T HE coexistence of hydrocephalus and

Transcription

1 J. Neurosurg. / Volume 31 / November, 1969 Venous Factors in Infantile Hydrocephalus* HORAE NORRELL, M.D., HARLES WILSON, M.D., JOHN HOWIESON, M.D., LOYD MEGISON, M.D., AND VURAL BERTAN, M.D. Division o] Neurosurgery and Department o] Radiology, University o] Kentucky School of Medicine, Lexington, Kentucky T HE coexistence of hydrocephalus and elevated sagittal sinus venous pressure has been observed in clinical, 7,9,12 and laboratory studies. :~,' ~ However, the basis for venous sinus hypertension and its etiologic relationship to hydrocephalus lack satisfactory explanation. The present study was undertaken to correlate sagittal sinus venous pressure (SSVP) and cerebrospinal fluid pressure (SFP) in a series of hydrocephalic infants, and to investigate the possibility that elevated sagittal sinus pressure might have an anatomical basis. Materials and Methods We studied 30 infants, ranging in age from 8 days to 25 months, who had a provisional diagnosis of hydrocephalus; all subsequently were proven to have hydrocephalus. Eleven of the group had myelomeningoceles, all repaired earlier, and the remainder had no extracranial abnormalities. With the child lying supine and under pentobarbital sedation, bilateral ventricular punctures were performed at the lateral angles of the anterior fontanel with 20-gauge needles. The superior sagittal sinus was cannulated in the direction of the blood flow at the posterior angle to the anterior fontanel using a 20-gauge scalp vein needle. Patency was maintained by intermittent irrigation with a saline solution containing heparin (1 unit per ml). All needles were fixed in position with sterile gauze impregnated with flexible collodion. The right ventricular needle was connected to one arm of a Statham P23 differential pressure transducer, and the sagittal sinus cannula was connected to the opposite arm with the transducers calibrated to zero pressure at the estimated level of the right atrium. All recordings were made on a 4-channel Sanborn Received for publication November 7, Revision received April 14, * Supported by Easter Seal Research Foundation. 561 recorder. A second ventricular needle was connected by flexible tubing to a 50 cc saline-filled syringe mounted in a Harvard infusion pump. In the first 14 cases, saline was infused into the spinal subarachnoid space; in the remaining 16 cases, infusions were made directly into the left lateral ventricle. Base-fine ventricular cerebrospinal fluid pressure (SFP) and sagittal sinus venous pressure (SSVP) were recorded. Unavoidable loss of cerebrospinal fluid invalidated baseline in a few cases, although estimated loss was immediately replaced with saline. Saline was then infused into the left ventricle at a rate sufficient to elevate the SFP to mm water pressure within 2 minutes, an arbitrary pressure-time standard selected early in the study. During the saline infusion SFP and SSVP were recorded constantly. Sagittal sinus venography was performed when the pressure studies were completed. oincident with the completion of rapid (approximately 2 sec) hand injection of 4 cc of 60% meglumine iothalamate into the sagittal sinus cannula, five serial radiographs were taken at a rate of one exposure every 2 seconds with ventricular pressure at mm of water, and immediately repeated after elevating SFP to mm of water. ontrol of SFP was achieved by infusing saline into the lateral ventricle with a Harvard infusion pump. Sagittal sinus emptying time refers to the time lapse from completion of contrast injection until all contrast material has disappeared from the intracranial sinuses. Sinograms were followed by ventriculography in each child. Results Of the 30 infants, 20 had communicating and 10 noncommunicating hydrocephalus. Baseline ventricular pressure exceeded sinus pressure in 12 infants, whereas the pressures were equal in 10 cases, and the relations:hip of pressures reversed (SSVP > SFP) in

2 562 Norrell et al. TABLE 1 Pressure studies in hydrocephalic children* ase No. Diagnosis SFP Baseline Pressure (mmh20) SSVP Pressure After Saline Infusion (mm H~O) SFP SSVP O (M) (M) (M) (M) 0 (M) O (M) O (M) O 0 (M) O (M) O (M) O O (M) O O O0 60O 6O0 6O0 60O O 30O 3OO 56O 49O 65O 25O * Abbreviations: --ommunicating O--Obstructive (M)--Accompanying Myelomeningocele SFP-- erebrospinal fluid pressure SSVP--Sagittal sinus venous pressure eight cases (Tables 1 and 2). With SF pressure elevated to mm, ventricular pressure exceeded sinus pressure in 24 infants, the pressures were equal in four cases and reversed in only two cases. Baseline ventricular pressure exceeded mm water pressure in 13 of the 30 cases studied. Baseline sinus pressure was greater than mm water pressure in eight patients, whereas after elevation of ventricular pressure to mm, sinus pressure exceeded mm in 19 patients (Table 1). Intraventricular or intrathecal saline infusion produced a parallel but not necessarily identical rise between ventricular and sinus pressure in 12 of the 30 patients (Table 3, and Fig. 1 ). In the group not demonstrating a parallel pressure rise, the mean sinus pres-

3 Venous Factors in Hydrocephalus 563 TABLE 2 TABLE 3 Relationship between SFP and SS VP before and hanges between SFP and SS VP in response to after saline infusion in 30 infants intrat;entricular saline infusion in 30 infants Pressure Ratio Baseline Post-Infusion* SSVP > SFP 8 2 SSVP = SFP 10 4 SSVP < SFP Total * SFP = mm water sure remained constant or changed only moderately during the rise of SF pressure (Fig. 2). In all cases, the intraventricular and sagittal sinus pulse wave amplitude increased progressively during the elevation of ventricular pressure. The increasing amplitude of the sagittal sinus pulse wave occurred independent of any changes in mean sinus pressure. Sagittal sinograms adequate for interpretation were obtained in 28 infants. Both lateral sinuses filled in 18 instances, and only one lateral sinus filled in the remainder. Among 10 cases with filling of only one lateral sinus, the right side filled in seven and the left side in three instances. All sinograms in hydrocephalic patients without accompanying mye- Parallel Non-Parallel Myelomeningocele Pressure Rise Pressure Rise Present 10 1 Absent 2 17 Total lomeningoceles demonstrated a normal lateral sinus configuration. The torcular Herophili was located in a midline position at the internal occipital protuberance on the concave aspect of the squamous portion of the occipital bone, approximately half-way between the lambda and the posterior lip of the foramen magnum. The lateral sinus followed a horizontal course with a slight anterosuperior convexity from the torcular to the junction of the petrous and mastoid portions of the temporal bone (Fig. 3). The sigmoid sinus and jugular bulb were normally placed and in only one case was there radiographic evidence of constriction of the lateral sinus. The lateral sinograms of all patients (11 ) with myelomeningoceles demonstrated a Fro. 1. The upper tracing represents SFP while the lower tracing shows SSVP; both curves show a parallel rise in response to intraventricular saline infusion.

4 564 NorreU et al. Fro. 2. Mean SSVP does not change in response to saline infusion, although the amplitude of the pulse waves within sagittal sinus and SF increase. low-lying torcular Herophili at the posterior extremity of the foramen magnum with the lateral sinuses taking a direct course to the jugular foramen along the flattened base of the skull. The lateral sinus failed to reach the lateral aspect of the petrous bone and the vertical portion of the sigmoid sinus was entirely lacking. On the half-axial view the lateral sinuses coursed around the foramen magnum in a basal position (Fig. 4). The normal upward convexity of the transverse portion of the lateral sinus was either lacking Fro. 3. Left: Lateral sinogram in an infant with communicating hydrocephalus with a normal sinus configuration. Right: Towne projection showing the wide sweep of the normal lateral sinuses reaching the mastoid portion of the temporal bone.

5 Venous Factors or replaced by a concavity, suggesting that the sinus lay close by the foramen magnum. The horizontal portion of the sigmoid sinus was deformed and shortened, reflecting the failure of the lateral sinus to reach the lateral extremity of the petrous pyramid (Fig. 5 left). There were four cases of myelomeningoceles among the 12 patients having sinograms performed with the SFP at mm and repeated at mm of the water (Table 4). All demonstrated parallel rises between SF and sinus pressure and when the SFP was elevated to mm of water in the myelomeningocele group, all sinograms demonstrated partial or complete obliteration of the lateral sinuses never observed when the sinogram was performed with SFP at mm. Associated with basilar venous sinus obstruction, there was opacification of the parietal, occipital, and mastoid emissary veins, indicating collateral routes of venous drainage (Fig. 5 right). In no case was collapse of the superior sagittal sinus observed when intracranial pressure was artificially increased. No occlusion of the sinuses developed in the sinograms in infants without myelomeningocele when the intracranial pressure was elevated. FIG. 4. Sinus pattern seen in association with the Arnold-hiari malformation shows the lowlying torcular and lateral sinuses coursing near the foramen magnum. in Hydrocephalus 565 TABLE 4 Relationship between basefine SFP and SS VP in myelomeningoeele and non-myelomeningoeele patients Pressure Ratio Myelomenin- Non-Myelogocele meningocele SSVP > SFP SSVP=SFP SSVP < SFP Total I Segmental obstruction of the superior sagittal sinus accompanied sinus hypertension in one case (ase 23) producing filling of extensive cortical venous channels leading to delayed opacification of the lateral sinus. Another case with venous hypertension showed extensive parietal emissary veins, with fixed stenosis of one lateral sinus and non-filling of the opposite sinus (Fig. 6). Sagittal sinus emptying time was determined in 13 patients, four having myelomeningoceles (Table 5). With intraventricular pressure at mm water pressure, the sagittal sinus emptying time was 2 seconds or less in 10 cases (two with myelomeningocele). Sagittal sinus emptying time was delayed to 6 seconds or longer in the remaining three cases. With SFP elevated to mm water pressure, sagittal sinus emptying time was 6 seconds or longer in five cases. Increased intracranial pressure producing lateral sinus obstruction resulted in delayed sinus emptying time in each of the four patients with a myelomeningocele. However, no change in emptying time was seen in the nine patients without a myelomeningocele. Subdural extravasation of contrast material occurred in five cases with a major extravasation in only one. No ill effects were noted as a result of extravasation and no other complications were encountered. Discussion Normally the pressure of the cerebrospinal fluid is slightly higher than that of the sagittal sinus venous pressure. 15 Variable resuits have been reported equating sinus pressure to ventricular pressure in normal animals. Weed and Flexner reported no change in sinus pressure accompanying abrupt elevation of the ventricular pressure, 1~ whereas 19

6 566 Norrell et al. FIG. 5. Left: Basal displacement of the lateral sinus occurring with the Arnold-hiari malformation. The lateral sinus is patent with a SFP of ram. Right:Following elevation of SFP to mm there is basilar obstruction of the sinus with filling of collateral venous channels. Wright 17 reported an early, limited rise in sinus pressure with a rapid fall to a lower level during the elevation of intracranial pressure to near systolic arterial blood pressure levels. Bedford ~ observed that rapid increases in subarachnoid pressure are invariably accompanied by a fall in sinus pressure. No complications in our series were attributable to saline infusion or elevation of the intracranial pressure. Foldes and A r r o w o o d G demonstrated the innocuous effects of elevating the ventricular pressure to m m water pressure in humans by infusing saline into the spinal subarachnoid space. Maintenance of the pressure at this level for 10 to 20 minutes produced mild symptoms in about one-half of their patients. Shulman and Ransohoff 1~ showed that disregarding the etiology of the hydrocephalus, hydrocephalic infants frequently had an elevation of the sagittal sinus venous pressure (SSVP ---- or > S E P ). In our series the sinus pressure was elevated in only 18 of the 30 cases studied. Elevated sinus pressure oc- FiG. 6. Towne and lateral projection of the sagittal sinogram showing stenosis of the right lateral sinus and non-opacification of the left lateral sinus.

7 Venous Factors in Hydrocephalus 567 TABLE 5 Sagittal sinus emptying time in 13 patients SFP mm SFP mm Emptying Time 2 scc 6 sec 2 sec 6 scc Myelomeningocele Patients Non-Myelomeningocele Patients Total Patients curred in only nine of the 19 non-myelomeningocele patients in the group, whereas in the patients with myelomeningoceles, elevated sinus pressure was a frequent finding. In further considering the non-myelomeningocele group we found that none of the patients with aqueductal atresia exhibited elevated sinus pressure, but of the 14 cases with communicating hydrocephalus, nine had sinus hypertension (Table 1). Shulman's pressure recordings suggested that the SF and sinus pressures were causally related in hydrocephalic infants and, basing his calculations upon the pressure regression slope in a single infant, he demonstrated that sinus pressure was directly dependent upon SF pressure. We observed a pressure dependency between ventricular and sinus pressure in only 12 of 30 cases, manifest as a parallel rise between SFP and SSVP in response to saline infusion; 10 of the 12 children had previously repaired myelomeningoceles (Table 3). In the two remaining children without myelomeningocele the parallel rise is explained by technical errors. In one (ase 15) the parallel rise occurred following misplacement of the sagittal sinus needle into a cerebral cortical vein, a structure directly reflecting intracranial pressure. 1" The only other case suggesting a parallel rise was considered technically unsatisfactory. Our initial protocol did not provide for the performance of sinograms with SF pressure at mm and at mm. Based upon the finding of parallel pressure rises between ventricular and sinus pressures in children with hydrocephalus and myelomeningocele early in the series, we predicted a dynamic obstruction of the sinuses related to SF pressure, causing us to repeat the sinograms with SF pressure elevated to mm. We also predicted that the sinus obstruction would be incomplete or that collateral pathways of venous drainage must exist since venous pressure did not invariably exceed mm with the SF pressure elevated to mm. Only one child with a myelomeningocele failed to exhibit a parallel rise in SFP and SSVP. The sagittal sinogram in this patient was not typical of the usual pattern seen in the children with myelomeningoceles since the lateral sinus on one side occupied a normal position. Shulman and Ransohoff 1~ demonstrated narrowing of the lateral sinus in one case, predicted that the obstruction was pressure dependent, but failed to show this. No mention was made of whether or not this child had a myelomeningocele. Kinal 9 observed sagittal sinus venous hypertension in four hydrocephalic infants and attributed this to constrictions of the lateral sinus and jugular bulb, but made no mention of the dynamic component of the obstruction. The significance of Kinal's findings has not been confirmed by subsequent investigators. Anatomical stenosis of both lateral sinuses or jugular bulbs appears to be an uncommon finding associated with hydrocephalus, since we demonstrated the abnormality only once (ase 6) in 30 sinograms (Fig. 6). Apparent narrowing of the sigmoid sinus, as illustrated in Kinal's report, is a frequent finding, but when visualized in two planes, true stricture is absent. Failure of one lateral sinus to fill during sagittal sinography cannot be interpreted as an abnormal finding. The anatomical study of Woodhal116 indicated that the superior sagittal sinus drains exclusively into one lateral sinus in approximately 35 % of cases. Most pathological accounts of the Arnold-hiari malformation, a deformity al-

8 568 Norrell et al. Fro. 7. Photograph of base of skull (Arnold- hiari malformation). The tentorium cerebelli is abnormally attached to the petrous ridges producing a small posterior fossa. The lateral sinuses are not seen, but the right jugular foramen is immediately adjacent to the foramen magnum. most invariably coexisting with myelomeningocele, al contain little information concerning accompanying abnormalities of the bony posterior fossa, s tentorium, and dural venous sinuses. 4 In a detailed anatomical study, Daniel and Strich ~ described the small posterior fossa produced by an anomalously low attachment of the tentorium cerebelli. As a consequence of inferior and medial displacement of the tentorial attachment, the lateral sinuses course along the floor of the posterior fossa near the foramen magnum (Fig. 7). The sinus occupies a position near the foramen magnum as it follows an abnormal, low and straight course to the jugular foramen. The sinuses conform to the radiographic configuration of the normal occipital sinus leading to the speculation that the true lateral sinuses may be entirely absent. However, the position of the sinuses within the tentorial attachment adds strength to the belief that they are truly misplaced lateral sinuses. With rising intracranial pressure, this structural deformity predisposes to sinus collapse or kinking from the extrinsic pressure exerted by the crowded neural tissue within the small posterior fossa. This hypothesis is supported by our findings of delayed emptying of the sinuses, sinus occlusion, and opacification of collateral channels in sino- grams performed during the period of raised intracranial pressure. The dynamic sinus obstruction accompanying the Arnold-hiari malformation may aggravate the severity of hydrocephalus caused by other anatomical abnormalities in this disorder, and the development of adequate collateral venous channels may account for the late spontaneous arrest of hydrocephalus in some children with the Arnold-hiari malformation. In our study, increasing intracranial pressure was invariably accompanied by an increasing amplitude of the pulse wave within the sagittal sinus. The walls of the triangular sagittal sinus are abutted on two sides by brain and cerebrospinal fluid. The flexible walls of the dural sinus are capable of transmitting pulsations from the external environment, and we have demonstrated that the sagittal sinus does not collapse. The pulse wave generated by choroid plexus and intracranial arteries is dampened by a combination of volume-controlling variables such as emptying of the cortical veins, distention of the dura in the region of the fontanels or the mobile sutures of infants. As the intracranial pressure rises, the dampening capacity of these mechanisms is reduced, reflected by larger pulse waves within the cerebrospihal fluid; the intracranial pulse waves in turn are transmitted to the column of blood within the sagittal sinus. The intracranial pressure wave transmitted to the column of blood within the sagittal sinus is negligible when the intracranial pressure is normal or only slightly elevated, but as intracranial pressure increases the pulse wave amplitude increases. In the cases not demonstrating pressure-dependent lateral sinus occlusion, the pulse wave amplitude within the sagittal sinus remained smaller than that of the cerebrospinal fluid. We postulate that the pulse wave within the sinus was dampened by the venting of the blood column through the jugular foramen. In the case demonstrating a parallel rise between sinus and ventricular pressure with gradual occlusion of the lateral sinuses there was a progressive and identical rise in the amplitude of the cerebrospinal fluid and sagittal sinus pulse waves as intracranial pressure increased (Fig. 1 ) indicating that the extracranial venting source was lost and the pulse waves remained identical.

9 Summary Sagittal sinus venous hypertension does not invariably accompany infantile hydrocephalus. A pressure dependency between cerebrospinal fluid pressure (SFP) and sagittal sinus venous pressure (SSVP) is rare except in cases with myelomeningocele. We predicted a dynamic venous sinus obstruction in children with myelomeningocele causing the parallel rise between SFP and SSVP in response to the intrathecal infusion of saline, and showed a pressure dependent point of obtruction to be at the level of the abnormally positioned lateral sinus accompanying the Arnold-hiari malformation. References 1. BEDFORD, T. H.B. The effect of variations in the subarachnoid pressure on the venous pressure in the superior longitudinal sinus and in the torcular of the dog. J. Physiol, London, 1942, 101: BERING, E. A., JR. horoid plexus and arterial pulsation of cerebrospinal fluid. Archs Neurol. Psychiat., hicago, 1955, 73: BERING, E. A., and SALIBI, B. Production of hydrocephalus by increased cephalic-venous pressure. Archs Nearol. Psychiat., hicago, 1959, 81: AMERON, A. H. The Arnold-hiari and other neuroanatomical malformations associated with spina bifida. J. Path. Bact., 1957, 73: DANIEL, P. M., and STmH, S. J. Some observations on the congenital deformity of the central nervous system known as the Arnold- hiari malformation. J. Neuropath. exp. Neurol., 1958, 17: Venous Factors in Hydrocephalus FOLDES. F. F., and ARROWOOD, J. (;. hanges in cerebrospinal fluid pressure under the influence of continuous subarachnoidal infusion of normal saline. J. clin. Invest., 1948, 27: HOOPER, R. Hydrocephalus and obstruction of the superior vena cava in infancy. linical study of the relationship between cerebrospinal fluid pressure and venous pressure. Pediatrics, Springfield, 1961, 28: KRUYEE, E., and JEFFS, R. Skull abnormalities associated with the Arnold-hiari malformation. Acta radiol., 1966, 5: KINAE, M. E. Hydrocephalus and the dural venous sinuses. J. Neurosurg., 1962, 19: NOLLE, W., and SHNEIDER, M. Zur Hamodynainik der Gehirndurchblutung bei Liquordrucksteigerung. Arch. Psychiat. Nervenkr., 1948, 180: i 1. RUSSELL, D. A., and DONALD,. Mechanisms of internal hydrocephalus in spina bifida. Brain, 1935, 58: SHULMAN, K., and RANSOHOFF, J. Saginal sinus venous pressure in hydrocephalus. J. Nearosurg., 1965, 23: SHULMAN, K., YARNELL, P., and RANSOHOFF, J. Dural sinus pressure in normal and hydrocephalic dogs. Archs Neurol., hicago, 1964, 10: WEED, L. H., and FLEXNER, L. B. The relations of the intracranial pressures. Am. J. Physiol., 1933, 105: WEED, L. H., and HUGHSON, W. Intracranial venous pressure and cerebrospinal fluid pressure as affected by the intravenous injection of solutions of various concentrations. Am. J. Physiol., 1921, 58: WOODHALL, B. Anatomy of the cranial blood sinuses with particular reference to the lateral. Laryngoscope, St. Louis, 1939, 49: WRIGHT, R.D. Experimental observations on increased intracranial pressure. A ust. N.Z.J. Surg., 1938, 7: