Article Text
Abstract
Objective To determine the relationship between normal physiologic and pathologic venous sinus pressures in patients with idiopathic intracranial hypertension (IIH), which is poorly understood.
Methods Retrospective analysis was performed to identify patients with medically refractory IIH who were evaluated by angiography and retrograde venography with venous manometry. Patients were further subdivided into groups based on anatomic factors.
Results 104 patients met inclusion criteria for the study. In the absence of non-invasive venographic screening, 58% of patients in this series were found to have pressure gradients of ≥8 mm Hg; 93% were located near the transverse-sigmoid sinus junction. Opening pressure (OP) is strongly predictive of superior sagittal sinus (SSS) pressures (p<0.001) and also of the presence of a pressure gradient ≥8 mm Hg (p<0.001). Twenty-three percent of patients with an OP <25 had a pressure gradient ≥8 mm Hg compared with 77% of patients with an OP ≥35. Analysis of patients with OP ≤20 suggests that SSS pressures in patients without IIH should be less than 16–18 mm Hg with total cranial gradients <5 mm Hg. Across all patients, a pressure decrement of approximately 1 mm Hg occurs with progressively more caudal transition across anatomic points of measurement.
Conclusions This study describes intracranial and extracranial venous pressure measurements and gradients in different subgroups of patients with IIH. OP is highly predictive of intracranial venous pressures and significant venous pressure gradients.
- idiopathic intracranial hypertension
- pseudotumor cerebri
- venous sinus stenting
Statistics from Altmetric.com
Introduction
Idiopathic intracranial hypertension (IIH), also referred to as pseudotumor cerebri or benign intracranial hypertension, is a medical condition characterized by a combination of intractable headaches, papilledema, visual symptoms, tinnitus, and an elevated cerebrospinal fluid (CSF) opening pressure (OP) on lumbar puncture (LP) in the absence of an intracranial mass. The often cited Dandy criteria require an OP of 25 cm H2O for diagnosis, but most clinicians now realize that this threshold is obsolete as many patients have debilitating symptoms at pressures <25 cm H2O and obtain significant symptomatic relief from pressure-lowering treatments. In the past decade, venous sinus stenting has matured as a promising treatment for patients with IIH with associated venous sinus or internal jugular vein stenosis (sometimes referred to as ‘secondary intracranial hypertension’). Systematic reviews and meta-analyses demonstrate 78–83% improvement in headache, 87–97% improvement in papilledema, 74–85% improvement in visual symptoms, and 95% improvement in tinnitus following stenting.1–4 Evidence-based recommendations have been published supporting selection of patients for this treatment,5 which requires documentation of a trans-stenosis pressure gradient of substantial magnitude.
Increasing evidence supports stenting as a treatment for IIH with symptomatic venous sinus stenosis, but our understanding of the normal physiology of cerebral venous drainage and the pathophysiology of stenosis remains poorly understood. Studies have demonstrated that unidirectional flow of CSF through arachnoid granulations requires a 3–5 mm Hg pressure gradient from subarachnoid space to venous sinus.6 Patients with venous sinus stenosis may have pronounced venous congestion upstream from the stenosis, resulting in elevated intracranial pressures due to high venous pressures. However, normal physiologic venous pressures in individuals without elevated intracranial pressures have yet to be reported. In addition, normal or pathologic venous pressure decrements along the venous drainage pathway have not been reported. This study seeks to describe venous pressures and gradients in a large series of patients with IIH during retrograde venography and manometry and provide insight into the relationship between patient factors such as gender, age, body mass index (BMI) and LP OP with venous pressure measurements.
Methods
A database of patients evaluated for IIH was retrospectively queried after obtaining institutional review board approval. Criteria for inclusion in the study were presentation with clinical symptoms of IIH without evidence of intracranial mass lesion on imaging, a lumbar puncture opening pressure >20 cm H2O in the lateral decubitus position, and either medically refractory symptoms or medication intolerance. All included patients underwent diagnostic catheter angiography and retrograde venography with venous manometry. Non-invasive venous imaging was not considered in determining candidacy for venography and manometry. Patients who had previously undergone venous sinus stenting or a cerebrospinal fluid shunting procedure with an indwelling shunt system were excluded. Patients with a known diagnosis of sinus thrombosis were also excluded.
Chart review of each patient was performed to identify demographic features (age, gender, and BMI) and OP on LP. In all cases, OP was recorded in cm H2O. If several LPs had been performed, the most recent value was used. Any patient with a LP performed within 7 days before the venogram procedure was excluded given the known temporary effect of CSF diversion on venous pressure gradients.7–9 Patients who had a LP within the 7 days after the venogram procedure were included. A minority of patients with high suspicion for IIH, or previously documented elevated OP, underwent venography followed by LP and were discovered to have OPs <20 cm H2O (three patients). These patients were included in the study to provide more information on venous pressures in individuals with lower intracranial pressures.
Procedural details
Diagnostic angiography, venography, and venous sinus manometry were performed in all patients under minimal conscious sedation with fentanyl. Procedures are performed under minimal conscious sedation owing to known confounding effects of general anesthesia and ventilator changes on venous pressure measurements.10–12 The femoral artery and vein were accessed in all patients and 5F sheaths placed. Cerebral arteriography was performed with a 5F diagnostic catheter to determine the presence of venous sinus stenosis, venous outflow patterns, and to rule out arteriovenous fistulae. After arteriography, a 5F catheter was positioned in the right and/or left internal jugular veins, depending on the side of venous dominance, near the jugular bulb. A 0.027 inch Rebar microcatheter (Medtronic, Minneapolis, Minnesota, USA) was then navigated over a 0.014 inch microwire into the superior sagittal sinus (SSS), and supraselective diagnostic venography was subsequently performed. After venography, manometry was performed in the SSS, torcula, transverse sinus (TS), sigmoid sinus, internal jugular vein just below the bulb (IJ), and superior vena cava-atrial junction (central venous pressure (CVP)). In all cases, the venous manometry was performed in the dominant (or co-dominant) transverse-sigmoid-jugular pathway. All pressure measurements were obtained using standardized techniques and standardized anatomic locations by a single operator; in all cases a single mean pressure was recorded measured in units of mm Hg. Pressures obtained in the non-dominant, contralateral transverse-sigmoid pathway were not included in this analysis.
Definition of pressure gradients
Pressure gradients at adjacent anatomic locations were calculated for each patient. Table 1 shows the definition of each pressure gradient and the methodology for calculation. Adjacent anatomic gradients are defined as the difference between venous sinus pressure measurements at the next anatomical location in which manometry was performed. Summative gradients are defined as those that include multiple adjacent anatomical gradients in their calculation.
Stratification of patients
The presence and magnitude of venous sinus pressure gradients found on manometry as well as the details of intervention were noted for each patient. Pathologic venous sinus pressure gradients were defined as those of at least 8 mm Hg based on established convention.5 13
The patients were further stratified into three separate groups based on presentation and venous pressure measurements: (1) all patients (group A); (2) those with at least one pathologic adjacent anatomic pressure gradient (≥8 mm Hg) to summarize the subgroup of patients that would potentially benefit from stenting (group B); and (3) patients with an OP of ≤20 and all adjacent anatomical pressure gradients <4 mm Hg (group C). Summative gradients (total cranial pressure gradient or overall pressure gradient) were not considered in stratification. The final group, group C, was selected to represent the subgroup of patients without venous outflow obstruction and marginally elevated or normal intracranial pressures to better understand normal venous pressure decrements throughout the cerebral and cervical venous system in patients without intracranial hypertension.
Statistical analysis
All analyses were conducted using R: a language and environment for statistical computing (R Foundation for Statistical Computing, version 3.5.1, Vienna, Austria and RStudio: Integrated Development for R, version 1.1.456 RStudio, Inc., Boston, Massachusetts, USA). Descriptive statistics were calculated, with mean (SD) used for normally distributed variables and median (range) for non-parametric data. In some instances, both mean and median are presented for reference. For all analyses, two-tailed hypothesis testing was used with p<0.05 interpreted for statistical significance. To investigate relationships among patient characteristics and various pressures, Pearson (r) or Spearman (ρ) correlations were performed depending on the normality of the data. To examine the difference in BMI between patients with and without an adjacent anatomical pressure gradient of ≥8 mm Hg, a t-test was performed. Individual linear models were used to predict torcula, TS and SSS pressures by OP and to predict CVP by BMI. Logistic regression was used to predict pathologic adjacent anatomic pressure gradient by age, BMI, gender and OP. Model fit was determined by likelihood ratio tests and fit statistics.
Results
A total of 120 patients underwent angiography with venography to evaluate venous sinus outflow obstruction. Of these, 14 were excluded owing to the presence of a venous sinus stent or a CSF shunt at the time of the procedure. An additional two patients were excluded owing to the presence of known sinus thrombosis.
Group A: All included patients
Of the remaining 104 patients, all had recorded data detailing pressure measurements at all locations (SSS, dominant TS, dominant SS, dominant IJ, and CVP) except for 17 patients (16.3%) where CVP was not recorded. These patients were all included in group A. A total of 87.5% of patients were female with a mean age of 37.3 years (SD 13.1). Mean BMI was 37.8 (SD 9.8). Most recent OP on LP for these patients was a median of 30.0 (IQR 11.0). Median, mean, and IQR descriptive data for all anatomical locations and gradients among all included patients are shown in figure 1.
Table 2 demonstrates the number and percentage of all included patients with identified adjacent anatomic pressure gradients. Measurements were performed in the right transverse-sigmoid pathway in the majority of cases (80, 76.9%).
Group B: Presence of any adjacent anatomic pressure gradient ≥8 mm Hg
Of the 104 patients, 60 (57.7%) had at least one adjacent anatomical pressure gradient of ≥8 mm Hg and were included in group B. This group included 55 women (91.7%) with mean age of 36.6 (SD 12.8), mean BMI of 38.3 (SD 7.9), and median OP of 32.4 (IQR 14.0). Median, mean, and IQR descriptive data for all anatomical locations and gradients are shown in figure 2.
Group C: OP of ≤20 and all adjacent anatomic pressure gradients <4 mm Hg
Of the 104 patients, 3 (2.9%) had adjacent anatomical pressure gradients of ≤3 mm Hg and OP ≤20 and were included in group C, with two being women with a median age of 41 (IQR 12.5) and median BMI of 19.4 (IQR 11.07). Median, mean, and IQR descriptive data for all anatomical locations and gradients are shown in figure 3.
Opening pressure
The largest adjacent anatomical pressure gradient was highly associated with OP (ρ=0.40; p<0.001; figure 4A). Table 3 demonstrates the percent of patients with adjacent gradients of ≥8 mm Hg who would potentially be amenable to stenting based on OP stratification.
Correlation of OP with torcula, TS, and SSS pressures
Transverse sinus, torcula, and SSS pressures all demonstrated significant correlations with OP on LP. Torcula pressures were highly correlated with OP (ρ=0.42; p<0.001). Torcula pressures can be predicted by OP using the formula: torcula=8.67 + 0.59*OP (adjusted R2=0.19; p<0.001). TS pressures were also highly correlated with OP (ρ=0.43; p<0.001). TS pressures can be predicted by OP using the formula: TS=8.30 + 0. 57*OP (adjusted R2=0.19; p=0.039).
However, of the three locations, SSS pressures best correlated with OP (ρ=0.44; p<0.001). SSS pressures can be predicted by OP using the formula: SSS=6.80 + 0.69*OP (adjusted R2=0.24; p<0.001; figure 4B).
CVP was not significantly correlated with OP (ρ=−0.09; ρ=0.42).
Body mass index
There was no relationship between BMI and the presence of an adjacent anatomical pressure gradient of ≥8 mm Hg (t=−0.55, df=69, p=0.58).
Correlation of BMI with CVP
CVPs were correlated with BMI (r=0.37, p<0.001). CVP can be predicted by BMI using the formula: CVP=5.26 + 0.18*BMI (adjusted R2=0.13; p<0.001; figure 4C).
Predictors of pathologic adjacent anatomic pressure gradient ≥8 mm Hg
The variables OP, age, gender, and BMI were analyzed as potential predictors of a pathologic adjacent anatomic pressure gradient. Age and BMI were not significant predictors (p=0.23 and 0.64, respectively) and were removed as they did not improve model fit. Both OP and female gender were significant predictors of a pathologic pressure gradient (p<0.001 and p=0.01, respectively). The expected increase in odds of having a pathologic pressure gradient is 1.14 times higher for each cm of water increase in OP (OR=1.14; 95% CI (1.07 to 1.24)). A woman is 7.2 times more likely to have a pathologic adjacent pressure gradient than a man (OR=7.2; 95% CI (1.64 to 40.25)).
Discussion
This study was designed to quantitatively describe venous pressures and gradients throughout the cerebral and cervical venous systems in patients with IIH. With a sample size of over 100 patients, this study is one of the largest samples to date of patients with IIH who have undergone venography and manometry. A number of important, new observations are described. Among all patients with IIH, SSS pressures are highly correlated with intracranial CSF pressures as measured by OP on LP. Higher OP is also highly predictive of having a pathologic adjacent anatomic pressure gradient, with roughly one-quarter of patients having an OP of ≤24 harboring a pressure gradient ≥8 mm Hg compared with over three-quarters of patients with an OP of ≥35. In all patients, there is a consistent venous pressure decrease from the SSS to CVP, with a median decrement of approximately 1 mm Hg with each progressive point of caudal measurement. In those patients with ‘normal’ OP and absence of an adjacent anatomic pressure gradient of ≥4 mm Hg, a total cranial gradient amounting to a decrease of 3 mm Hg is seen from the SSS to the IJ bulb, with an overall pressure gradient from SSS to CVP measuring a median of 4 mm Hg.
There is a relative consensus among neurointerventionalists that a pressure gradient of ≥8 mm Hg should be present when selecting patients for venous sinus stenting.5 It is important to note, however, that this threshold was arbitrarily selected by Ahmed et al 13 without studies validating it and without outcome studies supporting the magnitude of the gradient. Some authors have reported stenting pressure gradients as low as 4 mm Hg.14–17 Many interventionalists use the presence of ≥50% stenosis of the venous sinuses on non-invasive venographic imaging, such as magnetic resonance venography, before selecting patients for venography and venous manometry. We recently demonstrated that non-invasive venography has very low negative predictive value in ruling out patients with pathologic trans-stenosis pressure gradients18 and therefore do not use these imaging modalities in selecting patients for invasive manometry. The sample of patients in this study therefore represents an ‘all-comers’ group of patients with IIH with medically refractory IIH but without preselection for venography. Importantly, in the absence of magnetic resonance venography screening more than 50% of patients were found to have a transverse-sigmoid sinus pressure gradient of at least 8 mm Hg, with only 1% patients having an 8 mm Hg adjacent anatomical pressure gradient at any other location (4% total). This study indicates that physiologic venous outflow obstruction, almost always at the transverse sinus, may be a contributing factor to elevated intracranial pressures in over 50% of patients being evaluated for surgical management of IIH. It is important to note that another large study of patients with IIH reported pathological pressure gradients in just over one-third of patients, which may be a reflection of varying referral and practice patterns at different institutions.19
These data also strongly suggest that the presence of an adjacent anatomical pressure gradient of ≥8 mm Hg can be predicted based on simple patient presenting factors, which can help clinicians to counsel patients and select patients for invasive manometric testing. Opening pressure is highly predictive of the presence of a pathologic pressure gradient, with a 14% increase in odds for every 1 cm of water elevation. Furthermore, women have a seven times greater odds of harboring a pathologic pressure gradient than men, although it should be noted that men comprised only 13% of the total sample in this study. In the series by Levitt et al, gender was not found to be a significant predictor of a pathologic pressure gradient during venography.19 BMI was expected to be a predictor of a pathologic pressure gradient, but this was not the case. When BMI is low, CVP would be expected to be low, and venous stenosis would therefore be expected to contribute more to elevated venous pressures in those with higher OP. In fact, Raper and colleagues previously demonstrated that patients with higher BMI were more likely to have higher pressure gradients and greater improvements in the gradient following stenting than those with lower BMI.20 No such relationship was identified in this study as BMI had little influence in the regression model (p=0.64). Finally, using the formulas provided, both CVP and SSS pressures can be predicted using BMI and OP, respectively, which can help to identify patients with a higher chance of having a pathologic venous sinus outflow obstruction amenable to stenting.
Published data on normal venous pressures and normal pressure gradients in individuals without elevated intracranial pressure are limited. These data are desperately needed to aid in our understanding of the pathophysiology of intracranial venous hypertension and the resultant manifestations of IIH. Levitt et al previously reported that none of the 11 patients with intracranial pressures of ≤24 cm H2O had pathological pressure gradients at the time of venography; however venous sinus pressure measurements in these individuals were not reported.19 By selecting patients with an OP of ≤20 cm H2O and all adjacent pressure gradients of ≤4 mm Hg (group C), we attempted to capture patients at the milder end of the IIH disease spectrum who might have venous pressure measurements approximating (or at least closer to) those in normal individuals. Although limited by the small sample size, the data from these three patients are homogeneous and potentially suggest that in normal individuals SSS pressures should probably be less than 16–18 mm Hg with overall pressure gradients of <8 mm Hg and total cranial gradients <5 mm Hg. Even in these individuals, there is a median adjacent anatomic gradient approximating 1.0 mm Hg across most measured locations. As normal venous pressure data have not yet been clearly established, these data may help to describe venous sinus pressures and gradients in individuals with normal or near-normal intracranial pressures.
Pathologic trans-stenosis pressure gradients may occur in patients with venous sinus stenosis measuring as little as 30–35% on venous phase arteriography or venography.18 As there is no formal, widely accepted method of calculating percent stenosis on two-dimensional venous phase arteriography or venography and the gold standard of determining candidacy for stenting remains the trans-stenosis pressure gradient, this study specifically avoided using angiographic measurements of the venous sinuses or evaluating the degree of angiographic stenosis as a clinical outcome.
This study has other important limitations. The patients represent a single-center practice which may represent a different patient population than centers elsewhere. Data on the concurrent use of acetazolamide, furosemide, or other potential intracranial pressure-lowering agents were not considered. Early in the patient series CVP was not recorded; however, all data collection was otherwise standardized using the same microcatheter, measurement landmarks, and conscious sedation dosages.
Conclusions
This study describes the magnitude of pressure decrements across anatomical locations throughout the venous system in over 100 patients with IIH. OP on LP is strongly predictive of SSS pressures and also of the presence of a pathologic pressure gradient; approximately one-quarter of patients with a normal OP (≤24) have a pressure gradient ≥8 mm Hg compared with over three-quarters of patients with an OP of ≥35. Across the IIH disease spectrum, a normal pressure drop of approximately 1 mm Hg occurs with progressively more caudal transition across anatomic points of measurement.
References
Footnotes
Contributors Concept design: KMF. Data collection: RMG. Data review: KMF, CK, SQW. Manuscript composition: KMF, SQW, RMG. Final approval of article: all authors.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Ethics approval IRB00042737.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement There are no additional data available.
Patient consent for publication Not required.