Endovascular treatment of idiopathic intracranial hypertension

Idiopathic intracranial hypertension (IIH) is a rare disorder, nine times more common in women than in men, in which the incidence in both adolescents and adults seems to be proportionate with the prevalence of obesity. IIH is defined as the syndrome of raised intracranial pressure without clinical, laboratory, or radiologic evidence of intracranial pathology. In particular, a lack of venous obstructive disease is emphasized in the criteria of Dandy and in the International Headache Society classification.¹ The classic treatment is symptomatic, mostly focusing on normalization of intracranial pressure. The major morbidity of the disease is blindness or permanent visual impairment caused by prolonged papilledema with secondary optic atrophy.² Acetazolamide is often an efficient treatment and the most commonly used.³ However, the side effects of acetazolamide, such as paresthesias, taste perversion, somnolence, and depression, often limit its use.⁴ Surgical intervention could be considered as soon as medical treatment fails because there are no clear guidelines in the management of IIH.⁵ A common management error is to prolong delay before recommending surgery. Corbett emphasized that there is no acceptable level of visual field or acuity loss that should be awaited.⁶ A persistent visual loss despite optimum medical therapy is sufficient to recommend a more aggressive therapeutic option. For patients with failing medical therapy, two procedures are considered: optic nerve sheath fenestration, performed if visual loss is the main morbidity; and shunting procedures, performed if headache is the main symptom. Both medical and surgical treatments are insufficient and complicated by side effects.⁷ The recent review by Lueck et al. underlines that there is insufficient information to generate an evidence-based management strategy for IIH.⁵ The significantly higher role of venous outflow obstruction as an etiologic factor of IIH has been previously suggested on manometry and neuroradiologic findings and has led to therapeutic implications. Transverse sinus (TS) stent placement seems to represent a promising therapeutic alternative, but its long-term safety and efficacy have yet to be proven.^8,9

We report the results of intracranial venous sinus stenting in 10 consecutive patients with refractory IIH.

METHODS

Ten patients, eight women and two men, with a mean age of 41.8 years (range 28 to 60 years), were referred to the Neurosurgical Department of the Timone Hospital with a diagnosis of refractory IIH. Patient informed consent was obtained in all cases. Because the technique and the principle of the treatment of IIH using venous stenting have been previously described,⁹ the procedures were approved by the local ethics committee advisory board.

All patients had criteria for IIH and failed medical therapy (acetazolamide). All patients undertook paracetamol and/or nonsteroidal anti-inflammatory drugs for headaches. Medical treatment failure was defined as persistent visual symptoms or persistent headaches in spite of treatment with acetazolamide. The duration of the medical treatment depended on the severity of visual symptoms and the intensity of headaches. A clinical examination, including funduscopic examination, measurement of best corrected visual acuity, and Goldmann perimetry, was performed in all patients. Optical coherence tomography was performed only in three recent patients (data not shown). The diagnosis of sinus abnormalities was performed with three-dimensional rotational gadolinium-enhanced magnetic resonance (MR) venography (figure 1). In our early experience, we performed conventional cerebral angiography before the direct retrograde cerebral venography (DRCV) (figure 2A). CSF and TS pressures were recorded during DRCV, before stent placement (figure 2B). DRCV and manometry were performed while the patient was awake because induction of general anesthesia causes artificially elevated venous sinus pressures.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Magnetic resonance venography showing a bilateral venous sinus stenosis predominating on the left side

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Direct retrograde cerebral venography

(A) Anteroposterior cerebral angiography at the late venous phase showing a right lateral sinus stenosis (double arrows). (B) Anteroposterior subtracted venogram showing the site of stenosis via a direct retrograde cerebral venography (DRCV) in the same patient. (C) Stent deployment and balloon angioplasty (double arrows) during a DRCV procedure. (D) Anteroposterior subtracted venogram showing the stent after deployment with normal flow through the lateral sinus (double arrows). Pressure gradient is no longer detected by manometry.

Stent placement was performed under general anesthesia. A guide catheter was directed into the TS during the retrograde cerebral venography from a femoral vein puncture and the self-expanding stent deployed with or without associated balloon angioplasty (figure 2, C and D). This technique was the same as used by our team for treatment of dural arteriovenous fistulas involving the transverse and sigmoid sinus using a self-expanding stent with balloon angioplasty.¹⁰ An antiplatelet therapy (clopidogrel) was given 3 days before endovascular stent placement. The endovascular procedure was performed under heparin and antiplatelet therapy. After TS stenting, the patient was monitored for one night in the neurosurgical intensive care unit and discharged after 48 hours. The treatment with clopidogrel was continued for 3 months, the time of presumed endothelialization of the stent. In nine patients, only one side was treated. For one patient, a second stent was placed in the contralateral lateral sinus in a subsequent procedure. Follow-up venography and manometry were undertaken and CSF pressures monitored at 3 months. The mean follow-up was 17 months (range 6 to 36 months).

RESULTS

Patient characteristics, treatment, and outcome are summarized in tables 1 and 2. All patients had intractable headaches and visual disturbances. All patients had papilledema. Eight of 10 patients had visual acuity impairment, 5 of 10 patients had pulsatile tinnitus, and none had previous neurosurgical treatment. Duration of symptoms ranged from 4 to 46 months (mean 17.4 months). Body mass index values ranged from 22 to 37 kg/m² and are summarized in table 1. Only two patients were obese, and none of them lost weight during the follow-up period.

CSF pressures were high, ranging from 31 to 59 cm H₂O (mean 40.2 cm H₂O). Venous sinus anatomy analysis revealed left TS hypoplasia in six cases with contralateral sinus stenosis. In two cases with bilateral cerebral venous sinus drainage, the more significant stenosis involved the right TS. In the two last cases with bilateral drainage, stenosis was predominant on the left TS. Among our 10 patients, 6 had focal stenosis and 4 had more homogenous sinus compression involving all or at least two-thirds of the TS length. The venous pressures ranged from 20 to 44 mm Hg (mean 29.2 mm Hg) in the proximal prestenotic TS and from 5 to 14 mm Hg (mean 10.1 mm Hg) in the distal poststenotic TS. The pressure gradient ranged from 14 to 34 mm Hg (mean 19.1 mm Hg).

There was no serious complication after stenting. Seven of the 10 patients experienced transient headache lateralized to the treated side, which resolved with paracetamol or nonsteroidal anti-inflammatory drug.

After stenting, six patients had no headache, two were improved, and two were unchanged. These two patients had chronic tension-type headaches, and one of these patients, who developed a reactive depression, was under antidepressant therapy that may be a factor in chronic headache. All patients had documented papilledema. After stenting, papilledema resolved for all patients. In two patients, an optic atrophy was noted. At the last follow-up, visual acuity was normal for six patients (normalized for four), improved for three, and unchanged for one. The patient who did not show visual improvement had both severe and long-standing symptoms (>3 years). Pulsatile tinnitus resolved for all patients. In all patients, there was a normalization of CSF and sinus pressures. Three-month post-treatment CSF pressures varied from 9 to 19 cm H₂O (mean 16 cm H₂O).

Stent patency was evaluated at 3 and 12 months post-treatment. The dural sinus stents were patent at 3 months in all patients and at 12 months for all patients with a follow-up superior to 12 months. Stent patency was evaluated systematically by DRCV and MR venography at 3 months and by either MR or CT scan venography at 12 months (figure 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Multidetector row three-dimensional CT angioscan showing the patency of the stent and the cortical drainage of veins through the stent wall

DISCUSSION

The therapeutic management of IIH remains problematic. Indeed, the various treatments attempting to reduce intracranial pressure constitute only symptomatic rather than pathogenic approaches that do not allow reproducible and long-term satisfactory results to be obtained.⁵ The lack of efficacy of current medical and surgical treatments proposed in IIH reflects the incomplete comprehension of the pathophysiologic mechanisms underlying this pathology.

Since 1996, some authors have proposed that elevated intracranial venous pressure constitute a universal mechanism of IIH.¹⁰ They conclude that whether secondary to intracranial venous outlet obstruction or idiopathic, raised intracranial hypertension in this pathologic entity may have increased venous pressure as its final common pathway. An association between venous sinus obstruction and IIH is not new. Thrombosis of venous sinuses was originally postulated as a cause of so-called idiopathic intracranial hypertension syndrome as early as the 1930s.¹¹ Support for such a mechanism was provided by other authors who were able to obtain images of the venous sinuses.^12,13 In 2002, Johnston et al. reported that the incidence of cranial venous outflow obstruction was 4.2% before 1979 and increased to 31% after 1989.¹⁴ More recently, dynamic three-dimensional gadolinium-enhanced MR venography studies have found bilateral TS stenosis in more than 90% of patients with IIH.^15–17 The discrepancy between venous sinus obstruction incidences in IIH over the past three decades underscores the limitations of the previous imaging methods, namely digital subtraction angiography (DSA) and time-of-flight (TOF) MR venography. Indeed, DSA is usually performed in two static planes, making difficult the assessment of the transverse sigmoid junction, which is a frequent site of stenosis. Only rotational three-dimensional DSA with injection of a large amount of contrast medium could approach the quality of dynamic three-dimensional gadolinium-enhanced MR venography. In the same way, TOF MR venography presents technical limitations, specifically the artifactual signal loss that occurs at predictable locations in the venous system due to in-plane flow and turbulence.^17,18 It is now well known that transverse and sigmoid sinuses are locations routinely plagued by such artifacts.¹⁷ In a recent review of IIH mechanism and treatment, it was recommended that standard MRI and MR venography should be performed in any patient with suspected IIH.

Since intracranial venous outlet obstruction has been demonstrated in IIH, several groups have undertaken catheter studies to analyze and record intrasinus pressure (manometry) in this pathology.^10,19,20 These groups have documented high intracranial venous pressure in patients with IIH, occasionally secondary to systemic venous hypertension but more often apparently the result of stenotic lesions of the venous sinuses. These observations have led some authors to propose, successfully, in patients with IIH, transstenotic venous sinus stenting to restore a normal intracranial venous pressure regimen.²¹ To date, including our series, only 28 cases of patients with IIH treated by this method have been reported.^8,9,22–24 Higgins et al. recently published their results in 12 patients.⁸ Among these patients, only 7 had no previous treatment and 5 were improved (71%). If we consider only patients with pretreatment visual impairment and with no previous therapeutic procedure, 80% were improved. These data compare favorably with our results. Indeed, in our 10 patients treated with venous sinus stenting, 6 were cured and 4 were improved. Among the 8 patients with visual impairment, 4 recovered normal visual acuity, 3 were improved, and 1 was unchanged. The patient who did not show visual improvement had both severe and long-standing (46 months) symptoms. This case underlines the difficulty in determining the duration of medical treatment before aggressive management. To date, there is no consensus on the timing of surgery or endovascular management in IIH. We consider, as others, that there is no acceptable level of visual field or acuity loss that should be awaited to indicate surgery.⁶ Another issue needs to be raised regarding functional results. It is well known that invasive procedures have a high placebo rate, and one can wonder whether there is a placebo effect in the results obtained with venous sinus stenting. In our patients, 9 of 10 had improvement of objective symptoms such as papilledema or visual acuity besides relief of headaches. Furthermore, in all patients, the decrease in CSF pressure was documented. We acknowledge that regarding headaches, there is probably a placebo effect; however, the visual and CSF pressure results obtained are objective.

In the authors' opinion, the exact site where pressures are recorded may be an important issue. Certainly to make different series comparable and to define a reliable gradient threshold for stenting, it is essential for physicians to assess the pressure gradient using the same method. In fact, intrasinus venous pressures varied, decreasing from the superior sagittal sinus to the jugular bulb even in the absence of stenosis. In our series, the pressure gradients were systematically calculated from the proximal prestenotic TS to the poststenotic distal TS.

With regard to the morbidity of the technique used, once again insufficient data are currently available in the literature to appreciate its true extent. However, rare severe and permanent complications associated with this endovascular procedure have been reported in the literature.²⁵ In only two cases, retrograde venous angiography demonstrated a thrombosed stent in the immediate postoperative course, necessitating a thrombolysis.⁸ In these two cases, it was not mentioned whether the procedure was performed under anticoagulation only or associated with antiplatelet therapy.⁸ In our experience with venous sinus stenting in patients with IIH, we did not encounter any complication during the follow-up period. Recently, we reported the treatment of 10 dural arteriovenous fistulas of the transverse sigmoid sinus using self-expanding stent placement and balloon angioplasty.²⁶ In this series, at the beginning of our experience, endovascular procedure was performed under heparin therapy only, which was discontinued 24 hours postoperatively, and one patient experienced an intraparenchymal hematoma secondary to the occlusion of a superficial cortical draining vein. Since then, we have used systematically an associated antiaggregation therapy by clopidogrel 3 days before and 3 months after, and so far, no other complication has occurred in our population of stented patients. We recommend, as others, continuing the antiaggregation therapy by clopidogrel 3 months postoperatively because this corresponds with the time of stent endothelialization.^9,26

In addition to the results obtained, this treatment seems particularly interesting because the therapeutic strategy used possibly improves patients' clinical status via a pathogenic rather than a symptomatic approach. However, pathophysiologic mechanisms underlying IIH remain an ongoing source of debate. If, within the past few years, increased venous sinus pressure has been hypothesized to be the main cause of IIH, the role and mechanisms of venous sinus stenoses are still controversial. Some authors have claimed that lateral sinus stenoses are, in a probably underestimated proportion of cases, the cause of IIH.²¹ They treated their first patient who presented with a typical IIH and bilateral venous sinus stenoses, in 2002, with a stent that reduced both pressure gradient and intracranial venous hypertension, effecting immediate clinical improvement that has been maintained over the long-term follow-up period.²¹ They speculated that given the difficulty of establishing venous sinus stenosis, its role in the etiology of IIH was probably underestimated.²¹ In 1995, some authors measured intracranial venous pressures with transducers in patients with IIH.¹⁹ They found high pressures in the venous sinus with a sudden decrease in pressure distally that suggested some type of venous obstruction or stenosis. In 2002, the same team found in patients with IIH that when transducer-measured intracranial venous pressure is high, reduction of CSF pressure by removal of CSF predictably lowers the venous sinus pressure.²⁰ They concluded that the increased venous sinus pressure was caused by the elevated intracranial pressure and not the reverse. In a recent study on the prevalence and the morphology of sinovenous stenosis in IIH, it has been shown that more than 90% of patients have TS stenotic appearance compared with those of normal patients and that there were two types of sinovenous narrowing.¹⁷ The first corresponds to a long, smooth, tapered narrowing of external compression caused by brain parenchyma, and the second corresponds to an acutely marginated apparent intraluminal filling defect of an enlarged, partially obstructing, arachnoid granulation swollen by elevated CSF pressure. Based on these observations and the fact that, in the confined intracranial space, the vascular compartment gives way to an expanding parenchymal and CSF compartment, some authors concluded that sinovenous stenosis was secondary to raised intracranial hypertension of unknown origin.²⁰ Thus, they speculate that in IIH, there is a prepresentation clinical status in which the raised intracranial pressure of unknown origin (first event) remains asymptomatic or poorly symptomatic. During this phase, progressively, the dural walls of the lateral sinuses slowly stretch and collapse, yielding to a long, smooth, extraluminal compression or a focal protrusion of the arachnoid granulation. This intraluminal or extraluminal obstruction leads secondarily to a flow-limiting stenosis (second event), resulting in an upstream increased venous pressure responsible for a pressure gradient.

Actually, even if the factors that initiate the pathologic process remain unknown, from a therapeutic standpoint, whether the sinus stenosis is the first or the second component of the chain of events leading to increased CSF pressure may not be the most important issue.²⁷ Venous sinus stenting seems to constitute, in some cases, an effective and safe technique able to alleviate the exacerbating role of the venous hypertension and return the patient to a nonsymptomatic state. However, the safety and efficacy of this technique should be evaluated further in larger series with longer follow-up to be definitely validated.

ACKNOWLEDGMENT

The authors thank Maryna Blankenstein-Gabert for help with English language editing.