Changes in the Cerebral Vascular Bed in Experimental Hydrocephalus: An Angioarchitectural and Histological Study.

Changes in the Cerebral Vascular Bed in Experimental Hydrocephalus: An Angioarchitectural and Histological Study

Summary. The angioarchitectural and histological changes of small cerebral blood vessels in experimental hydrocephalus were studied in order to assess changes of the vascular bed in the cerebral mantle.

Changes of the microvasculature assessed from microcorrosion casts by scanning electron microscopy (SEM) and histological changes shown by light and electron microscopy were compared before and after shunting for hydrocephalus. Changes of the regional cerebral blood flow (rCBF) were also evaluated by the hydrogen clearance method.

In hydrocephalus, a reduction in the number and caliber of the capillaries was noted in both the white and gray matter in the SEM study, but the capillaries were preserved and changes were mild and nonspecific in the elec­ tron microscopic examination. Shunting resulted in the reversal of all these changes to normal, along with recovery of the rCBF, which had decreased in hydrocephalus.

These observations suggest that changes of the vascular bed participate in the alteration of cerebral mantle width in the hydrocephalic process, and that changes of the microvasculature result not only from damage to the capillaries themselves, but also from changes of the perivascular structures.

Keywords. Hydrocephalus - Ventriculo-peritoneal shunting - Scanning electron microscopy -Electron microscopy -Cerebral blood flow

Introduction

We have previously reported changes of the microvasculature in hydrocephalic rats, as shown by scanning electron microscopy (SEM), in an investigation of the effects of hydrocephalus on the vascular bed in the cerebral mantle (Oka et al. 1986). In order to study changes of the microvasculature after cerebrospinal fluid (CSF) shunting, rabbits were used in the present investigation.

Histological changes were assessed using light and electron microscopy and changes of the rCBF were determined by the hydrogen clearance method. All parameters were evaluated before and after shunting to determine changes of the vasculature produced by the relief of hydrocephalus.

Materials and Methods

Hydrocephalus Model and Ventriculo-Peritoneal Shunting

Rabbits weighing 1.5-2.0kg were used for this experiment. Under intravenous pentobarbital sodium anesthesia (20-25 mg/kg), a 2-cm sterile midline skin incision was made at the occipital region. After exposing and incising the atlanto-occipital membrane, about 1.0-1.5ml CSF was removed. Then 0.6- 0.8 ml of a kaolin suspension (250 mg/ml) was slowly injected into the cisterna magna. To prevent reflux of the kaolin suspension, a piece of muscle with Aron-a was used to cover the incision.

Ventricular size was measured on coronal computed tomography (CT) scans one week, one month, one and a half months, and two months after the kaolin injection. The degree of ventricular dilatation was divided into three categories (mild, moderate, and severe) in accordance with the ratio of the maximum distance between the bilateral anterior horns and the inside diameter of the skull in the same slice (Fig. 1).

In 11 of the 22 rabbits with severe hydrocephalus, ventriculo-peritoneal shunting was performed two months after the kaolin injection.

Scanning Electron Microscopic (SEM) Study

Seven rabbits. with hydrocephalus (three moderate and four severe), four rabbits after shunting , and nine untreated control rabbits were used for this experiment. Two months after the injection of kaolin, thoracotomy was per­ formed under pentobarbital sodium anesthesia (25 mg/kg). After ligation of the innominate artery and the aortic arch, a 19-gauge elastic needle was used to cannulate the ascending aorta from the left ventricle. The rabbits were then perfused with physiological saline solution, followed by the slow injection of cooled polyester resin (Mercox), at 120mmHg pressure, into the ascending aorta until the resin emerged bilaterally from the cut jugular veins.

The heads were decapitated at the neck and the skin and mandibula were removed. Heads were then embedded in 6% carboxylmethyl cellulose and frozen at -70°C with hexan and dry ice. Each head was cut into serial sagittal sections to nearly the midline with a microtome (LKB 2250, PMV450MP). The rest of the hemisphere was soaked at about 40°C in 2% Triton X-100 solution with 25mM NarEDTA and 2N sodium hydroxide, for about three weeks, to digest the muscle, cranium, and brain tissue. The microvasculature of the gray and white matter in the parietal and frontal areas was then examined by SEM (Hitachi X-650).

Light and Electron Microscopic Study

Eight rabbits with hydrocephalus (four moderate and four severe), four rab­ bits after shunting, and five untreated control rabbits were used for this study. After thoracotomy under intravenous pentobarbital sodium anesthesia (25 mg/kg), they were perfused via the left ventricular route with 0.5% glutaraldehyde/4% paraformaldehyde in 0.01 M phosphate buffer. The brains were then sectioned coronally into slices of about 3mm in thickness. For the electron microscopic study, specimens of gray and white matter in the parietal region were soaked in the same fixing fluid for one night, postfixed in buffered 1% osmium tetroxide for 1 hour, dehydrated in ethyl alcohol, and embedded in Epon 812. Ultrathin sections were double-stained with uranyl acetate and lead citrate and examined under an electron microscope (Hitachi H-300). For the light microscopic study, the sections were stained with hematoxylin and eosin (H & E) and periodic acid-methenamine (PAM).

Regional Cerebral Blood Flow (rCBF)

Regional cerebral blood flow was measured in the parietal gray and white matter by the inhalation method in eight rabbits with hydrocephalus (four moderate and four severe), four rabbits after shunting, and five untreated control rabbits.

Results

Scanning Electron Microscopy

Corrosion casts of the rabbits with hydrocephalus showed that the vasculature was sparse; the main trunks of the vessels could be seen more clearly compared with the controls (Fig. 2). In the SEM study, there was an obvious reduction in both the number and the caliber of the capillaries in hydrocephalus, and this change tended to increase in proportion to the severity of the hydrocephalus. The capillaries were about 8-10 Jlm in diameter in the controls and decreased to a diameter of about 5-8 Jlm in severe hydrocephalus. The number of capillaries returned to normal and their caliber also recovered to about 6- 11Jlm after shunting (Figs. 3 and 4, Table 1). These changes were seen almost equally in the gray and white matter.

Cortical straight and parallel vessels showing a typical "palisade" pattern were distorted in hydrocephalus and returned to normal after shunting (Fig. 5).

Light and Electron Microscopic Study

The light microscopic study showed that the spongiomatous change which was found in the periventricular white matter in hydrocephalus disappeared after shunting. Pathological vessels could not be found in either the hydrocephalic or the post-shunting groups.

In the electron microscopic study, swelling of astrocytes in the perivascular area was observed in hydrocephalus (Fig. 6). Vacuoles, microvilli, and webs were noted more frequently in the endothelial cells of the capillaries in hydro­ cf'phalus as compared to controls (Fig. 7), and these changes were found more frequently in the white matter than in the gray matter. Following shunting, these perivascular and endothelial changes disappeared (Fig. 8). Opening of tight junctions, degeneration and reactive proliferation of endothelial cells, and abnormal vessels indicating neovascularization were not observed in this study.

Regional Cerebral Blood Flow (rCBF) The rCBF of the gray and the white matter in five normal controls was 34.7 ± 4.5ml/100g per min and 17.8 ± 3.3ml/100g per min, respectively. Eight

hydrocephalic rabbits had a rCBF of 24.2 ± 6.7 ml/100 g per min in the gray matter and 14.8 ± 3.7ml/100g per min in the white matter. A decrease in rCBF was thus noted in both the gray and the white matter, and this change became more marked in severe hydrocephalus when compared to moderate hydrocephalus . The rCBF of the gray matter was significantly (P < 0.05) reduced in hydrocephalus compared with that in the normal controls . After shunting, the rCBF recovered to 37.1 ± 9.7 ml/100 g per min in the gray matter and to 21.3 ± 4.2 ml/100 g per min in the white matter. These were significant increases (P < 0.05) from the values seen in hydrocephalus (Tables 2 and 3).

Discussion

An early report on the cerebral vasculature in hydrocephalus was made by Penfield (1929) who performed a macroscopic necropsy study. For the inves­ tigation of the cerebral angioarchitecture several methods have been reported,

including microradiography (Hassler 1964; Sato et al. 1984; Wozniak et al. 1975), microangiography (Okuyama et al. 1987; Plets 1986; Sato et al. 1984), histological microscopic examination (De 1950; Del Bigio and Bruni 1988), and the use of vascular corrosion casting (Oka et al. 1986).

Microradiography involves exposing sections of the brain to X-ray films. Although this method produces excellent stereoscopic pictures of the spatial distribution of the vessels, it has several disadvantages: (1) artifacts may be produced during the injection procedure, (2) obstructed vessels cannot be demonstrated, and (3) most of the capillaries and veins do not fill with X-ray contrast medium and are not visualized.

The microangiographic method, which involves observation, under the microscope, of the injection of various dyes into the cerebral vessels, allows the

study of smaller vessels than the microradiographic method, but is not so good at showing the spatial distribution of the vessels, because of the thin sections used.

Several histological studies of the angioarchitecture in hydrocephalus have been reported. De (1950) observed the vascular pattern of the smaller vessels using Pickworth's stain. Del Bigio et al. (1988) examined periventricular blood vessels by light microscopy and assessed them by a quantitative method.

Using microvascular corrosion casting, the cerebral vasculature can be observed more stereoscopically and with a better understanding of the suc­ cessive levels of the vascular tree than is possible by the microradiographic method. Also, the caliber of vessels can be measured more exactly by this method than by microangiography and histological studies, which reveal only sectioned vessels. In addition, this method can demonstrate the capillaries and veins by the injection of polyester resin into the venous system.

Reports on the cerebral angioarchitecture in hydrocephalus have been some­ what varied, probably owing to differences in the methods used and differences in the caliber of the vessels which were observed. Hassler (1964) used micro­ radiography to observe vessels above the precapillary level in experimental hydrocephalus, and described an increase in the number of arteries and veins in the periventricular white matter and in the cortex. Plets (1986) also stated that the arterioles in the subependymal region showed a relative hyper­ vascularization. On the other hand, Wozniak et al. (1975) found a decrease in the number and caliber of the microvessels in advanced hydrocephalus in the congenitally hydrocephalic Hy-3 mouse. There are several other reports that indicate poor visualization of the microvessels in the periventricular region.

As for the capillaries, only a few studies have been published. Hassler (1964) observed the capillaries in the hydrocephalic necropsied brain and in experi­ mental hydrocephalus by the peroxidase staining method, and reported them to seem normal in the atrophic white matter. In contrast, Del Bigio and Bruni (1988) described a significant decrease in the number of capillaries with lamina of 10 11m or less in the paraventricular area one week after silicon injection. De (1950) also reported a decrease of the capillaries in hydrocephalus.

After shunting, most studies reported that the microvasculature was restored to normal, although in several studies (Del Bigio and Bruni 1988; Plets 1986) in which CSF shunting was performed in the late stages of hydrocephalus, changes of the microvasculature were irreversible.

The most marked change noted in our study was the reduction in the number and caliber of the capillaries in the white and gray matter in chronic hydro­ cephalus. This change reversed completely after shunting. However, shunting was performed only two months after the induction of hydrocephalus, so a longer period of observation would be necessary to better assess the reversi­ bility of changes of the cerebral angioarchitecture in hydrocephalus.

Regarding morphological studies of the capillaries using microscopic methods, changes of the capillaries in the hydrocephalic brain seemed to be mild. Nakagawa et al. (1984) found clefts and vacuoles between the tight junctions of the capillaries and postcapillary venules in kaolin-induced hydro-56 N. Oka et a/.

cephalus, and suggested that the tight junction acted as a shunt pathway for interstitial edema fluid and cerebrospinal fluid to enter the microvessels. Regarding changes of the tight junctions themselves and changes of the endo­ thelial cells in hydrocephalus, Okuyama et al. (1987), from the finding of stenotic or occluded capillaries in the late stage of hydrocephalus, suggested the possible disruption of the blood-brain barrier. In our study, vacuoles, microvilli, and webs, which are thought to be an early and nonspecific response to various stresses, were found in the endothelial cells of the capillaries in hydrocephalus. All these changes reversed after shunting.

The reduction in rCBF in hydrocephalus improved after shunting, which was a similar result to that cited in previous reports (Higashi et al. 1986; Hochwald et al. 1975; Nakamura and Hochwald 1983). The marked decrease in rCBF in the gray matter found in our study, which Murata et al. (1980) also described, contrasted with another report (Sato et al. 1984) of reduced rCBF in the white matter. Aseptic meningitis induced by the injection of kaolin suspension might have been related to the reduction in cortical blood flow.

In hydrocephalus, the vascular bed seemed to decrease with ventricular dilatation and to increase again after shunting. These changes are suggested to participate, not in a small way, in the restoration of cerebral mantle width.

Regarding our results using microcorrosion casting, some questions are raised with regard to the capillaries: (1) Was the sparsity of the vascular cast in hydrocephalus caused by the disappearance of capillaries which actually changed irreversibly? (2) Was this changes caused by the lack of flow of the polyester resin into narrowed capillaries which actually still existed? (3) Was the increase in capillaries after shunting due to the same vessels that were present previously or due to neocapillaries? We also investigated histological changes in order to answer these questions. Our histological studies showed that capillaries were found even in the thinned-out cerebral mantle and that neocapillarization did not occur in the re-expanded mantle after shunting. Therefore, the apparent changes of the capillaries which we observed in hydro­ cephalus may not have been caused by changes of the vessels themselves, but may have been secondary to the influence of changes in the periventricular structures. Enlargement of the extracellular space due to interstitial edema and the hypertrophy and accumulation of astrocytes may have caused the capillaries to collapse in hydrocephalus, and these changes were then nor­ malized after shunting, leading to capillary re-expansion.

Conclusions

1. A reduction in the number and caliber of the cerebral capillaries was the most marked change seen in hydrocephalus. This change was found in both the gray and the white matter.

2. The number and caliber of the capillaries returned to normal with re­ expansion of the cerebral mantle after shunting.

3. Capillary endothelial cells in hydrocephalus showed an increase of vacuoles, microvilli, and webs. These changes were reversed by shunting, and neo­ vascularization was not seen.

4. The rCBF was reduced in hydrocephalus and improved after shunting.

5. These findings suggest that, in the hydrocephalic process, changes of the cerebral vascular bed may participate in the alterations of the cerebral mantle. In addition, the reversal of the changes in the capillaries was apparently brought about not by the vessels themselves, but by changes of the perivascular elements.

References

De SN (1950) A study of the changes in the brain in experimental internal hydro­ cephalus. J Pathol 112: 197-209

Del Bigio MR, Bruni JE (1988) Changes in periventricular vasculature of rabbit brain following induction of hydrocephalus and after shunting. J Neurosurg 69: 115-120 Hassler 0 (1964) Angioarchitecture in hydrocephalus. An autopsy and experimental

study with the aid of microangiography. Acta Neuropathol (Berl) 4: 65-74

Higashi K, Asahisa H, Ueda N, Kobayashi K, Hara K, Noda Y (1986) Cerebral blood flow and metabolism in experimental hydrocephalus. Neurol Res 8: 169-176

Hochwald GM, Boa! RD, Martin AE, Kumer AJ (1975) Changes in regional blood flow and water content of brain and spinal cord in acute and chronic experimental hydrocephalus. Dev Med Child Neural [Suppl] 17: 42-50

Murata T, Yamagata S, Mori K, Handa H, Nakano Y (1980) Computed tomography in experimental canine hydrocephalus. Part 4: Periventricular lucency (PVL) and regional blood flow in the chronic stage of hydrocephalus (in Japanese). No To Shinkei 32: 219-227

Nakagawa Y, Cervos-Navarro J, Artigas J (1984) A possible paracellular route for the resolution of hydrocephalic edema. Acta Neuropathol (Berl) 64: 122-128

Nakamura S, Hochwald GM (1983) Effect of arterial PC02 and cerebrospinal fluid volume flow rate changes on choroidal plexus and cerebral blood flow in normal and experimental hydrocephalic cats. J Cereb Blood Flow Metab 3: 369-375

Oka N, Nakada J, Endo S, Takaku A (1986) Angioarchitecture in experimental hydrocephalus. Pediatr Neurosci 12: 294-299

Okuyama T, Hashi K, Sasaki S, Sudo K, Kurokawa Y (1987) Changes in cerebral microvasculature in congenital hydrocephalus of the inbred rat LEW/Jms: light and electron microscopic examination. Surg Neural 27: 338-342

Penfield W (1929) Cerebral pressure atrophy. In: Penfield W (ed) Proceedings, Asso­ ciation for Research in Nervous and Mental Disease, Vol. 6, Internal pressure in health and disease. Williams and Wilkins, Baltimore, pp 346-361

Plets C (1986) Influence of experimental hydrocephalus on cerebral vascularization. In: Baethmann A, Go KG, Unterberg A (eds) Mechanism of secondary brain damage. Plenum, New York, pp 169-178

Sato 0, Ohya M, Nojiri K, Tsugane R (1984) Microcirculatory changes in experimental hydrocephalus: morphological and physiological studies. In: Shapiro K, Marmarou A, Portnoy H (eds) Hydrocephalus. Raven, New York, pp 215-230

Wozniak M, MeLone DG, Raimondi AG (1975) Micro- and macrovascular changes as the direct cause of parenchymal destruction in congenital murine hydrocephalus. J Neurosurg 43: 535-545