Objective To characterise the topographic patterns of thalamic infarcts associated with various stroke syndromes and aetiologies.
Methods In this study, 168 consecutive patients with acute infarcts involving the thalamus were investigated by use of diffusion weighted MR imaging. Involved thalamic territories were classified into four vascular territories: anterior (polar artery), posteromedial (thalamoperforating artery), ventrolateral (thalamogeniculate artery) and posterolateral (posterior choroidal artery) territory. The distribution of thalamic infarcts involving specific vascular territories in association with various stroke syndromes and aetiologies were analysed.
Results There was a significant association between involvement of the ventrolateral thalamus and isolated thalamic infarcts (49 patients (73%), p<0.01) or posterior cerebral artery infarcts (29 patients (76%), p=0.02), and between involvement of the posteromedial thalamus and top of the basilar artery syndrome (17 patients (77%), p<0.001) or extended posterior circulation infarcts (29 patients (71%), p<0.001). The ventrolateral territory was most commonly affected in association with small vessel disease (43 patients (72%), p=0.03) and the posteromedial territory in association with large artery disease (19 patients (63%), p<0.01). In cardioembolic stroke, the ventrolateral, posteromedial and posterolateral territories were equally affected.
Conclusions The results of the present study indicate a specific association between the topographic patterns of thalamic infarcts and the stroke syndromes and aetiologies.
Statistics from Altmetric.com
Thalamic infarcts can be divided into four groups based on the territories supplied by the four main arteries: polar, thalamoperforating, thalamogeniculate and posterior choroidal arteteries.1–3
The thalamus is affected in various stroke syndromes, including isolated thalamic infarct, posterior cerebral artery (PCA) territory infarct, top of the basilar artery (TOB) syndrome and extended posterior circulation (EPC) infarct. However, the patterns of thalamic infarcts involving specific vascular territories in various stroke syndromes and aetiologies have been poorly defined. A few studies addressing these issues were limited to small number of patients.3–5
Diffusion weighted MR imaging (DWI) provides more accurate detection of very small or multiple acute ischaemic lesions than other brain imaging modalities.6–8 In addition, DWI enables the differentiation between acute and chronic lesions.6 Thus the lesion pattern and underlying mechanism of acute ischaemic stroke can be more precisely estimated with DWI.
The aim of this study was to characterise the topographic patterns of thalamic infarcts in various stroke syndromes and aetiologies by using DWI in patients with acute ischaemic stroke.
A total of 168 consecutive patients with acute infarcts involving the thalamus who were admitted to the Neurology Department of the Dankook University Hospital between March 2005 and April 2010 were retrospectively studied. The medical records, stroke registry and MR images were reviewed. Patients were included if they had acute ischaemic lesions involving the thalamus alone or with concomitant lesions outside the thalamus on DWI within 1 week of stroke onset.
All patients underwent MRI, MR angiography, transcranial Doppler, transthoracic echocardiograpy, 12 lead electrocardiography and standard blood tests. In selected patients, transoesophageal echocardiography, 24 h electrocardiography (Holter) monitoring and conventional angiography were performed as required.
MRI consisted of axial T1 weighted (1.5 T; repetition time 417 ms; echo time 10 ms) and T2 weighted images (repetition time 5000 ms; echo time 80 ms) with 5 mm sections. DWI was performed using a single shot echo planar, spin echo pulse sequence (1.5 T; repetition time 6000 ms; echo time 84 ms; b=1000 s/mm2; field of view 260 mm; matrix 128×128; thickness 5 mm; 20 slices). Acute lesions were differentiated from chronic lesions using DWI. DWI and T2 weighted images were used to delineate the distribution of acute infarcts.
Thalamic territories were classified according to their arterial supply based on the previously reported template9: (1) anterior territory, supplied by the polar or tuberothalamic artery; (2) posteromedial territory, supplied by the thalamoperforating artery; (3) ventrolateral territory, supplied by the thalamogeniculate artery; and (4) posterolateral territory, supplied by the posterior choroidal artery (figure 1).
Stroke syndromes were categorised according to the lesion topography as follows: (1) isolated thalamic infarct was defined as an infarct limited to the thalamus; (2) PCA territory infarct was defined as a unilateral infarct in the regions supplied by the PCA, including the thalamus, midbrain, medial temporal and occipital lobes; (3) TOB syndrome was defined as a bilateral infarct in the regions supplied by the rostral basilar artery and the superior cerebellar artery, including the thalamus, midbrain, occipital and medial temporal lobes, and superior cerebellum; and (4) EPC territory infarct was defined as multiple infarcts extending beyond the TOB territory.
Stroke aetiologies were assigned to five categories according to the TOAST (Trial of ORG 10172 in Acute Stroke Treatment) criteria10: (1) small vessel disease (SVD) in patients with a small (<15 mm) infarct in the thalamus without significant stenosis (>50%) in the large artery of the posterior circulation or cardiac sources of embolism; (2) large artery disease (LAD) in patients whose infarction size was >15 mm with an occlusion or severe stenosis (>50%) in the artery supplying the infracted areas or proximal to stroke lesions in the absence of cardiac sources of embolism; (3) cardioembolism; (4) cryptogenic cause; and (5) two or more causes. Other causes, including vasculitis, venous infarct and infarction related to an angiographic procedure, were excluded from the analysis because of the small number of patients.
The association between involved thalamic territories and stoke syndromes and between those and stroke aetiologies were analysed using the χ2 test with Fisher's exact test. Statistical significance was established at p<0.05. The descriptive statistics were also used to analyse the overall distribution of thalamic lesion patterns. SPSS software (V.12.0; SPSS, inc) was used for statistical analysis.
There were 99 men and 69 women with a mean age of 66 years (range 27–91). Stroke risk factors included hypertension (blood pressure >160/90 mm Hg at least twice before the stroke) in 118 (70%) patients, diabetes mellitus in 56 (33%) patients, hypercholesterolaemia (serum cholesterol level >6.5 mmol/l) in 42 (25%) patients, current cigarette smoking in 44 (26%) patients and cardiac sources of embolism in 44 (26%) patients (atrial fibrillation in 29 patients, hypokinetic or akinetic ventricular segments in 16 patients, congestive heart failure in 10 patients and mitral valve disease in four patients). Thirty-four (20%) patients had a previous history of stroke.
Stroke syndromes and topographic patterns of thalamic infarcts
Isolated thalamic infarcts were found in 67 (40%) patients, PCA infarcts in 38 (23%) patients, TOB in 22 (13%) patients and EPC infarcts in 41 (24%) patients.
The topography of thalamic involvement was confined to one thalamic vascular territory in 116 (69%) patients, extended to multiple thalamic territories unilaterally in 27 (16%) patients (two (3%) patients in isolated thalamic infarcts, 15 (39%) patients in PCA infarcts, three (14%) patients in TOB and seven (17%) patients in EPC infarcts) and bilaterally in 25 (15%) patients (14 (64%) patients in TOB and 11 (27%) patients in EPC infarcts).
The most commonly affected thalamic territory was the ventrolateral territory, observed in 103 (61%) patients, followed by the posteromedial territory in 65 (39%) patients, posterolateral territory in 42 (25%) patients and anterior territory in 19 (11%) patients.
The patterns of thalamic territory involvement among different stroke syndromes are shown in table 1. There was a significant association between involvement of the ventrolateral thalamus and isolated thalamic infarcts (48% (49/103) vs 28% (18/65); p<0.01) or PCA infarcts (28% (29/103) vs 14% (9/65); p=0.02), and between the posteromedial territory involvement and TOB (26% (17/65) vs 5% (5/103); p<0.001) or EPC infarcts (45% (29/65) vs 12% (12/103); p<0.001).
Taken together, the most common pattern of thalamic lesion was the single lesion restricted to the ventrolateral territory in isolated thalamic infarcts (47 patients (67%), p<0.01) and in PCA infarcts (14 patients (38%)), bilateral lesions involving both posteromedial territories in TOB (five patients (23%)) and unilateral lesion confined to the posteromedial territory in EPC infarcts (14 patients (34%)) (figure 2).
Stroke aetiologies and topographic patterns of thalamic infarcts
Stroke aetiologies were SVD in 60 (36%) patients, LAD in 30 (18%) patients, cardioembolism in 31 (18%) patients, cryptogenic cause in 39 (23%) patients and two or more causes in eight (5%) patients. Of the 30 patients with LAD, four (13%) had a stenosis or occlusion in the proximal PCA, 18 (60%) in the basilar artery, five (17%) in the vertebral artery and three (10%) in both the basilar and vertebral arteries.
The patterns of thalamic territory involvement among different stroke aetiologies are shown in table 2. The ventrolateral territory was most frequently affected in association with SVD (43 patients (72%), p=0.03). The posteromedial (19 patients (63%), p<0.01) and ventrolateral territories (15 patients (50%)) were commonly involved in LAD. In cardioembolic stroke, the posteromedial (14 patients (45%)), ventrolateral (16 patients (52%)) and posterolateral (16 patients (52%)) territories were almost equally affected.
Consequently, the most common pattern of thalamic lesion was the single ventrolateral lesion in SVD (43 patients (72%), p<0.01) and unilateral (seven patients (23%)) or bilateral (six patients (20%)) posteromedial lesion in LAD. In cardioembolism and cryptogenic cause, however, various lesion patterns were commonly observed without any specific dominant pattern, including the unilateral lesion involving the posteromedial or ventrolateral territory and unilateral lesion involving the combined ventrolateral and posterolateral territories.
To our knowledge, this is the largest series of thalamic infarcts focusing on the topographic patterns of thalamic lesions. Furthermore, DWI used in this study has the advantage in detecting acute small and multiple thalamic lesions compared with CT and traditional MRI.
Most studies on thalamic infarcts were limited to isolated thalamic infarcts. In this series, however, thalamic infarcts with concomitant infarcts outside the thalamus were also included. The data showed that 60% of thalamic infarcts were accompanied by infarcts in other structures. In other studies, the percentage of thalamic infarcts associated with other territory infarcts was 21%, 38% and 61%.3–5 These variable results may be due to the difference in the age of the study population and in the inclusion methods among the studies.
Topographic patterns of thalamic infarcts associated with stroke syndromes
In the isolated thalamic infarcts, most of the lesions were restricted to one thalamic territory (97%) whereas infarcts tended to involve multiple thalamic territories in other stroke syndromes (39% in PCA, 77% in TOB and 44% in EPC infarcts).
Consistent with previous reports, the ventrolateral territory was most commonly involved in isolated thalamic infarcts.2–5 However, there are only a few reports with respect to the specific thalamic territories involved in association with other stroke syndromes. In studies on patients with PCA infarcts, the lesions involving the thalamus were largely limited to the ventrolateral and posterolateral territory,11 12 while in another study the lesions most frequently involved the posteromedial and/or ventrolateral territory.13 In this study, involvement of the ventrolateral territory was the most common (14 patients (37%)), followed by involvement of combined ventrolateral and posterolateral territories (11 patients (29%)). Although the results are different among the different studies, the ventrolateral thalamus was the consistent region involved predominantly after PCA infarcts.
In TOB and EPC infarcts, the posteromedial territory was more commonly involved (77% and 71%, respectively) than other vascular territories and often bilaterally involved in TOB. In the literature, both territories in the medial thalamus were frequently infarcted after TOB, but the detailed frequency was not available.14 None of the previous studies have reported the pattern of thalamic involvement in EPC infarcts. The results in this study suggest that the pattern of thalamic involvement is similar between TOB and EPC infarcts. However, involvement of multiple or bilateral thalamic territories was more frequent in TOB than in EPC infarcts (multiple, 77% vs 44%; bilateral, 64% vs 27%).
Topographic patterns of thalamic infarcts associated with stroke aetiologies
In the stroke caused by SVD, the predominant vascular territory of thalamic infarcts was the ventrolateral territory (72% (43/60)). This may be attributed to the small vessel changes related to hypertensive arteriolopathy that are frequently observed in the thalamogeniculate arteries.15
In cardioembolic stroke, various lesion patterns were commonly found, affecting the ventrolateral, posteromedial or posterolateral territory. These results contrast with other reports in which the posteromedial thalamic infarcts were most commonly caused by cardioembolism.16 17 Those studies, however, included mainly patients with isolated thalamic infarcts. Although the posteromedial territory was the frequent region (45% (14/31)) of cardiogenic embolism, the ventrolateral (52% (16/31)) and posterolateral territory (52% (16/31)) were also frequent areas of cardioembolism. Therefore, the posteromedial territory was not the specific region of cardioembolism. Moreover, the paramedian territory was most frequently involved in a cryptogenic cause (31% (20/65)) and LAD (29% (19/65)), and the most frequent site of thalamic infarction caused by LAD (63% (19/30)).
Thalamic lesions were classified into four territories based on the template reported by Tatu et al.9 In some patients, however, the thalamic lesions did not correspond exactly to any of the categories. In these cases, the lesions were categorised into one of the most similar territories or into combined territories if the lesion size was larger than 15 mm. These cases may represent individual variations of the territories supplied by the thalamic arteries or a borderzone mechanism of stroke between adjacent territories.17
The patients with an incomplete work-up were not included in this study. Most of those patients did not undergo MR imaging because of impaired consciousness after severe stroke. This may have resulted in selection bias in the group of TOB and EPC infarcts. However, the proportion of such patients was relatively small and the results nevertheless showed that the posteromedial territory, which is known to be associated with consciousness, was mainly involved in TOB (77%) and EPC (71%) infarcts.
In conclusion, the present study provides information on the topographic patterns of thalamic infarcts involving specific vascular territories on DWI and highlights the association of the involvement of a particular thalamic territory and the specific stroke syndromes and aetiologies.
Funding The study was supported by the research fund of Dankook University in 2010.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.