This study characterizes the regional changes in vascularity, which accompanies chronic progressive hydrocephalus. Fifteen dogs underwent surgical induction of hydrocephalus and were used for histologic studies. Animals were divided into 4 groups: surgical control, short term (< or = 5 weeks), intermediate term (8 weeks), and long term (10 to 12 weeks). Vessel diameter, density, and luminal area were calculated by imaging quantification after manual vessel identification in the cortical gray, white matter, and caudate nucleus. Capillary vessel diameter decreased 23.5% to 30.2% (P < 0.01) in the caudate, but then returned to normal at 12 weeks. Capillary vessel density decreased 53.5% (P < 0.05) in the cortical gray, but then increased to 234.8% (P < 0.01) over surgical controls at 12 weeks. There was no initial decrease in capillary density in the caudate; however, the long-term group capillary density was significantly greater (172.8% to 210.5%, P < 0.01) than surgical controls. Overall, there was a short-term decrease in lumen area, with recovery in the longer term. Glial fibrillary acidic protein (GFAP) immunohistochemistry demonstrated the pattern of GFAP staining and reactive astrocytes differed in the caudate compared with the occipital cortex. This data suggest that an increase in capillary density and diameter may be an adaptive process allowing maintenance of adequate cerebral perfusion and metabolic support in the hypoxic environment of chronic hydrocephalus.

No abstract

In the large canine model of acquired obstructive hydrocephalus that we have developed recently, computer-assisted 3-dimensional morphometry has been performed on T1-weighted Spin Echo MRI images from adult dogs before and after the induction of hydrocephalus. To date, 7 hydrocephalic animals have been analyzed that survived 7-83 days (median = 54) after receiving injections of cyanoacrylate glue into the anterior fourth ventricle. Measurements were obtained from lateral, 3rd, and 4th ventricles. The volumes of the left and right lateral ventricles were symmetrical before and after induction. Mean lateral ventricle volume increased 424% from a baseline of 0.63 cc to a post-induction value of 3.30 cc (p < 0.01 with unpaired t-test). In contrast, the 3rd ventricle expanded 187% from a mean of 0.15 cc to 0.43 cc (p < 0.05). The combined volume of the lateral and 3rd ventricles increased 369% from a mean of 0.78 cc to 3.69 cc (p < 0.01). Evans' ratios, which are used routinely in the clinical setting, were also obtained from linear measurements of the lateral ventricle width divided by brain width at the level of the foramen of Monro. These values exhibited only a 94% increase from mean baseline ratios of 0.17 to post-induction ratios of 0.33 (p < 0.05). These findings indicate that in mechanically-induced obstructive hydrocephalus the relative expansion of the lateral ventricles is greater than that of the 3rd ventricle. In addition, volumetric measurements of the lateral and 3rd ventricles suggest that the extent of ventriculomegaly is 3-4 times greater than estimated by Evans' ratios.

Although neonatal hydrocephalus often results in residual neurological impairments, little is known about the cellular mechanisms responsible for these deficits. The immediate early gene, fos (c-fos), functions as a “third messenger” to regulate protein synthesis and is a good marker for neuronal activation. To identify functional changes in neurons at the cellular level, the authors quantified fos RNA expression and localized fos protein in the H-Tx rat model of congenital hydrocephalus. Tissue samples from sensorimotor and auditory regions were obtained from hydrocephalic rats and age-matched, normal litter mates at 1, 6, 12, and 21 days of age (four-six animals in each group) and processed for immunohistochemical analysis of fos and Northern blot analysis of RNA. At 12 days of age, hydrocephalic animals exhibited significant decreases in the ratio of fos immunoreactive cells to Nissl-stained neurons from both cortical regions, but no statistical differences were noted in fos expression. At 21 days of age, both the ratio of fos immunoreactive cells to Nissl-stained neurons and fos expression decreased significantly. The number of fos-positive neurons decreased in all cortical layers but was most prominent in layers V through VI. This decrease did not appear to be caused by neuronal death because examination of Nissl-stained sections revealed many viable neurons within the areas where fos immunoreactivity was absent. These results suggest that progressive neonatal hydrocephalus reduces the capacity for neuronal activation in the cerebral cortex, primarily in those neurons that provide corticofugal projections, and that this impairment may begin during relatively early stages of ventriculomegaly.