Background and Purpose Controversy exists regarding whether lacunar infarction is due to embolism or whether it is always due to lipohyalinosis of small penetrating arteries. We hypothesized that emboli can enter penetrating arteries in relation to the blood flow to these arteries and to the diameter of the emboli.
Methods We injected agarose spheres of three different mean diameters (31±4, 68±14, and 92±28 μm [n=50 for each]) into one internal carotid artery of 3 monkeys for each sphere size (total, n=9 monkeys). After injection of spheres, monkeys were killed, the brains were removed and fixed in formalin, and serial hematoxylin and eosin sections of three coronal sections of the cerebrum were examined by light microscopy. Sphere diameter (n=25 for each territory and sphere size) and distribution in circumferential and penetrating artery territories were measured with the use of an image analyzer. Corrections were made for shrinkage of spheres during fixation and for the effect of random sampling of 10-μm sections through spheres of different diameter.
Results Mean numbers of spheres for each size were significantly higher in circumferential than penetrating artery territories (P<.05, t test). When correction was made for the volume of brain supplied by each territory, there was no significant difference in the number of spheres in circumferential versus penetrating artery territories for the two smaller sphere sizes. For spheres of mean diameter of 92 μm, significantly more spheres entered circumferential rather than penetrating artery territories (P<.05, t test). The percentage of the total number of spheres that entered penetrating artery territories was 5%, 6%, and 1.4% for beads of 31±4, 68±14, and 92±28 μm mean diameter, respectively.
Conclusions Small emboli can enter penetrating arteries and could therefore produce lacunar infarction. The majority of emboli, however, enter circumferential arteries. The larger the emboli, the more likely that they will enter circumferential arteries rather than penetrating arteries.
Fisher1 described relatively distinct clinical syndromes due to small cerebral infarcts in the territory of penetrating arteries originating from the major basal arteries of the circle of Willis. These lacunar infarcts or lacunes were thought to be related to lipohyalinosis secondary to hypertension, a process that occluded the arteries and caused stroke. Bamford and Warlow2 asked two questions about the lacunar hypothesis. First, are clinical lacunar syndromes usually caused by lacunes? Second, are lacunes usually due to local disease (lipohyalinosis) of the small perforating arteries? They concluded that the second question could not be answered because of a lack of supporting evidence. Other investigators hypothesized that lacunes might be due to embolism as well as small intracerebral hemorrhages or blood abnormalities.3 4 5 6 There has been reluctance, however, to accept the specific hypothesis that lacunes may be due to embolism.7
This experiment was designed to test the hypothesis that small emboli can enter the small perforating arteries of the brain and therefore potentially cause lacunes. We theorized that the smaller the embolus, the more likely it
is to be distributed in the small perforating arteries, in direct relation to the blood flow to that artery. Large emboli should preferentially enter the larger circumferential cerebral arteries. Monkeys were used because the arterial anatomy approximates that observed in humans.
Materials and Methods
Nine cynomolgus monkeys (weight, 2.5 to 4.0 kg) that were being killed for another experiment were used. They were randomly divided into three groups to undergo injection of one of the following types of agarose spheres into the left internal carotid artery: (1) Cibacron blue (Sigma Chemical Co), (2) Sephacryl (Sigma Chemical Co), or (3) Fractogel (Merck Darmstadt). The number of beads injected was as follows: (1) Cibacron blue, 108 000±7000; (2) Sephacryl, 69 000±1500; and (3) Fractogel, 1 025 000±54 000 (n=3 for each estimate).
Animals were sedated with ketamine HCl (6 to 10 mg/kg IM). They were endotracheally intubated and ventilated on O2 and 1% to 2% isofluorane administered by a variable-phase animal respirator (Harvard Apparatus). End-tidal Pco2 was monitored continuously and adjusted to approximately 40 mm Hg (N-100 pulse oximeter, Nellcor Corp). Monkey body temperature was maintained at 37°C by a water-filled heating pad (TP-200, Gaymar). Blood pressure and heart rate were monitored noninvasively (Dinamap 847XT, Critikon, Inc). Atropine (0.6 mg/kg IM) was administered.
The right axillary artery was exposed and catheterized with a 20-gauge polyethylene catheter. A single anteroposterior, arterial-phase cerebral angiogram was obtained by manual injection of 5 mL of iothalamate meglumine. Exposure factors were uniform (70 kV, 1.6 mA), and a magnification control standard was used in each radiograph for correction to constant magnification. The right femoral artery was exposed and catheterized with a Tracker 18 catheter (Target Therapeutics). This catheter was advanced under fluoroscopic control into the left internal carotid artery. A mixture of 250 μL of the agarose spheres and 9 mL autologous, heparinized blood was injected into the catheter over 10 minutes. Each monkey was killed by exsanguination under anesthesia 10 minutes after injection of the emboli.
Procedures involving animals were approved by the Animal Care and Use Committee of the University of Chicago. They were carried out in compliance with recommendations of the US Department of Health and Human Services and the National Institutes of Health.
Specimen Processing, Determination of Size, and Distribution of Emboli
Brains were removed with preservation of the leptomeninges. The middle cerebral and basilar arteries were removed for pharmacological and biochemical study, and the brains were placed in phosphate buffer containing 10% formalin for 1 week. The brain stems were removed, and the cerebral hemispheres were cut into 11 coronal sections of equal thickness (5 mm). Five serial 10-μm sections of each coronal section were dehydrated, embedded in paraffin, and stained with hematoxylin and eosin. To determine the size of the spheres before fixation, the diameters of 50 of each type of sphere were measured on an image analyzer. The shrinkage of spheres with fixation was quantified by measuring the diameters of 25 spheres of each type in brains from monkeys in each group. To correct for the effect of obtaining random, 10-μm sections through the spheres, we applied correction factors based on the mean diameters of the spheres of each type and on calculations based on integration of random sections through the spheres. The calculated correction factors were 0.799 for Cibacron blue spheres, 0.803 for Sephacryl, and 0.823 for Fractogel. Since there was significant additional shrinkage of Sephacryl spheres with fixation (see “Results”), an additional correction factor was applied to correct these spheres to actual size.
The distribution of emboli in the brains was determined by counting the number of spheres in each of three sections through the cerebrum that contained structures irrigated by penetrating arteries (caudate nucleus, internal capsule, thalamus, globus pallidus, and putamen). The corresponding sections from the right hemisphere were cut into two pieces representing the penetrating and circumferential vascular territories, blotted dry, and weighed. Since there was some variation in size of spheres within each sphere type, we measured the diameters of 25 spheres in each of the penetrating and circumferential territories to determine whether there was a predilection for spheres of different sizes to enter these territories. We measured the diameter of 15 penetrating arteries at their point of entry into the brain.
To determine the percentage of the total number of emboli injected that entered penetrating arteries, we integrated the mean number of emboli of each size that were observed per section over the number of sections that contained the penetrating artery territories. This estimate of the total number of emboli in penetrating arteries was expressed as a percentage of the total number of emboli injected.
Diameters of the following cerebral arteries were measured at predetermined points six times with a calibrated optical micrometer and a mean value calculated: extradural internal carotid artery, intradural internal carotid artery, middle cerebral artery, anterior cerebral artery, and basilar artery. In addition to these data, sphere diameters, distributions, and arterial diameters were entered into a computer and edited. Pairwise comparisons were made by unpaired t test. All values are mean±SD, and significance was taken as P<.05.
The mean diameters of the major cerebral arteries and of the penetrating arteries of the middle cerebral artery as they enter the brain parenchyma are given in Table 1⇓. There were no significant differences between the groups. The mean diameters of unfixed spheres, fixed spheres, and spheres in each arterial territory are shown in Table 2⇓. After correction for sectioning with the sectioning factor, there was a significant decrease in size of Sephacryl spheres after fixation (P<.05, unpaired t test). Subsequent sizes of Sephacryl spheres were corrected for this shrinkage in addition to the correction for sectioning applied for every type of sphere. Comparisons between diameters of spheres in penetrating and circumferential arterial territories showed that there were no significant differences in mean diameters of spheres of Cibacron blue, Sephacryl, or Fractogel between the two territories. There was a trend for a larger mean diameter of the larger types of spheres (Cibacron blue, Sephacryl) in the circumferential territories compared with penetrating territories.
The mean number of each type of sphere in penetrating and circumferential artery territories is shown in Table 3⇓. Mean numbers of spheres for each size were significantly higher in circumferential than penetrating artery territories (P<.05, t test). When correction was made for the mass of brain supplied by each territory, there was no significant difference in the number of spheres in circumferential versus penetrating artery territories for the two smaller sphere sizes. For spheres of mean diameter of 92 μm (Cibacron blue), significantly more spheres entered circumferential rather than penetrating artery territories (P<.05, t test). After correction for mass of brain supplied, the ratios of spheres in circumferential to penetrating artery territories were 5.0, 1.2, and 1.6 for Cibacron blue, Sephacryl, and Fractogel spheres, respectively.
The percentage of the emboli that were observed in penetrating artery territories was 5% of the Fractogel spheres, 6% of the Sephacryl spheres, and only 1.4% of the Cibacron blue spheres (Table 3⇑).
Erythrocytes are 7 μm in diameter. Since erythrocytes enter all arteries, emboli between 7 μm in diameter and the diameter of an artery itself could theoretically enter that artery, including the small perforating arteries that arise from the major vessels of the circle of Willis. Despite this, it was believed that infarcts in the territories of the small penetrating arteries could not be embolic but were due only to local small artery disease.3 7 This study shows that small emboli may enter any cerebral artery that is larger than the embolus. There is a significant tendency for larger emboli to enter larger arteries. Overall, the number of emboli that enter the penetrating arteries is small compared with the number that enter circumferential arteries. There are no previous systematic studies of the distribution of emboli of different sizes in the brains of primates. Swank and Hain8 injected wax emboli between 4 and 60 μm in diameter into the common carotid artery of dogs and studied the locations of the emboli and of infarcts produced by them. Emboli were found mainly in cortical arterial territories, and there was no specific mention of emboli or infarcts in penetrating artery territories. Hill et al9 injected autologous clot into the carotid artery of dogs and produced infarcts in cortical gray, thalamostriate, and white matter areas. Infarction confined to the thalamostriate territory occurred in 25% of dogs. In most cases clot was located in the proximal middle and anterior cerebral arteries, and it seems likely that infarction sometimes spared cortical territories because of extensive leptomeningeal collateral circulation found in the dog rather than because of specific embolic occlusion of small penetrating arteries to the thalamostriate area. Futrell et al10 11 produced lacunar infarctions in rats by generation of platelet emboli in the cervical carotid artery by photochemical injury. The rats were normotensive, and this was taken as strong evidence that emboli can enter small penetrating arteries and produce lacunes, at least in rats.
Entrance of emboli into penetrating arteries is consistent with several clinical observations. Occlusions of small arteries of 55 to 900 μm in diameter in systemic organs contained cholesterol clefts, suggesting that the occlusions were due to embolism by cholesterol fragments from large-vessel atheromata.12 The brains were not examined. Case reports document lacunar infarction developing after cardiac angiography, a situation in which the most likely cause of stroke is cerebral embolism.5 13 Reiber et al14 described a normotensive patient with carotid stenosis and transient ischemic attacks of pure motor hemiplegia/paresis that resolved after endarterectomy. Other investigations suggest that embolism is an uncommon cause of lacunar stroke. Soloway and Aronson15 examined the brains of 16 patients with atheromatous embolism to the brain and reviewed the literature on 14 additional cases. Hypertension was present in 53%. Most supratentorial lesions were in cortical territories, and only one artery in the basal ganglia was reported to contain an embolus.
What is the explanation for the discrepancy between these observations that emboli can enter small penetrating arteries and the clinical infrequency of this occurrence? First, embolism as a cause of lacunar infarction may be underrecognized or underreported. Among 108 consecutive patients with lacunar infarction, Horowitz et al6 found hypertension in 68%, a carotid artery source of embolism by noninvasive testing in 23%, and previously established criteria suggesting a risk of embolism in 18%. In the Lausanne Stroke Registry, more than 25% of patients with lacunes had a potential embolic source of stroke in the absence of another etiology for infarction.16 Second, to extrapolate from the data from this experiment, if the larger emboli are five times as likely to go to circumferential rather than penetrating artery territories, and larger emboli are even more likely to enter circumferential arteries, then it should not be surprising that most patients with embolic sources of stroke would experience cortical rather than lacunar ischemia or infarction. There are little data on the size of emboli produced by carotid stenosis.17 They may be as large as or larger than penetrating arteries. The problem is complicated by multiple potential causes of stroke that exist in some patients. Finally, since smaller emboli appear to be the only ones that enter penetrating arteries with any frequency, infarcts may be more likely to go undetected, or the emboli may be so small that they break up before producing clinically detectable cerebral ischemia and infarction.
These clinical observations, in conjunction with the experimental data presented here, suggest that lacunes may be caused by embolism. Larger emboli that are more likely to be clinically significant and to produce infarction, however, enter penetrating arteries with decreasing likelihood, suggesting that embolism is an uncommon mechanism for lacunar infarction.
- Received September 13, 1994.
- Revision received February 1, 1995.
- Accepted April 12, 1995.
- Copyright © 1995 by American Heart Association
Bamford JM, Warlow CP. Evolution and testing of the lacunar hypothesis. Stroke. 1988;19:1074-1082.
Millikan C, Futrell N. The fallacy of the lacune hypothesis. Stroke. 1990;21:1251-1257.
Bogousslavsky J. The plurality of subcortical infarction. Stroke. 1992;23:629-631.
Cacciatore A, Russo LS Jr. Lacunar infarction as an embolic complication of cardiac and arch angiography. Stroke. 1991;22:1603-1605.
Horowitz DDR, Tuhrim S, Weinberger JM, Rudolph SH. Mechanisms in lacunar infarction. Stroke. 1992;23:325-327.
Kappelle LJ, van Gijn J. The fallacy of the lacune hypothesis. Stroke. 1991;22:535. Letter.
Futrell N, Millikan C, Watson BD, Dietrich WD, Ginsberg MD. Embolic stroke from a carotid arterial source in the rat: pathology and clinical implications. Neurology. 1989;39:1050-1056.
Reiber J, Cote R, Ladouceur D. The fallacy of the lacune hypothesis. Stroke. 1991;22:534. Letter.
Soloway HB, Aronson SM. Atheromatous emboli to central nervous system: report of 16 cases. Arch Neurol. 1964;11:657-667.
Mohr JP, Pessin MS. Extracranial carotid artery disease. In: Barnett HJM, Mohr JP, Stein BM, Yatsu FM, eds. Stroke: Pathophysiology, Diagnosis, and Management. New York, NY: Churchill Livingstone, Inc; 1986:293-336.