Cavitation of Deep Lacunar Infarcts in Patients With First-Ever Lacunar Stroke
A 2-Year Follow-Up Study With MR
Background and Purpose—Studies in patients with lacunar stroke often assess the number of lacunes. However, data on how many symptomatic lacunar infarcts cavitate into a lacune are limited. We assessed the evolution of symptomatic lacunar infarcts over 2-year follow-up.
Methods—In 82 patients with first-ever lacunar stroke with a lacunar infarct in the deep brain regions (excluding the centrum semiovale), we performed a brain MR at presentation and 2 years later. We classified cavitation of lacunar infarcts at baseline and on follow-up MR as absent, incomplete, or complete. We recorded time to imaging, infarct size, and vascular risk factors.
Results—On baseline MR, 38 (46%) index infarcts showed complete or incomplete cavitation. Median time to imaging was 8 (0–73) days in noncavitated and 63 (1–184) days in cavitated lesions (P<0.05). On follow-up imaging, 94% of the lacunar infarcts were completely or incompletely cavitated, most had reduced in diameter, and 5 (6%) had disappeared. Vascular risk factors were not associated with cavitation.
Conclusion—Cavitation and lesion shrinkage were seen in almost all symptomatic lacunar infarcts in the deep brain regions over 2-year follow-up. Counting lacunes in these specific regions at a random moment might slightly, however not substantially, underestimate the burden of deep lacunar infarction.
Lacunar stroke is characterized by one of the clinical lacunar syndromes with a compatible lacunar infarct on brain imaging.1 Lacunar infarcts are small infarcts located deep within the brain caused by occlusion of a single perforating artery due to cerebral small vessel disease.2 Radiologically, old lacunar infarcts are considered to be represented by lacunes, which are focal cavities.2
Lacunes on brain imaging are often used as markers of cerebral small vessel disease. However, precise data on the proportion of lacunar infarcts that cavitate and become a lacune are limited. Former MRI studies found that only 28%3 to 60%4 of the symptomatic lacunar infarcts progress to definite cavitation, leading to the conclusion that many lacunar infarcts stay noncavitated white matter lesions. However, these studies included patients over a wide range of times to follow-up imaging and symptomatic lacunar infarcts in all subcortical brain regions. There is some evidence that although lacunar infarcts in the deep brain regions are caused by cerebral small vessel disease, lacunar infarcts in the centrum semiovale have different causes.5–7
We studied cavitation of symptomatic lacunar infarcts in the basal ganglia, pons, and internal capsule, for which there was no other cause than intrinsic cerebral small vessel disease, in patients with first-ever lacunar stroke, all of whom had a 2-year follow-up MRI.
Subjects and Methods
For a detailed description, see the online-only Data Supplemental Methods. In short, we included 82 patients with first-ever lacunar stroke who had a MRI at baseline and after 2 years (fluid-attenuated inversion recovery and T2 sequences); 59% also had diffusion-weighted imaging at baseline. An acute lacunar infarct was defined as a round T2-weighted hyperintense lesion (<20 mm) with, if diffusion-weighted imaging was performed, restricted diffusion, located in the basal ganglia, internal capsule, or pons, and compatible with clinical findings. We excluded all patients with symptomatic centrum semioval infarcts at baseline. Cavitation on imaging (fluid-attenuated inversion recovery), at baseline and follow-up, was classified into: absent, incompletely, or completely cavitated infarcts (Figure). The total diameter of each lacunar infarct was measured. We also assessed the presence of asymptomatic lacunes and graded white matter lesions according to the Fazekas scale.
Cavitation at Baseline MR
Table 1 shows baseline measurements. Median time between stroke onset and baseline MR was 22 days (0–184 days). At baseline, 44 of the 82 lacunar infarcts (53%) showed no cavitation, 27 (33%) were incompletely cavitated, and 11 (13%) were completely cavitated.
Longer time interval between stroke onset and baseline imaging was significantly associated with cavitation. No other patient-related or imaging-related variables (white matter lesions or asymptomatic lacunes) were associated with cavitation.
Cavitation on Follow-Up MR
Results are shown in Table 2. The follow-up brain MRI was performed at mean 759±58 days after the baseline MRI. In the group without cavitation at baseline (n=44), 16 (36%) showed incomplete cavitation, 24 (55%) were completely cavitated, and in 4 cases (9%), no lesion was seen on follow-up imaging. In the group with incomplete cavitation at baseline (n=27), 5 (19%) were still incompletely cavitated, 21 (78%) showed complete cavitation, and in one case, no lesion was seen on follow-up. There was no association between vascular risk factors and cavitation on follow-up.
The total diameter of the noncavitated lesions at baseline (11.5±3.7 mm) reduced to 6.5±3.4 mm on follow-up imaging (P<0.05). Infarct size in the group of 11 completely cavitated lesions at baseline was 8.8±4.0 mm and showed a significant reduction (6.6±3.7 mm; P<0.05) on follow-up.
We assessed the evolution of symptomatic lacunar infarcts in the basal ganglia and internal capsule due to presumed cerebral small vessel disease over a period of 2 years. We found that all symptomatic lacunar infarcts, except those no longer visible, cavitate over time and that longer time interval between stroke onset and imaging is the sole predictor of cavitation. Furthermore, we found a significant reduction of infarct size on follow-up.
Potter3 found definitive cavitation in one fifth of patients with symptomatic lacunar stroke and concluded that many lacunar infarcts long term may resemble white matter lesion. However, they included symptomatic lacunar infarcts both in the centrum semiovale and the basal ganglia, their time interval to follow-up imaging was generally shorter, they partly used CT for imaging, and their MR scoring methods might not completely correspond with ours. Koch4 found that cavitation occurred in 61% of patients with lacunar stroke, but their follow-up time interval was widespread and their patient sample with MRI follow-up was relatively small. Because there is a shortage of data on long-term appearances of symptomatic lacunar infarcts, our data almost double the number of the total published MRI cases so far, even if there are limitations such as the restricted definition of location of lacunar infracts in our study.
We did not include lacunar infarcts in the centrum semiovale in contrast to these 2 studies, because the centrum semiovale is generally supplied by medullary branches of the cortical arteries and these infarcts may have different causes than those arising in the territory of the deep perforators.5–7 It may also be that the deep gray and white matter respond differently to lacunar damage, gray matter being more likely to cavitate than white matter. It would be interesting to directly compare cavitation rates between small deep and centrum semiovale infarcts.
Even when cavitation has been reached, it seems that the lacune is still in a dynamic process because we showed reduction in size over time. In 5 cases the lesion even disappeared on follow-up MR. Although it is possible that the brain tissue healed without the formation of a lacune, it seems more likely that these infarcts have progressed into small, collapsed lacunes that can be missed by brain imaging, possibly corresponding to microinfarcts.8
The main strengths of our study are a relatively fixed long follow-up time and MRI in all patients, which is superior to CT for the detection of lacunar infarcts.9 A limitation is that we did not use the same MR field strength in all patients due to the clinical setting of this study; however, other MRI settings were standardized. Furthermore, not all patients had diffusion-weighted imaging, which could have caused that the wrong lesion was counted for the symptomatic lacunar infarct. However, we only included patients with a definitive clinical diagnosis of first-ever lacunar stroke and the lesion had to be compatible with clinical signs.
In conclusion, cavitation of symptomatic lacunar infarcts was present in all, except for those that disappeared over time. Cavitation seems to be a dynamic process with a reduction in size over time and with a definitive lacune as a final result. Because it takes time to cavitate, counting lacunes at a random moment might slightly, however not substantially, underestimate the burden of deep lacunar infarcts, but these results cannot be applied to lacunar infarcts in the centrum semiovale.
Source of Funding
This project was supported by the Dutch Pearl String Initiative (C.M.J.L.).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.112.660076/-/DC1.
- Received February 19, 2012.
- Revision received April 6, 2012.
- Accepted April 24, 2012.
- © 2012 American Heart Association, Inc.
- Donnan GA,
- Norring B,
- Bamford J,
- Bogousslavsky J
- Donnan GA,
- Norring B,
- Bammer R,
- Bogousslavsky J
- Fisher CM
- Potter GM,
- Doubal FN,
- Jackson CA,
- Chappell FM,
- Sudlow CL,
- Dennis MS,
- Wardlaw JM
- Koch S,
- McClendon MS,
- Bhatia R
- Bogousslavsky J,
- Regli F
- Uehara T,
- Tabuchi M,
- Mori E