The initial portion of the study was undertaken using T2 images. However, the T2 images were angled to the skull base and are thus not exactly perpendicular to the sinus. Unlike the T2 images, the 3D T1 images are able to be reconstructed perpendicular to the sinus. Measurements of both the T2 and post contrast T1 images in MS patients were performed to test for possible inaccuracies due to distortion in the T2 imaging and to gauge the inter observer reliability. There was no significant difference between the two methods with good inter observer reliability. The cross-sectional area in the controls i.e. 42.1 mm
2 compares favourably with recently published reference data showing an area of 43 mm
2 three cm from the torcular [
6]. The shape of the sinuses was tested by calculating the circularity. Circularity is defined by the formula
\(4\uppi{\text{A}}/{\text{L}}^{2}\) where A is the cross-sectional area in mm
2 and L the circumference in mm [
7]. It can be seen that a circle of radius R would return a circularity value of 1. An equilateral triangle of side length 1 unit would return a value of 0.6. We can see from Table
1 that the shape of the sinuses of all groups was midway between that of a circle and triangle and not significantly different to each other. The sagittal sinus circularity has been previously published in control subjects and is 0.77 [
7], which is not significantly different to the current study.
The cerebral sinuses lie between the fibrous dura mater and the endosteum. The sagittal sinus consists of a venous channel passing through a split in the dura as it passes from the falx cerebri to the skull [
8]. The dura at the base of the sinus is attached to the endosteum of the skull and is fixed. The other two walls of the sinus are attached to the falx cerebri and are relatively fixed at this point. Between the three fixed vertices, the two walls separating the subarachnoid space from the sinus lumen are free to move. In the transverse sinuses the free walls have been noted to be concave, straight or convex [
4]. The dural walls of the sagittal sinus are made up of collagen fibres interspersed with fibroblasts and elastin in a ground matrix [
9]. Therefore, the walls are expected to have viscoelastic properties [
10]. In human dura, small changes in stress (the applied force) directed along dura strips produces corresponding small changes in strain (the percentage change in length), in a linear relationship which follows Hooke’s law [
11]. The stretch is recoverable in an elastic fashion provided the elastic limit is not reached. The gradient of the stress vs strain graph is called the Young’s modulus [
11]. The Young’s modulus for human dura adjacent to the sagittal sinus in adults was measured at 61 MPa with the dura remaining elastic up to a 20% stretch [
10]. Therefore, the amount of stretch induced in the free walls of the sinus is expected to be proportional to the applied transmural pressure gradient provided the tensile strength of the wall remains a constant. Therefore a large transmural pressure should give more stretch and a lower pressure less stretch. However, the area change produced in the vessel by a change in wall length may not be a linear function because of the triangular shape and constraint of the sinus.
Correlation between transmural pressure and sinus size
The normal transmural venous pressure from CSF to sinus lumen is 4 mmHg [
12]. In hydrocephalus patients the CSF outflow resistance is said to be elevated [
2]. Therefore, the expected transmural pressure in chronic hydrocephalus should be higher than normal. It follows the wall deflection in hydrocephalus should be larger and cross-sectional area lower than in the control group. From Table
1 we can see that the cross-sectional area of the sinus in hydrocephalus is 25% smaller than the matched controls.
In idiopathic intracranial hypertension there is an elevation in ICP above 18 mmHg, which is not related to an intracranial mass, meningeal process or cerebral thrombosis [
13]. Pickard et al. simultaneously measured the ICP and sagittal sinus pressure in 9 patients with IIH and found the transmural sinus pressure to be 1.8 mmHg [
14]. King et al. simultaneously measured the CSF pressure at C2 and the SSS pressure in 21 IIH patients and found a gradient of 5 mmHg [
15]. These studies are somewhat divergent but added together suggest the transmural pressure may be normal in IIH. In the current study there was no significant difference in sinus size, perhaps correlating with the normal pressure gradient. Similarly, Rohr et al. found no significant difference in the cross-sectional area of the sagittal sinus measured from T2 images 1 cm above the Torcular when comparing IIH with controls [
3].
Spontaneous intracranial hypotension (SIH) patients present with positional headaches which are worse in the upright position. The cause is typically a CSF leakage, most frequently within the spinal canal. An opening pressure of less than 4.4 mmHg is diagnostic for SIH [
16]. In SIH there is enlargement of the venous sinuses [
4] and also dural thickening [
17]. In the current cohort, an overall 22% increase in the size of the sinus compared to normal was found. Extensive dural thickening was also found on other imaging (not shown). Given the normal sinus pressure is 7.5 mmHg at age 45 years [
2], an opening pressure of 4.4 mmHg would give a reversed pressure gradient accounting for the venous dilatation.
The cross-sectional area of the sinuses in MS is 16% larger than the controls. This is unlike either hydrocephalus or IIH but closest to SIH. The increase in size could favour a decrease in transmural pressure, like SIH. However, this would need to be due to an increase in venous pressure rather than a decrease in ICP. The notion that there is an elevation in venous pressure in MS has been extensively investigated, reviewed and disputed [
18‐
20]. Therefore the cause for the dilatation is undefined and awaits further study.
The dilatation of the sinus walls in MS is similar to the findings of a whole brain 3T susceptibility-weighted study into MS which showed that although the intralesional parenchymal veins were 24% smaller, the extralesional veins in the normal appearing white matter were 20% larger by diameter [
21]. This raises another paradox. The compliance of the venous system in MS was noted to be reduced by half when compared to controls [
1]. Compliance is the change in volume divided by the change in pressure which brings this about or
\({\text{C}}=\Delta{\text{V}}/\Delta{\text{P}}\). Note the change in volume is in absolute terms not relative. If the change in pressure across the venous wall and the tensile strength of the veins were constant, a much larger venous volume in MS should bring about a much larger
\(\Delta{\text{V}}\). Therefore, the current findings suggest the venous compliance in MS should be larger not smaller. One possible solution to this problem would be if the wall strength were much larger than normal. Coen et al. noted that a change in the type 1 collagen component of the jugular veins occurs in MS perhaps [
22] correlating with a possible change in wall strength but this would require further enquiry to ascertain.
Limitations
The estimation of the cross-sectional area is likely to be more accurate using the post contrast images compared to the T2 images. Unfortunately, post contrast imaging was only available in the MS patients. The wall of the sinus being fibrous is of low signal on T2, similar to the flowing blood, so it is possible part of the wall may be added to the lumen in T2 based cross-sectional area studies. Despite this, the difference between both techniques was found to be negligible. In addition, the plane of the T2 studies is not exactly perpendicular to the direction of flow of the sagittal sinus. A review of all cases showed the average angle of the T2 slices directed away from perpendicular was 8.8° ± 6.2° which would affect the measurement of the antero-posterior dimension of the sinus by the cosine of the angle giving a 1.2% error. The transverse dimension would be unaffected.