Article Text
Abstract
Background Transverse sinus (TS) stenting is a valid treatment alternative for patients with intracranial hypertension caused by underlying bilateral TS stenoses. Its mid-term patency has, however, not been well documented.
Objective To assess the 6-month patency of TS stenting using subtracted CT venography (CTV).
Methods A retrospective analysis of a prospectively collected database of patients undergoing TS stenting was performed. The cohort was a single-center, single-operator series of 125 consecutive patients treated between 2008 and 2018. Mid-term follow-up 320-row detector CTV was available for review in 104 patients.
Results Follow-up CTV was obtained on average 6 months after stenting. Stents in all patients (100%) were patent. Subtracted reconstructions showed no intraluminal thrombus or neointimal hyperplasia. Native reconstructions confirmed the structural integrity of the stents. De novo stenosis proximal to the stent was noted in 10 cases (10%). A total of 10 patients (10%) received additional treatment due to recurrent symptoms. In univariate analysis, both high body mass index and stent size (>6 mm) were associated with development of de novo stenoses: OR 1.12 (95% CI 1.01 to 1.25, p=0.037) and OR 5.63 (95% CI 1.16 to 27.22, p=0.032), respectively. In multivariate analysis, only stent size (>6 mm) remained significant: OR 7.19 (95% CI 1.03 to 50.01, p=0.046).
Conclusion TS stenting is an effective treatment for intracranial hypertension secondary to dural sinus stenosis in an appropriately selected patient population. A 320-row dynamic CTV is a high-quality non-invasive imaging method that can assess both the physical integrity of the stent and its patency. At mid-term follow-up, all imaged stents were patent. The occurrence of de novo stenoses proximal to the stent (10%) correlated with stent size (>6 mm).
- stenosis
- CT angiography
- intracranial pressure
- stent
Statistics from Altmetric.com
Introduction
Nonne coined the term pseudotumor cerebri in 1904 to describe an entity characterized by increased intracranial pressure (ICP) of unknown etiology.1 The syndrome was later renamed benign intracranial hypertension by Foley (1955)2 and idiopathic intracranial hypertension (IIH) by Buchheit and colleagues (1969).3 The latter authors emphasized that the term IIH should be used only after conditions known to produce papilledema and increased ICP—for example, dural sinus thrombosis—have been ruled out. IIH is about 10 times more common in women, with an incidence of 3 in 100 000 per year; obesity is a strong risk factor.4 Patients usually present with headache, nausea, vomiting, visual field disturbances, and papilledema.4 5 Unilateral or bilateral pulsatile tinnitus is common.6 The diagnosis relies on modified Dandy criteria.5
Oral acetazolamide, often at a high dose, is the primary medical treatment for IIH; side effects such as paresthesia, excessive fatigue, and kidney stones limit its long-term use.7 Surgical options, such as ventriculoperitoneal (VP) or lumboperitoneal shunt placement and optic nerve sheath fenestration (ONSF), are palliative methods aimed at reducing the ICP and preserving vision.8
Transverse sinus (TS) stenting was first performed to treat a patient with pulsatile tinnitus by Marks and colleagues in 1994.9 The subsequent demonstration of the role of increased venous pressure in patients with pseudotumor cerebri10 11 led to the application of TS stenting in that population, first reported by Higgins and coauthors.12 13 Stenting has now been proved to be a valid alternative for patients with intracranial hypertension secondary to dural sinus stenosis that is refractory to medical management.14 It can achieve symptomatic improvement and papilledema resolution in 87% and 90%, respectively,15 16 and has the advantage of being curative rather than palliative for patients with TS stenoses.
However, mid- to long-term efficacy of TS stenting remains debated. One systematic review revealed a pooled re-treatment rate of 10.6% (22/207 patients), of which 72% (17/22 patients) underwent stenting of a de novo stenosis, either adjacent to a prior stent or occasionally involving the contralateral TS.15 A more recent meta-analysis reported stent survival rates and proximal restenosis of 84% and 14%, respectively.17 Our study evaluates mid-term stent patency in a large homogeneous series using subtracted dynamic CT venography (CTV).
Methods
Institutional review board approval was obtained both for data collection and result publication. This series consists of a total of 125 adult patients with intracranial hypertension refractory to medical management treated with TS stenting at a single center by a single operator between January 2008 and December 2018. The cohort was divided into two patient groups (2008–2010) and (2011–2018) separated by a 1-year pause aimed at evaluating safety in the initial group.
Mid-term follow-up CTV was available in 104 patients; the remainder either had magnetic resonance venography or declined additional imaging. All patients satisfied the modified Dandy criteria for IIH and showed imaging evidence of dominant unilateral or bilateral TS stenosis prior to stenting. Failure of medical treatment included allergy or intolerance to escalating doses of acetazolamide as well as worsening papilledema, despite daily doses of up to 3000 mg. Patients’ demographics included age, gender, race, body mass index (BMI), the most recent cerebrospinal fluid (CSF) opening pressure on lumbar puncture prior to TS stenting, relevant medical and surgical history, and need for additional treatment. All patients received 320-row CTV within 2 hours after obtaining a lumbar puncture, with reduction of CSF opening pressure to <15 cm H2O, in order to quantify the degree of TS stenosis with minimum external compression due to increased intracranial pressure, during work-up for the TS stenting procedure.
Transverse sinus stenting
All procedures were performed under general anesthesia. Femoral 8F venous and 4F arterial accesses were obtained. Four-vessel diagnostic cerebral angiography was performed to rule out concomitant vascular anomalies; two dural arteriovenous fistulas and one saccular aneurysm undiagnosed by MR and/or CT angiography were documented and treated prior to TS stenting. Cerebral venous pressure measurements were performed by placing a guide sheath (6F Neuron Max, Penumbra, USA) into the internal jugular vein (IJV) ipsilateral to the targeted stenosis. A biplane venous roadmap was obtained from an optimal venous image selected from the venous phase of the cerebral angiogram (Roadmap; Artis Zee, Siemens) and used to advance a 2.8F microcatheter (Renegade Hi Flo; Boston Scientific, USA) over a 0.018-inch wire (Transend; Boston Scientific, USA) across the midline through the torcula to the opposite IJV. An exchange wire (300 cm Luge; Boston Scientific, USA) was then placed in the contralateral IJV and pressure measurements were taken at multiple sites while the microcatheter was progressively pulled back, including both IJVs, sigmoid sinuses, distal and proximal TS, and the torcula.
Patients were heparinized (targeted activated clotting time 250 to 300 s) during the procedure and overnight. Stenting was performed when the measured pressure gradient across the TS stenosis (i.e. between proximal and distal transverse sinus) was 4 mm Hg or higher.
The only difference between the two treatment groups concerned the size of the stents. In the first group (2008–2010), stent sizes were selected based on TS diameter measurement. In the second group (2011–2018), only 6×40 mm or occasionally 5×40 mm stents were used, without TS measurement.
All the stents used were self-expanding: a tapered stent in one early case (7–10 mm by 40 mm Acculink, Abbott, Lake Bluff, Illinois, USA) and non-tapered stents in all the others (5 to 10 mm by 40 mm Precise stent; Cordis, Fremont, California, USA). Wider diameters (7 to 10 mm)—used early in our experience—were replaced by 6 mm stents for later cases (91 cases).
Long stenotic segments required the use of two (17 cases) or three (3 cases) overlapping stents. Pre-stenting and post-stenting angioplasty was used in 12 and 10 cases, respectively. Post-stenting venous manometry confirmed gradient reduction in all instances. Daily clopidogrel (75 mg) and aspirin (325 mg) were initiated 7 days before procedure and continued for at least 6 months.
320-Row dynamic volume CT venography
Follow-up CTV using a 320-detector subtracted dynamic volume CT scanner (Aquilion ONE; Toshiba, Tustin, California, USA) was performed on average 6 months (median 6; range 2 to 20) after stenting. Technical parameters were as follows: 0.5 mm detector width, 0.25 mm reconstruction interval, 512×512 matrix, 160 mm field-of-view, 0.75 s scan rotation time, 80 kV tube voltage, and 150–280 mA tube current. After acquisition of a native volume scan for mask subtraction, 50 mL of non-ionic contrast material (iopamidol; Isovue 370, Bracco Diagnostics, Italy) was injected intravenously at a rate of 6 mL/s followed by a saline flush. The dynamic acquisitions included arterial and venous phases respectively obtained 15 to 25 s and 30 to 45 s post-injection.
Postprocessing
The precontrast acquisition was used both as a native series and as a mask image for the subtracted series. The data were transferred to a 3D workstation (Carestream; Rochester, New York, USA). Subtracted dynamic 3D CTV images were generated using a volume rendering technique. The acquisition that provided the best depiction of the dural venous sinuses and the stent was selected for evaluation. The respective benefits of native and subtracted images are illustrated in figure 1. Recorded parameters included stent patency, in-stent intimal hyperplasia or thrombosis, stent kinking or fracture, and de novo dural sinus stenosis. The imaging studies were analyzed on a consensual basis by an experienced neurointerventionalist (10-years' experience; MSP) and a senior neurointerventional fellow (AEM), who were not involved in the procedures.
Statistical analysis
Data are reported as mean (SD; SD and range) or median (25th-75th percentile IQR; IQR) as appropriate for continuous variables and as frequency (percentage) for categorical variables. The stented population was analyzed as a whole and as separate treatment groups (Group 1=2018–2010, group 2=2011–2018). Fisher’s exact test, chi-square (χ2) (for categorical variables), unpaired Student t-test and Mann-Whitney U test (for continuous variables) were used, as appropriate, for univariate analysis of demographic variables. A univariate logistic regression analysis to evaluate predictors (OR; OR and 95% CI; CI) of significant anatomical findings on CTV was performed. Relevant variables found on univariate analysis were then examined with multivariate logistic regression to account for independent predictors of significant findings on CTV. BMI was reported as a continuous variable and also based on the WHO classification18 as an ordinal variable (normal weight: 18.5–24.9 kg/m2; overweight: 25–29.9 kg/m2; obese class I: 30–34.9 kg/m2; obese class II: 35–39.9 kg/m2; and obese class III≥40 kg/m2). Statistical analysis was performed with SPSS version 24 (IBM, USA). A p value of <0.05 was considered significant.
Results
Patient characteristics
We included 104 patients in this analysis, 95 (91%) women and 9 (9%) men. Their mean age was 36 years (SD 11, range 18–65) and the mean BMI 31 kg/m2 (SD 6, range 20–50). The BMI WHO class was as follows: (a) normal weight: 10 (10%), (b) overweight: 38 (37%), (c) obesity class I: 32 (31%), (d) obesity class II: 12 (11%), (e) obesity class III: 12 (11%). Ethnic distribution was as follows: 75 patients (72%) were Caucasian, 23 (22%) African American, and 6 (6%) from other groups. The mean CSF opening pressure was 35 cm H2O (SD 9, range 22–65). Five patients (5%) had prior non-medical treatment for IIH: three patients had a VP shunt, one had ONSF, and one had both a VP shunt and ONSF. In addition, three patients (3%) had prior repair of a CSF leak (one rhinorrhea, one otorrhea, and one rhinorrhea/otorrhea). One patient had Rosai-Dorfman disease.
Venous manometry and transverse sinus stenting
Pre-stenting four-vessel diagnostic angiography, venous manometry, and stent placement was performed in all cases without neurological complications. The median pretreatment pressure gradient across the TS stenosis on the dominant side was 9 mm Hg (IQR 6–13). Stents were successfully deployed in a single session in all cases but one. The exception was a group 1 patient in whom a focal subdural intraprocedural contrast extravasation led to procedural abortion without clinical repercussion; stenting was then successfully completed 3 weeks later using a different type of stent. Pre-stenting and post-stenting balloon angioplasty was used in 12 and 10 instances, respectively. Ninety-one patients (88%) received a stent with a diameter of 6 mm or less, and 13 (12%) with a diameter greater than 6 mm. Twenty patients (19%) received multiple stents (two stents in 17 cases, three stents in three cases) to cover longer lesions. A comparison of patient characteristics between group 1 and group 2 is listed in table 1. The only significant difference between the two groups was the BMI: mean 35 (SD 8, range 25–50) in group 1 versus 31 (SD 6, range 20–50) in group 2 (p=0.031). Stents were deployed on the right side in 80 patients (77%). The median reduction in pressure gradient was 7 mm Hg or 88% (IQR 5–12 and 76–100%, respectively) with a median pressure gradient of 2 mm Hg (IQR 0–2) after stenting.
Findings on 320-row subtracted dynamic CTV follow-up
The follow-up CTV—performed on average 6 months after stenting (SD 3, range 2 to 20), revealed patent and structurally intact stents in all 104 patients (100%), without evidence of thrombosis, intimal hyperplasia, kinking, or migration. Focal de novo dural sinus stenoses adjacent and proximal to the stent were noted in 10 patients (10%) (group 1: 4/13 (31%) vs group 2: 6/93 (6%), p=0.001). See examples in figure 2 and Supplementary Figure 1, including one patient who developed an interval diffuse dural sinus narrowing due to severe pachymeningeal thickening from fulminant flaring of known Rosai-Dorfman disease. In addition, two patients (2%) showed mild regrowth of an arachnoid granulation through the stent mesh (online supplementary figure 2). A comparison of demographics, manometry findings, and treatment in patients who developed focal de novo dural sinus stenosis with the rest of the cohort is shown in table 2.
Supplemental material
A univariate logistic regression analysis of predictors of the development of de novo dural sinus stenosis was performed (table 3). In this primary analysis, both BMI (OR 1.12; 95% CI 1.01 to 1.25, p=0.037) and a stent size over 6 mm (OR 5.63; 95% CI 1.16 to 27.22, p=0.032) were significant predictors. Clinically relevant variables found in the univariate analysis (BMI, CSF opening pressure, and large stent size) were then entered in a multivariate logistic regression model analysis, which revealed that larger stent size (>6 mm) was independently associated with the development of de novo dural sinus stenosis after TS stenting (OR 7.19; 95% CI 1.03 to 50.01, p=0.046). Overall 10% (10/104) of patients required re-treatment after TS stenting owing to recurrent symptoms; the re-treatment rate was higher in group 1 (4 of 13, or 31%) than group 2 (6 of 91, or 7%). Eight of the 10 patients who required re-treatment had evidence of de novo dural sinus stenosis on CTV. The other two patients had recurrent symptoms without evidence of de novo dural sinus stenosis on CTV, or pressure gradient across the stent on repeat endovascular measurement, both received a VP shunt.
Discussion
A substantial proportion of patients with suspected IIH are diagnosed with cerebral venous hypertension secondary to stenosis of a dominant TS or both TSs.17 19 The benefit of dural sinus stenting in this population has been established: an overall improvement of symptoms caused by intracranial hypertension may be expected in 87% of cases,15 with a major complication rate of 1.9%.20 But the durability of that treatment option, notably for stent patency, is not well documented. The aim of our investigation was to assess the appearance of stents deployed in TSs of 104 patients with intracranial hypertension secondary to dural sinus stenosis based on review of routine 6-month follow-up CTVs. Dynamic subtracted CTV is considered, in our practice, the modality of choice in this indication as it provides precise vascular imaging devoid of metallic artifacts as well as a fine analysis of the stent structure, including kinking, fracture, or migration.
There was no instance of in-stent thrombosis and/or stenosis caused by endothelial hyperplasia in our study. All the deployed stents were intact, without structural defect or migration. These findings, which are in agreement with studies reporting a 100% stent patency on TS stenting followed up by DSA,21–24 appear to indicate a better stent survival than suggested by a recent meta-analysis, probably due, in part, to improved devices.17 Of note, we observed two instances of minor in-stent regrowth of an arachnoid granulation at 5 and 6 months. Although this finding had no clinical impact in our patients, the possibility of further growth, leading to delayed flow impairment, cannot be excluded.
The most significant finding in our study was the development of a new focal dural sinus stenosis immediately proximal to the stent in 10/104 (10%) patients. Such focal de novo stenoses have been reported in 12% to 26% of stented cases14 20 21 23 24; their etiology remains unclear.
An initial univariate analysis showed that both BMI (OR 1.12; 95% CI 1.01 to 1.25, p=0.037) and larger stents (>6 mm; OR 5.63; 95% CI 1.16 to 27.22, p=0.032) were statistically significant predictors of development of a de novo stenosis. However, in a multivariate regression analysis, only the stent size remained an independent predictive factor (OR 7.19; 95% CI 1.03 to 50.01, p=0.046). This finding is in agreement with the clinical suspicion of the role played by stent size during the treatment of our first group of patients, which empirically led us to use only 5 mm or 6 mm stents in our second group. Different approaches have been suggested to deal with this problem—for example, by extending the stented segment from the torcula to the upper sigmoid sinus.21 In our series, multiple overlapping stents were used only to treat longer diseased segments (20 of 104, or 19%). Our overall re-treatment rate was 10% (10/104): six patients underwent additional stenting (including one patient who was later treated with a VP shunt after developing a second proximal de novo stenosis) and four patients treated with a VP shunt. Our results fall within the recently reported rate of recurrent intracranial hypertension symptoms after TS stenting of 9.8% (95% CI 6.7% to 13%).20
This study confirms the safety of TS stenting, with no instance of neurological complications related to either diagnostic angiography or stent placement in our cohort. It also emphasizes the role played by subtracted dynamic volume CTV in the assessment of the stent physical integrity, its patency, and the patency of cortical veins draining through the stented segment (online supplementary figure 1).
This study has several limitations. It is a retrospective analysis of a cohort treated by a single operator. The 6-month delay for follow-up CTV—chosen according to our planned antiplatelet regimen—might also be viewed as a limiting factor since one of our patients showed progression of a de novo stenosis at 24 months (online supplementary figure 1). Regrowth of an arachnoid granulation through the stent mesh—if confirmed—could potentially also manifest clinically with longer delays.
Conclusion
Transverse sinus stenting is a safe and clinically durable technique for the treatment of intracranial hypertension secondary to dural sinus stenoses. A 320-row subtracted dynamic volume CTV provides a high-quality, non-invasive imaging method for the assessment of stent integrity and patency. At the 6-month follow-up evaluation, 100% of the stents in our cohort were patent. The development of a de novo stenosis immediately proximal to the treated segment was found in 10% of our patients, which was independently correlated with a stent size larger than 6 mm.
References
Footnotes
Contributors All authors participated in the design, conceptual design and approved the final version of the draft.
Funding This work was supported by Swiss National Science Foundation (SNSF), grant number P2SKP3-164949, in the form of as a mobility career stipend for the first author (AEM).
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Institutional Review Board approval (no: NA_00026773) was obtained both for data collection and result publication.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information.