 28 to 87 years (mean 65 years). The National Institute of Neurologic Disorders and Stroke classification was used for classifying presumed stroke etiology (atherothrombotic, cardioembolic, lacunar, and others). Potential embolic sources were checked by 12-lead EKG, echocardiography, and MR angiography. The clinical deficit was assessed at the time of MRI using the NIH Stroke Scale (NIHSS). The clinical outcome was measured at 3 months using the modified Rankin Scale (mRS). The NIHSS and mRS scores were determined by a neurologist who was blinded to MR images at onset. Written informed consent was obtained from all patients or their next of kin. This study was approved by an institutional review board.
MRI.
Images were obtained with a whole-body 1.5 T MR system (Gyroscan Intera; Philips Medical Systems, Best, the Netherlands) and a six-channel phased-array head coil. The routine stroke protocol at our institute consists of T1-weighted images (repetition time [TR] = 611 milliseconds, echo time [TE] = 13 milliseconds), T2-weighted images (TR = 4,754 milliseconds, TE = 100 milliseconds), fluid level–attenuated inversion recovery images (delay time T1 = 2,200 milliseconds, TR = 8,000 milliseconds, TE = 100 milliseconds), T2*-weighted images (TR = 666 milliseconds, TE = 23 milliseconds), MR angiogram (TR = 30 milliseconds, TE = 2.3 milliseconds), and DTI. DTI was performed using a single-shot echo-planar technique (TR = 6,000 milliseconds, TE = 88 milliseconds, flip angle = 90°). DTI data were obtained with a spin echo Stejskal–Tanner sequence12 with six motion-probing gradient orientations. A b value of 800 s/mm2 was used with an average of six images. The 128 × 53 data points were recorded using the parallel imaging technique.13 This allows the image to be reconstructed in half as many encoding steps and thus reduces the unique geometric image distortion of echo-planar imaging. Therefore, the true resolution of the images was equivalent to 128 × 106 pixels. Thirty-six sections were obtained with a thickness of 3 mm, without intersection gaps. The field of view was 230 mm. The total DTI scan time for tractography was 4 minutes 24 seconds. Diffusion-weighted images for interpretation and apparent diffusion coefficient maps were calculated from the DTI protocol mentioned above.
Data processing.
DTI data were transferred to an off-line workstation for analysis. Data were analyzed with Philips Research Integrated Development Environment (PRIDE) software written in Interactive Data Language. Diffusion tensor elements were determined by means of multivariate least-squares fitting weighted by the signal-to-noise ratio.14 The anisotropy maps were obtained by means of orientation-independent fractional anisotropy (FA).15 The translation of the vectors into neuronal trajectories was achieved by a technique known as the Fiber Assignment by Continuous Tracking method.10,16⇓ To map the neural connections, three arbitrary regions of interest (ROIs) in the three-dimensional space on PRIDE software were designated. Tracking was terminated (stop criterion) when it reached a pixel with low FA (<0.3) or a predetermined trajectory curvature between two contiguous vectors (inner product < 0.85). Fiber tracts that passed through ROIs (such as the basilar pons, cerebral peduncle, and precentral gyrus) were designated as the final tract of interest, namely, the CST (figure 1). All ROIs were determined by one of the authors, and validation of the depicted CST was obtained by consensus with reference to atlases of neuroanatomy.7,8⇓
Figure 1. Visualization of the corticospinal tract (CST). The first region of interest (ROI-1) was placed at the superior level of the basilar pons. The second ROI (ROI-2) was placed at the ipsilateral cerebral peduncle. Tractography depicted from these ROIs often contained tracts to or from areas other than the primary motor cortex. Therefore, to further specify the location of CST, an additional ROI was placed on the precentral gyrus (ROI-3), identified based on the sulcal pattern at the vertex. Fiber tracts that passed through these ROIs were designated as the CST. A typical appearance of bilateral CSTs (Patient 5) was demonstrated in this figure by superimposing these tracts on the representative slices of axial diffusion-weighted images.
Data analysis.
Subdivisions of LSA territory.
To classify the lesion location, we divided the LSA territory into the anterior and the posterior segments, according to previous research.4–6⇓⇓ The coronal plane that crosses through the foramen of Monro was selected as the boundary between the anterior and posterior segments (figure 2). The anterior segment included the superior part of the head of the caudate nucleus, the anterior limb of the internal capsule, and the anterior putamen. The posterior segment included the body of the caudate nucleus, the posterior putamen, and the superior part of the posterior limb of the internal capsule.
Figure 2. Segmentation of the lenticulostriate artery (LSA) territory. The LSA territory is divided into four subsegments by the coronal and axial planes that cross through the foramen of Monro, and these subsegments are color-coded as follows: anterosuperior (orange), posterosuperior (blue), anteroinferior (green), and posteroinferior (yellow).
To further specify the lesion’s location, each of the two segments was further subdivided into two subsegments by the axial plane that crosses through the foramen of Monro. Thus, the LSA territory was finally divided into four subsegments: the anterosuperior, the anteroinferior, the posterosuperior, and the posteroinferior subsegments (see figure 2).
Assessing anatomic location of CST.
First, we made a side-to-side comparison of the depicted CSTs on tractography so as to assess their symmetry. Tractography on the contralesional side was considered to be the gold standard for the tract’s location. Second, the location of the depicted CSTs in relation to the four subdivisions of the LSA territory was assessed. This evaluation was made by two neuroradiologists.
Correlation between LSA infarct and clinical features.
First, we evaluated the correlation between the volume of the LSA infarct and the clinical features (NIHSS and 3-month mRS scores). Infarct volumes were determined off-line after the images were transferred to a workstation for analysis. Two observers manually outlined the hyperintense area on diffusion-weighted images in each slice and determined the volume by multiplying the area by the slice thickness. The results of the two observers were averaged.
Second, the correlation between the extent of CST involvement within the infarcts and the severity of motor deficits was analyzed. To evaluate the extent of CST involvement, we devised an “Involvement Scale” (IS). The extent of CST involvement on an axial plane was classified on a scale from 0 to 2: 0 for no involvement, 1 for partial involvement, and 2 for complete involvement (infarcts engulfing the depicted CST) (figure 3). The IS score of each patient was determined by a sum of the scores for each slice, taking the average of measurements performed by two of the authors who were blinded to the clinical information.
Figure 3. Involvement Scale (IS) score. The extent of corticospinal tract (CST) involvement by the infarct was evaluated by using the IS. The presented representative case is similar to the one shown in figure 1 (Patient 5). On an axial diffusion-weighted image, IS was classified on a scale from 0 to 2. The IS score was determined by a sum of the scores for each slice.
Statistical analysis.
The degree of interrater agreement for each slice on the IS was determined by calculation of the κ statistic. The correlations between the NIHSS and 3-month mRS scores with the infarct volume and between the NIHSS and 3-month mRS scores with the IS scores were analyzed by Spearman rank correlation coefficients. Significance was determined at the level of p < 0.05.
Results.
All of the acute infarcts in this study were located in the posterior segments of the LSA territory, and there were no patients that exhibited a lesion in the anterior segment (figure 4).
Figure 4. Lesion location in the lenticulostriate artery territory. Acute infarcts (white) of all patients on diffusion-weighted images are drawn on the schematic sagittal plane.
Anatomic relationship between CST and LSA territory.
Tractography was able to depict the CST pathways in all patients. The CST pathways were noted to lie in a symmetric manner with respect to the contralesional side. There were no patients in whom the CST appeared to be disrupted or to deviate from its normal location. All the CST pathways originating from the precentral gyrus descended through the centrum semiovale and reached the lateral aspect of the LSA territory (see figure 1). They then crossed the territory only at the posterosuperior segment. Below this area, the CST did not have direct contact with the LSA territory.
LSA infarct related to clinical features.
All the CSTs depicted on tractography were involved at least in part by the infarct. All of our patients experienced motor disturbances. A representative case is illustrated in figure 5. The infarct volume and IS score of each patient are shown in the table. The strength of interrater agreement for the IS score was considered to be very good (κ = 0.834, 95% CI 0.694 to 0.973). The NIHSS score and 3-month mRS scores were not significantly correlated with the infarct volume (r = 0.371, p = 0.199 for NIHSS score; r = 0.336, p = 0.244 for 3-month mRS score). There was, however, a significant correlation between the NIHSS score and the IS score (r = 0.758, p = 0.009). Furthermore, the 3-month mRS score and the IS score also had a significant correlation (r = 0.576, p = 0.046).
Figure 5. Diffusion tensor imaging findings in Patient 3. A 28-year-old man presented with abrupt onset of left arm weakness, which rapidly progressed to left hemiplegia over the next 24 hours. Diffusion-weighted images obtained 2 days after symptomatic onset are shown in this figure with superimposed corticospinal tracts (CSTs) indicated by the purple cables. Transparent yellow planes through the foramen of Monro divide each segment. The infarct is located in the posterior segment of the lenticulostriate artery (LSA) territory, and the CST is shown to cross the LSA territory in the posterosuperior subsegment, where the infarct engulfs the CST. The measured fractional anisotropy of this infarct was 0.5064 ± 0.1571 (mean ± SD), which indicates that there were pixels within this infarct that exceeded the stop criterion used for fiber tracking.
Table Demographic data, lesion location, and clinical scales
Discussion.
This study had two major findings. First, the CST was shown to cross the LSA territory only at the posterosuperior subsegment (figure 6). Second, the degree of CST involvement within this subsegment was shown to have significant correlation with stroke severity measured by NIHSS and the outcome measured by mRS.
Figure 6. A schematic representation of the corticospinal tract (CST) and the lenticulostriate artery (LSA) territory. The CSTs are shown as the purple cables. The LSA territory is subdivided into four subsegments, and the CSTs cross this territory only at the posterosuperior segment (blue). (Top) Bilateral CSTs and the LSA territory shown in left anterior oblique (left) and left posterior oblique (right) views; (bottom) right CST and the LSA territory shown in anterior (left), inside (middle), and posterior (right) views.
The anatomic location of the CST relative to the LSA territory is an important clinical issue. Many previous anatomic reports have pointed out the close proximity of the CST to the LSA territory.1–3,7,8⇓⇓⇓⇓ Our study not only confirms these previous studies but also allows us to further assert that the CST crosses the LSA territory only through the posterosuperior quadrant. As the CST courses caudally, it passes through the LSA territory for a very short distance and quickly exits this territory to enter the area supplied by the anterior choroidal artery.6,17⇓ This spatial relationship of the LSA territory and the CST has not been well documented in vivo, and this study attempted to demonstrate this relationship using tractography.
We divided the LSA territory into anterior and posterior segments by identifying the foramen of Monro on coronal section. This method was based on previous investigations of the perforating branches of the MCA.3–6,18⇓⇓⇓⇓ This grouping of the perforators into anterior and posterior units is relevant as they often form a group of vessels that have similar morphologic trends. In the extracerebral (or cisternal) segment of the LSA, the anterior branches arise from the proximal portion of the MCA, whereas the posterior branches arise from the distal portion and make a steep turn before entering the anterior perforating substance.4,5,18,19⇓⇓⇓ Just after passing through the anterior perforating substance, the anterior branches of the LSA course toward the anterosuperior aspect, and the posterior group of branches ascends posterosuperiorly.4–6⇓⇓ There is a study that carefully investigated the angiographic and CT findings in patients with the LSA infarcts. In this report, the authors found that though infarcts can occur at the anterior, posterior, or both segments of the LSA territory, the posterior segment is the most commonly affected site.6 More recently, a study assessed infarcts in the LSA territory according to three subdivisions of the LSA territory (anterior, posterior, and medial) and reported that the clinical features of infarcts in the posterior and medial parts, which closely corresponded with our posterior segment, were significantly associated with symptoms suggesting the CST involvement.3
As we have shown, the posterosuperior quadrant of the LSA territory is the vital area where the motor tract crosses, and an infarct at this area may lead to significant motor symptoms. A question still remains as to whether an infarct involving other quadrants of the LSA territory would have less significant or no motor symptoms. It has been previously reported that some patients with infarcts involving the anterior LSA territory also have motor symptoms such as weakness and clumsiness.1,20,21⇓⇓ Motor symptoms in these patients may be the result of a disturbance of the circuit between the caudate nucleus and the motor cortex as well as the premotor cortex and the supplementary motor area or from a lesion in the anterior limb of the internal capsule that carries frontopontine motor fibers. This type of motor deficiency has been termed “nonpyramidal hemimotor syndrome”20 to distinguish it from pyramidal pathway lesions in which Babinski signs, hyperreflexia, paralysis, and “pyramidal distribution” weakness and altered tone are found.
There are two possible explanations for why LSA infarcts occur in the posterior segment but not in the anterior segment in so many patients. The first reason may be that atherosclerotic changes may be more severe in the posterior perforators, because they make a steep turn before they enter the anterior perforating substance,4,5,18⇓⇓ and this could predispose these small arteries to more significant sheer stress, resulting in more atheromatous change. The second reason may be somewhat speculative, but we presume that patients with an infarct in the posterior segment have a stronger motivation to visit the emergency room because of their more profound motor symptoms. The veracity of this hypothesis would need to be proven by analyzing patients with infarcts involving the anterior segment. Unfortunately, there were no such cases in our group of patients.
In the current study, the degree of CST involvement by the infarcts measured by our IS was highly correlated with the severity of the motor deficit and the clinical outcome. This IS takes into consideration of the degree of involvement at each plane and takes the sum of them through slices. This IS would be difficult to assess without tractography, which may imply the clinical impact of using this technique. It may be also important to note that infarct volume did not reach significance with clinical severity in our study, which may be due to the small number of our patient population as well as a rather homogeneous group of subjects with small infarcts limited to the LSA territory.
In all of our patients, the CST depicted by tractography passed through the infarcts without disruption. Although there is general trend toward reduction in FA within lesions at subacute and chronic stages of infarction, a slight but substantial increase in FA has been found to occur in a majority of the lesions immediately after the onset of ischemia. This increase in FA may have enabled the estimation of CST using tractography technique.22,23⇓ Even for those lesions with decreased FA, they seldom have values below 0.3, which is still higher than our regular stop criterion used for the fiber-tracking technique.
This study has several limitations. First, our study population was small; therefore, variations of the extent of the LSA territory could not be considered. It is well known that the major vascular territories have significant variation, and therefore it can be expected that there would be at least some variation in the extent of the LSA territory. Second, because of the small number of patients, a subgroup analysis of the patients by mechanism of infarct, such as atherothrombotic, cardioembolic, lacunar, and others, was not performed. All of our patients were Asians, in whom it is known that the prevalence of atheromatous disease is higher than embolic etiology.24 Therefore, whether our data apply to other populations may need further verification. Third, there is a technical limitation that tractography is capable of depicting only part of the tract. This is due mainly to a phenomenon known as the “crossing-fiber” problem. A detailed description of this issue is beyond the scope of this article and can be found elsewhere.25,26⇓ It is, however, expected to be solved by means of more sophisticated imaging techniques.25,26⇓ Further large-scale studies are necessary to validate our assumption that infarcts involving the posterior segment tend to have poorer outcomes.
- Received May 27, 2004.
- Accepted September 9, 2004.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan