Nearoradiology

Neuroradiology 16,207-211 (1978)

© by Springer-Verlag1978

Computer Analysis of Periventricular Lucency on the CT Scan M. Asada 1 , N. Tamaki 1 , Y. Kanazawa I , S. Matsumoto 1 , M. Matsuo 2 , S. Kimura2 , S. Fujiia , and Y. Kaneda a Departments of 1Neurosurgery and 2Radiology, Kobe University School of Medicine, and 3Department of System Engineering, Kobe University School of Engineering, Kobe, Japan

Summary. Of 53 patients with periventricular lucency, 70% revealed obstructive hydrocephalus, mainly due to infratentorial tumors, and the remainder were cases of communicating hydrocephalus, usually secondary to sub arachnoid hemorrhage. Of the patients with PVL, 95% had hypertensive hydrocephalus and 60% showed slightly dilated ventricles. Computer analysis was helpful in displaying PVL objectively a n d clarifying its nature. In experimentally induced obstructive hydrocephalus, PVL was observed at the acute hypertensive stage. We coneluded that the pathogenesis of PVL might be the passive diffusion or acute extravasation of CSF into white matter through the disrupted ventricular wall, rather than transependymal absorption of CSF.

Table 1. Classification of the cases with PVL 1. Obstructive hydrocephalus

37

acoustic neurinoma paraseUar tumor pontine tumor meduUoblastoma cerebellar tumor unverified infratentorial tumor aqueduct stenosis pineal tumor thalamic tumor intracerebral hematoma supratentorial tumor giant basilar head aneurysm

8 5 5 3 2 2 2 1 2 3 3 1

2. Communicating hydrocephalus subarachnoid hemorrhage diffuse meningeal earcinomatosis head trauma porencephaly Periventricular lucency (PVL), which is usually seen in the superior lateral margin of the dilated anteriorhorn, is one of the most striking phenomena detected b y computerized tomography (CT).The pathogenesis is, however, controversial. This paper describes the clinical characteristics of PVL and correlates it to the various pathologic factors associated with hydrocephalus. Experimental study was performed using animal models of obstructive hydrocephalus to clarify the mechanism and pathogenesis of PVL.

Materials and Methods CT scanning has been performed on about 2000 patients at Kobe University Hospital over the past 2 years, using EMI scanners 1000 and 1010. Fifty-three patients out of our series showed PVL of various degrees. The patients with PVL were divided into two groups: those of obstructive hydrocephalus and of communicating hydrocephalus. The composition of the two groups is shown in Table 1.

16 11 3 1 1 53 cases

The extent of hydrocephalus was graded as follows (Table 2). The hydrocephalic index was expressed on CT scan demonstrating the level of Monro's foramen by the greatest transverse diameter of anterior horn the greatest transverse diameter of brain A hydroeephalic index less than 0.30 was considered normal, greater than 0.3 and less than 0.4 mild hydrocephalus, greater than 0.4 and less than 0.5 moderate, and greater than 0.5 marked. Intracranial pressure was measured by a ventricular tap during operation or lumbar puncture in 18 out of 53 patients with PVL (Table 3). For the purpose of analyzing the characteristics of PVL, we devised the new method using a computer (FACOM 230-38); the ventricular image was delineated manually on the digital printout. First, the oblique line crossing the lucent zone was drawn, then the starting point on the line was decided in the ventricle. The original data were processed by computer using nine points smoothing with uniform weight. CT values on the above line were automatically calculated, and the calculated CT values at different points were dotted on the graph.

0028-3940/78/0016/0207/$ 01.00

208

M. Asada et al.: Periventricular Lucency on CT Scan Experimental obstructive hydrocephalus was produced using adult dogs weighing 8 to 10 kg. Under light Nembutal anesthesia and intubation, the dogs were positioned with the face flexed down as much as possible. After suboceipital eraniectomy, the eisterna magna was opened and the lower vermis was retracted upward to expose the lower half of the fourth ventricle. A silicone catheter with the tip attached to a latex balloon was introduced into the third ventricle through the fourth ventricle and aqueduct. The balloon was inflated. After operation, CT scanning was done every 12 h until the animal died.

Table 2. PVL and hydrocephalic index HI

No. of eases

-0.29 0.30-0.34 0.35-0.39 0.40-0.44 0.45 -0.49 0.50-

5 13 17 9 4 5

Hydrocephalicindex (HI) = the greatest transverse diameter of anterior horn the greatest transverse diameter of brain

Results

Table 3. PVL and intracranial pressure

Normal intraeranial pressure Increased intracranial pressure CSF pressure (200 mmH20 t) Choked disc Other symptoms of increased ICP Unknown

16 14 13 8

Fig. 1. Untreated aqueduct stenosis (4-year-old boy). This case showed the so-called compensated hydrocephalus. No PVL was noted e T number

50

"U'-'~.

~

\

'-..\.~\~ , perlventrlcular "-. ::~¢.. '\ latency (--~.-......

"\

per iveut ricular~'~,~'i, -lueency (+): 20

~ :~. . . . . . . ~ - ~ . ~

to skull ,---

lO location

/

ventricle

t ....

2o

Fig. 2. CT values in positive PVL and negative PVL. The solid lines of positive PVL group show slow decrease of CT values; the chain lines of negative group show a steep decrease

Of 53 patients, 37 (70%) had obstructive hydrocephalus, mainly due to infratentorial tumors. The degree of PVL was more prominent in patients with infratentorial and intra.axial tumors than with extra-axial tumors. Sixteen patients (30%)~,,showed communicating hydrocephalus. Of these cases, 12 were hydrocephalus secondary to subaraehnoid hemorrhage. Of 53 patients with PVL, 35 (60%) revealed normal or mildly enlarged ventricles, and 13 patients (25%) showed moderately enlarged ventricles. It was interesting that PVL was rarely observed in marked hydrocephalus, which was usuaUy seen in nontumorous aqueduct stenosis (Fig. 1). Of 18 patients who underwent lumbar puncture or ventrieular tap, 16 had high intracranial pressure over 200 mmH20. The remaining two of these 18 showed normal CSF pressure and had subarachnoid hemorrhage. They had shown a clinical picture of increased intracranial pressure in the past. Choked discs were observed in 14 patients, though their intracranial pressure had not been measured. Thirteen patients had symptoms of headache, nausea, and vomiting. As a result, 43 patients (95%) were considered to have increased intracranial pressure. Computer analyses of PVL are shown in Figures 2, 3, 4, and 5. On the figures, CT values are dotted on the vertical axis, and the distance from the skull on the horizontal axis, so the edge of the ventricle is situated toward the right of the horizontal axis. When CT values were dotted continuously on the horizontal axis to the left, a steep decrease of CT values was observed in the negative luceney group. On the other hand, a slow increase of CT values was apparent around the subependymal part in the positive lueency group (Fig. 2). There was no significant increase of CT values in PVL with intravenous contrast enhancement (Fig. 3). The PVL disappeared soon after shunting operation (Fig. 4). The curve shifted to the right side after operation. After injection of metrizamide into the lateral ventricle through external ventricular drainage, CT values of PVL increased with time (Fig. 5). These findings might be due to the influx of metrizamide into the PVL through the ventricular wall. CT scanning was performed on normal dogs and on dogs with experimentally induced hydrocephalus. The ventricle in a normal dog is usually too small to be visualized on a CT scan (Fig. 6a). Experimental obstructive

M. Asada et al.: Periventricular Lucency on CT Scan

209

Fig. 3. CT values in PVL and intravenous contrast enhancement. The solid line shows pre-enhaneed state, and the dotted line postenhanced state. There was no significant increase of CT values in PVL

Fig. 4. CT values in PVL and shunt operation. The solid line shows preshunted state, and the chain line postshunted state. The line has shifted to the right, which shows the disappearance of PVL

210

M. Asada et al.: Periventricular Lucency on CT Scan

Fig. 5. PVL and intraventricular injection of metrizamide. CT values increased with time in PVL. The solid thick line shows the prein-~ jected state, the dotted line the state at 3 h, the chain line at 12 h, and the solid thin line at 24 h

Discussion

Fig. 6a and b. a CT of a normal dog. The ventricle was small, b CT of the adult dog with experimental obstructive hydrocephalus. The dilated ventricle and PVL (arrows) were clearly observed

hydrocephalus developed rapidly within 78 h after inflation o f the balloon. Enlargement o f anterior horns and PVL were clearly demonstrated (Fig. 6b). The ventricle dilated rapidly in the first 24 h, and PVL was generally observed on the third or fourth day after inflation o f the balloon.

Since periventricular lucency was described in pediatric hydrocephalus by Naidich et al. [7] in 1976, a few papers [2, 6, 8] on PVL have been published. Naidich et al. [7] considered two mechanisms for the pathogenesis of PVL: transependymal absorption of CSF and CSF extravasation with acute ventricular rupture, on the basis o f clinical analysis and reviews of experimental results. Mori et al. [6] concluded that PVL represented acute edema caused by CSF extravasation due to increased intraventricular pressure, rather than transependymal absorption. On the other hand, Hopkins et al. [2] speculated that PVL might be transventricular absorption, since PVL was seen diffusely in the lateral ventricle, while ventricular rupture was usually localized, and PVL was observed in moderate hydrocephalus which was unlikely to lead to ventricular rupture. It is evident from our clinical analysis that PVL is the definite and direct sign o f increased intracranial pressure. It is often observed in obstructive hydrocephalus with slight or moderate ventricular dilatation, and rarely in marked hydrocephalus, the so-called compensated hydrocephalus. Pasquini et al. [8], using computer analysis, described the change of PVL in hydrocephalic patients before and after surgical treatment. Our method using computer

M. Asada et al.: Periventricular Lucency on CT Scan analysis o f PVL is also useful in displaying the degree and the nature of PVL objectively. PVL can be easily detected on the graph. The vascular component was not related to PVL because CT values in PVL were not significantly increased by intravenous contrast enhancement. The gradual increase of CT values o f PVL after injection o f metrizamide into the ventricles was suggestive of the influx o f metrizamide into PVL through the ventricular wall. CT scans made on experimental hydrocephalus showed that ventricular dilatation developed rapidly within 24 h and PVL was usually observed on the third or fourth day after inflation of the balloon. Milhorat et al. [1,4, 5] reported that ependymal disruption occurred in a few hours after inflation o f the balloon in the hydrocephalic monkey. It was most likely to occur at the superior lateral angle of the anterior horn, where PVL was demonstrated on the CT scan. We considered that PVL might be caused by acute extravasation of CSF into the white matter through the disrupted ventricular wall. Transependymal absorption o f CSF is unlikely to be the pathogenesis of PVL for the following reasons: 1. PVL was clinically observed in patients with hypertensive hydrocephalus, and rarely seen in patients with marked hydrocephalus. 2. The lucent zone is sometimes wider than the zone of transependymal absorption [3]. 3. Intraventricular metrizamide injection resulted in the influx o f metrizamide into the subependymal space. 4. PVL could be observed only at the acute stage in experimental study. It is hoped that experimental study may further elucidate the mechanism o f PVL:

211 Computer analysis gave the following information: objective display of PVL, change of CT values in PVL by contrast enhancement, effect o f shunt operation, and the influx o f metrizamide into the PVL during its intraventricular injection. We concluded that PVL might be caused by acute extravasation of CSF into the white matter through the disrupted ventricular wall.

References 1. Clark, R. G., Milhorat, T. H.: Experimental hydrocephalus. Part 3: Light mier0scopic findings in acute and subacute obstructive hydrocephalus in the monkey. J. Neurosurg. 32, 400-413 (1970) 2. Hopkins, L. N., Bakay, L., Kinkel, W. R., Grand, W.: Demonstration of transventricular CSF absorption by computerized tomography. Acta Neurochir. 39, 151-157 (1977) 3. Lux, W. E., Hochwald, G. M., Sahar, A., Ransohoff, J.: Periventricular water content. Effect of pressure in experimental chronic hydrocephalus. Arch. Neurol. 23, 475-479 (1970) 4. Milhorat, T. H.: Experimental hydrocephalus. Part 1: A technique for producing obstructive hydrocephalus in the monkey. J. Neurosurg. 32, 385-389 (1970) 5. Milhorat, T. H., Clark, R. G.; Hammock, M. K., MeGrath, P. P.: Structural, ultrastructural, and permeability changes in the ependyma and surrounding brain favoring equilibration in progressive hydrocephalus. Arch. Neurol. 22, 397-407 (1970) 6. Mori, K., Kurata, T.,"Nakano, Y., Handa, H.: Periventricular lucency in hydrocephalus on computerized tomography. Surg. Neurol. 8, 337-340 (1977) 7. Naidich, T. P., Epstein, F., Lin, J. P., Kricheff, I. I., Hochwald, G. M.: Evaluation of pediatric hydrocephalus by computed tomography. Radiology 119,337-345 (1976) 8. Pasquini, U., Bronzini, M., Gozzoli, E., Mancini, P., MenicheUi, F., Salvolini, U.: Periventricular hypodensity in hydrocephalus: A clinico-radiological and mathematical analysis using computed tomography. J. Comput. Assist. Tomogr. 1,443 to 448 (1977)

Condufion Periventricular lucency is the direct sign o f increased intracranial pressure and is often observed in obstructive hydrocephalus with mild or moderate ventricular dilatation.

M. Asada, MD Department of Neurosurgery Kobe University School of Medicine Kusonoki-cho 7-chome Ikuta-ku Kobe, Japan