Analysis of cerebral venous sinus stenosis by flat panel angiographic CT (FP-CT) to guide treatment for idiopathic intracranial hypertension

Abstract

Introduction

Cerebral venous sinus stenting (CVSS) is an effective treatment for idiopathic intracranial hypertension (IIH) secondary to dural venous sinus stenosis. Traditional selection of patients for CVSS has been made by microcatheter manometry, but pressure measurements are often equivocal. Here we present the results of a series of cases in which venous flat-panel CT (FP-CT) was used as an adjunct to microcatheter manometry to improve decision making and precise stent placement during CVSS.

Methods

Ten consecutive patients with IIH underwent angiography with microcatheter manometry and venous FP-CT, with CVSS if indicated by the results. Cross-sectional measurements of the narrowed sinus were obtained on FP-CT before and after stenting. After the procedure, clinical outcomes were tracked. Follow-up with quantitative MRA with sinus flow measurements was also performed, when available.

Results

There was an exponential correlation between measured pressure gradient and degree of stenosis calculated using venous FP-CT. All patients with both a high degree of stenosis measured by FP-CT and a high pressure gradient across the stenosis showed a clinical benefit from stenting.

Conclusions

True measurement of the cross-sectional area of the dural sinus, made by venous phase FP-CT, has a high degree of correlation with elevated venous pressure gradient across the point of stenosis. Even in a limited series of cases, we found an exponential decrease in flow with increasing severity of stenosis. Furthermore, patients with both an elevated venous pressure gradient and critical stenosis of the sinus on FP-CT showed symptomatic improvement after stenting.

Introduction

Cerebral venous sinus stenting (CVSS) is an effective treatment for idiopathic intracranial hypertension (IIH) accompanied by dural venous sinus stenosis. The etiology of elevated intracranial pressure in such cases remains controversial. Primary stenosis of a dominant dural sinus impedes cerebral venous drainage and increases intracranial pressure, while elevated intracranial pressure exerts an inward pressure on the dural sinuses and induces a secondary stenosis, which in turn impairs cerebral drainage further.^1–5 This positive feedback loop has been implicated in the onset and progression of the headaches, papilledema, and visual symptoms characteristic of IIH. Interruption of this pathway via CVSS has been shown to be a safe and effective treatment for IIH, with clinical response rates of greater than 80% reported.^6,7

However, patients with IIH secondary to dural venous sinus stenosis comprise only a subset of the overall IIH population. Furthermore, nearly all patients with IIH will show some degree of sinus narrowing on MRI or MRV imaging due to chronically elevated ICP. This stenosis is not necessarily the causative factor in the disease process, and likewise, its treatment may not reverse the clinical course of the disease.^8,9 It is therefore necessary to identify those patients who will benefit from CVSS, if for no other reason than to avoid dual antiplatelet therapy in patients who will need repeated lumbar punctures, CSF diversion, or optic nerve sheath fenestration.

It has therefore become common practice to perform dural venous sinus microcatheterization and direct manometry, with pressure gradient of >8 mmHg across the point of stenosis as an indication of a clinically significant stenosis that may benefit from stenting. Although greater than 90% of IIH patients may show sinus stenosis on imaging, as low as 30% to 40% may demonstrate a venous pressure gradient across the stenosis.^6,8 However, it is equally important to accurately identify the location and severity of the stenosis to perform effective stenting. Flat-panel CT (FP-CT) is an effective imaging modality to provide three dimensional, real-time feedback during neuroendovascular procedures. ¹⁰ Here, we describe our technique for 3D angiographic reconstruction of the dural sinuses using FP-CT and report our initial results using this technique to guide accurate and effective CVSS.

Methods

We retrospectively reviewed 10 consecutive cases in which FP-CT was used in conjunction with dural venous sinus pressure measurements to measure the degree of venous stenosis and plan endovascular intervention. All patients presented with suspected IIH including both papilledema on ophthalmological examination and elevated opening pressure on lumbar puncture, with MRA indicating likely stenosis of the dominant transverse sinus. All dural venous sinus pressure measurements were made in awake patients under moderate sedation, which is our protocol in all patients who are able to tolerate the procedure due to the alterations in central venous pressure that occur during general anesthesia. We measured the degree of stenosis in all cases both using conventional single-plane angiography (AP and lateral projections in the venous phase of ipsilateral internal carotid artery injection) and 3D-reconstructed venous FP-CT. In this technique (also called venous cone-beam CT), we perform an ipsilateral internal carotid artery injection and time the 24-second image acquisition to begin in the venous phase, generally 8 seconds after the injection. Contrast injection is performed at 4cc/second during the acquisition. Contrast is not diluted prior to injection, as is often done for arterial FP-CT, because the venous-phase acquisition naturally decreases the intensity of the contrast agent. In our experience this allows for the most complete visualization of the anatomy of the venous system, including the superior sagittal and transverse sinuses. This is in contrast to direct venous injection for single-plane or 3D venography, where only a focal portion of the venous system is investigated and, due to injection against flow and possibly against a point of stenosis, large variation in imaging results are often encountered.

On the single-plane images, the point of stenosis is estimated and the projection best displaying that location is used to measure stenosis, by percentage of the normal sinus caliber. On the 3D images, axial, coronal, and sagittal images are acquired and the projection showing the most severe point of stenosis is used. Cross-sectional area of the sinus and point of maximum stenosis are calculated from this data. Measured severity of stenosis, by cross-sectional area, is used along with microcatheter manometry of the dural sinus to guide the decision for stenting. After stenting, repeat FP-CT is performed to confirm that the point of stenosis had been adequately covered by the stent, and that the caliber of the sinus had increased accordingly. Figure 1 shows an illustrative case, including FP-CT images before and after stenting.

Figure 1.

Representative cases of FP-CT-guided CVSS. Pre-stent reconstructions are shown in panel A. Panel B shows FP-CT after stent placement at the right transverse-sigmoid junction. QMRA sinus flows before (C) and after (D) stenting are shown.

When available, subjects included in the study underwent quantitative magnetic resonance angiography (QMRA) with venous reconstruction for measurement of flow in the affected sinus (NOVA, Vassol, described elsewhere).^11–14 This was not used to guide treatment, but to evaluate for changes in flow after stenting and correlation with changes in symptoms. A representative case is included in Figure 1.

Results

Table 1 summarizes the consecutive patients included in the study. All subjects were adults between the ages of 25 and 50. Mean BMI was 38.7. Location of the dural sinus stenosis was at the transverse-sigmoid junction in all patients. Table 2 shows the subjects who underwent CVSS, including absolute change in sinus cross-sectional area after stenting and subsequent symptomatic response. When available, change in sinus flow measured by QMRA after stenting is also shown.

Table 1.

Demographics of consecutive subjects included in the study.

Subject demographics	n = 10
Age (mean)	36.9
Female (%)	100
BMI (mean)	38.7
Stent placed?
Yes	6
No	4
Presenting symptoms
Headache (n)	8
Vision loss (n)	9

Table 2.

Summary of clinical and radiographic outcomes in patients who underwent stenting.

Subject	Sinus CSA prior to stenting (mm2)	% stenosis by CSA prior to stenting	Absolute improvement in CSA after stenting (mm2)	Sinus flow (NOVA) before stenting (cc/min)	Sinus flow (NOVA) after stenting (cc/min)	Headaches after stenting	Visual symptoms after stenting
1	1.33	0.98	28.9	n/a	497	Resolved	Resolved
2	1.77	0.96	34.6	246	353	Improved	Resolved
3	0.20	0.99	28.1	n/a	403	Unchanged	Improved
4	0.28	0.99	26.1	66	353	Improved	Resolved
5	0.95	0.99	48.01	327	344	Resolved	Improved, mild photophobia
6	0.20	0.99	38.3	479	640	Improved	Improved

Figure 2 shows correlation between sinus stenosis, measured and calculated by the cross-sectional area of the sinus on FP-CT, and sinus pressure gradient measured by manometry. Patients who underwent stenting are marked. As the figure shows, patients with both a high dural sinus pressure gradient and a high degree of stenosis measured by FP-CT were most likely to have documented evidence of symptomatic improvement after CVSS.

Figure 2.

Relationship between dural sinus pressure gradient across the point of stenosis, made by microcatheter manometry, and degree of sinus stenosis measured by cross-sectional area of the sinus on FP-CT.

Discussion

Here we report our methods and initial results in using venous FP-CT as an important tool to guide decision making for CVSS in IIH. This method does not replace microcatheter manometry, which remains the mainstay of intra-procedural evaluation of CVSS. It should be noted, however, that manometry can provide equivocal results and stenting has been shown to be effective in carefully selected patients with low pressure gradients, as a vision-saving intervention. ¹⁵ It is for this reason that a better imaging modality is needed to truly evaluate a suspected point of stenosis. In some cases of low pressure gradients, FP-CT is useful to demonstrate the lack of a focal point of stenosis and perhaps to identify an arachnoid granulation as the suspected culprit, which may not be flow-limiting. In others, a true stenosis may be present that is not detected by manometry due to the numerous factors that can affect central venous pressure during the procedure (ETCO₂, patient sedation or pain level, inner diameter of the microcatheter used for manometry, and others). In cases that do have a significant pressure gradient, FP-CT is helpful to accurately identify the point of maximum stenosis (usually near the transverse-sigmoid junction) and be sure it is crossed by the stent, and then later to ensure that the stent has adequately opened the narrowed sinus.

In our initial series of cases, venous stenosis measured by FP-CT correlated in a nearly exponential fashion with the size of the measured pressure gradient. This provides a proof of principle and allows the identification of patients who are highly likely to benefit from stent placement. Although more data points are needed, all of our patients with a significant venous pressure gradient had over 80% stenosis by FP-CT measurement. Our data mirrors the well-known relationship between arterial stenoses and flow limitation, known at Spencer's law, which has long been applied to carotid artery disease and establishes the exponential decrease in flow once stenosis reaches 80% to 90%. ¹⁶ Even from our small sample size, this relationship becomes apparent when the true cross-sectional area of the sinus is measured by FP-CT and correlated with the measured pressure gradient. Furthermore, in those patients with QMRA, high-grade stenosis was correlated with lower flow within the sinus. In all 4 cases, flow increased on QMRA after stenting.

Perhaps most importantly, those patients in the group with both high-grade stenosis and a high pressure gradient (greater than 10 mmHg) that underwent stenting all reported improvement in headaches and visual symptoms. Of the two patients with high-grade stenosis but a lower pressure gradient that underwent stenting, one reported persistent headaches but improved visual symptoms. Although this sample size is quite small, the data in Figure 2 suggest that the combination of a high pressure gradient and high-grade stenosis measured by FP-CT portends a symptomatic response to stent placement. Whereas single-plane angiography in the venous phase often provides imprecise images of the dural sinuses and makes true measurement of a symptomatic stenosis difficult, our method is a simple way to ensure that the correct treatment is chosen and that the exact point of stenosis can be pinpointed for stenting.

Lasty, within our series we identified three patients with a high percentage of stenosis on FP-CT but moderate or low pressure gradient on manometry. We do not feel that these individual results reflect hypoplastic sinuses, as our procedures targeted the dominant sinus in all cases. Rather, the focal stenoses found in these patients are more likely large arachnoid granulations around which venous flow is preserved. 3D imaging is extremely helpful to identify these cases, which are unlikely to respond to stenting.

Conclusion

Treatment of IIH with CVSS is effective, but patient selection is difficult.^15,17–22 As an adjunct to microcatheter manometry, FP-CT angiography is a useful means to estimate the degree of stenosis and guide stenting. In our early experience using this combined technique, patients with severe stenosis by FP-CT in addition to elevated venous pressure gradient by manometry were likely to show immediate clinical response.