The Effect of Vertebral Venous Collaterals on Intracranial Hypertension and Related Symptoms in Patients with Bilateral Transverse Sinus Stenosis

Background: Vertebral venous collaterals (VVC) were often found in patients with bilateral transverse sinus stenosis (BTSS). The purpose of this study was to investigate the physiological role of VVC in BTSS patients. Methods: The index of TSS was used in the assessment of BTSS severity. Subjects underwent a standard lumbar puncture to measure the intracranial pressure (ICP). Papilledema and tinnitus were evaluated by using Frisén's grade and questionnaires for tinnitus handicap inventory (THI), respectively. The intensity and impact of headache was assessed by using 10-point Numeric Pain Rating Scale (NPRS) and six-item Headache Impact Test (HIT-6). Results: BTSS group had more patients with severe intracranial hypertension (IH) and less patients with normal ICP. BTSS patients had higher ICP than normal controls. Further analysis on VVC showed that VVC decreased the elevated ICP in BTSS patients and exerted no impact on ICP in normal controls. A similar incidence of VVC in BTSS patients and normal controls were found. BTSS patients with normal ICP and mild IH exhibited a higher incidence of VVC compared with those with severe IH. VVC, rather than the severity of BTSS, regulated the ICP. Furthermore, VVC alleviated IH-related clinical manifestations including papilledema and tinnitus in BTSS patients. Conclusions: The present study demonstrated BTSS is correlated with IH. The presence of VVC, which is postulated to be congenitally formed, normalizes the elevated ICP and alleviated IH-related symptoms in BTSS patients. VVC may be used as an indicator of relatively low ICP in BTSS patients.


Background
Bilateral transverse sinus stenosis (BTSS) is characterized by narrowing of bilateral transverse sinus which resulted in obstruction of venous re ux [1]. The majority of BTSS patients exhibited with intracranial hypertension (IH) which manifested as a triad of headaches, tinnitus and papilledema [2,3].
Stenting improved the symptoms in patients with transverse sinus stenosis (TSS) and reduced the elevated intracranial pressure (ICP) [4]. It is suggested that BTSS is one of the etiologies of IH.
Since an alternative pattern of venous re ux was observed in patients with occluded superior sagittal sinus, it is postulated that altered pattern of venous re ux also exists in the presence of BTSS [5]. The drainage of cerebral veins consists of two major pathways: transverse sinus-sigmoid sinus-internal jugular vein and vertebral venous plexus [6]. Vertebral venous plexus is an extensive paravertebral system that provides direct venous communication between peritoneum and cranial cavity [7]. The presence of vertebral venous collaterals (VVC) was reported in 108 consecutive patients [6]. However, whether VVC is of physiological signi cance in BTSS patients remained unknown. This article was aimed to investigate the physiological role of VVC and whether it is involved in the regulation of IH and related clinical symptoms in BTSS patients.

Methods
Subjects recruitment BTSS patients were recruited from the neurology department and neurosurgery department of Xuanwu Hospital from January 2014 to December 2018. The normal controls were recruited from health examination center of Xuanwu hospital. This study was approved by the Ethnic Board of Xuanwu Hospital.
The inclusion criteria were BTSS con rmed by two of magnetic resonance venography (MRV), computed tomography venography (CTV) or digital subtraction angiography (DSA). Normal controls had no history of neurological diseases and no neurological signs.
The exclusion criteria involved: 1) systemic in ammation; 2) medicine related intracranial hypertension; 3) moderate to severe stenosis in intracranial or jugular veins other than transverse sinus; 4) moderate to severe stenosis in intracranial, carotid or vertebral arteries; 5) intracranial lesions.

Experimental Design
Subjects underwent a standard lumbar puncture to measure ICP. It is reported that 95% reference interval for ICP was 100 to 250 mmH2O in healthy adults [8]. ICP above 250 mmH2O was considered IH [9,10].
However, there is an overlap between normal and elevated ICP ranging from 200 to 250 mmH2O [11,12]. In this article, Normal ICP was de ned as 100 to 200 mmH2O. Mild IH was de ned as an ICP ranging from 200 to 250 mmH2O whereas severe IH was de ned as an ICP above 250 mmH2O.
Triad of IH were also evaluated. Fundus photography was conducted and the papilledema was graded by using modi ed Frisén's grade [13,14]. The intensity of headache was assessed by using 10-point Numeric Pain Rating Scale (NPRS) [15]. The six-item Headache Impact Test (HIT-6) was used to measure the impact of headache [16]. The impact of tinnitus was evaluated by using questionnaires for tinnitus handicap inventory (THI) [17].
The index of TSS (ITSS) score was a useful tool for the assessment of BTSS severity. The degree of stenosis was graded from 0 to 4 based on the following scale: 0 = normal; 1 = stenosis up to 1/3; 2 = stenosis between 1/3 and 2/3; 3 = stenosis >2/3; and 4 = hypoplasia. ITSS was calculated as degree of right TSS × degree of left TSS [18].

Statistical analysis
All clinical data were collected by Dr. Chaobo Bai and Dr. Jingkun Sun. Dr. Min Li independently analyzed the clinical data in a blinded manner. The data are expressed as mean ± standard error (SE). Rates were compared by using chi-square test. Numerical values between two and three groups were compared by using Student's t test and one way-Analysis of Variance (ANOVA) followed by Student-Newman-Keuls (SNK) test, respectively. Correlations were performed with Pearson correlation coe cients or linear regression. P < 0.05 was de ned as statistically significant. All analyses were performed using the SPSS software (version 19.0, SPSS, USA) and GraphPad Software (Version 7.0, GraphPad Prism, USA).  [19][20][21]. Data from the present study con rmed the aforementioned results. BTSS patients had higher ICP compared with normal controls (Fig. 1A). ICP was signi cantly lower in BTSS patients with VVC compared with those without VVC. No signi cant difference in ICP was found between normal controls with VVC and those without VVC (Fig. 1A). It is suggested that VVC decreased the elevated ICP in BTSS patients and exerted no impact on ICP in normal controls. The incidence of VVC between BTSS patients and normal controls were compared. 51.6% of BTSS patients and 48.6% of normal controls had VVC (Fig. 1B). Interestingly, BTSS patients with normal ICP and mild IH exhibited a higher incidence of VVC compared with those with severe IH (Fig. 1B). These ndings indicated that severe IH in BTSS patients may be due to the absence of VVC.

Results
Since BTSS was correlated with IH, it is hypothesized that severity of BTSS may in uence the ICP.
However, further analysis revealed that ICP was not affected by the severity of BTSS graded by ITSS score. VVC was found to be the only factor that in uence the ICP (Table 2). Furthermore, we assessed the correlation between ITSS score and incidence of VVC. Results showed that incidence of VVC was irrelevant with ITSS score (Fig. 2). It is demonstrated that VVC, rather than the severity of BTSS, plays the key role in regulating ICP. The correlation between VVC and triad of IH symptoms was analyzed. The papilledema was assessed by Frisén's grade. Pearson correlation analysis showed that ICP was positively associated with the score of Frisén's grade (Fig. 3A). BTSS patients with VVC had signi cantly lower score of Frisén's grade than those without VVC (Fig. 3B). However, VVC exerted no in uence on headache either assessed by NPRS or HIT-6 score in BTSS patients (Table 3). THI was used to evaluate tinnitus. BTSS patients with VVC had signi cantly lower THI score than those without VVC (Table 3). Taken together, the presence of VVC attenuates papilledema and tinnitus in BTSS patients.  [22]. In this study, BTSS group had more patients with severe IH and less patients with normal ICP (Table 1). Consistent with previous ndings [11,12], our results also demonstrated that ICP ranging from 200 to 250 mmH2O is an overlap between normal and elevated ICP (Table 1). BTSS patients had an average ICP of 246.6 ± 80.7 (Fig. 1A). It is demonstrated that BTSS is associated with elevated ICP.
The presence of VVC was observed in our clinical practice. However, it has been very confusing that whether these collaterals were congenitally formed or acquired after BTSS. A higher incidence of VVC in BTSS patients compared with normal controls and in BTSS patients with elevated ICP compared with those with normal ICP suggests that VVC may be acquired after BTSS. Otherwise, it is indicated that VVC may be congenital. The present study showed a similar incidence of VVC between BTSS patients and normal controls. Moreover, the incidence of VVC in BTSS patients with severe IH was no higher than those with mild IH and normal ICP (Fig. 1B). A plausible explanation for this observation is that VVC was a congenital and physiological existence.
In the meanwhile, the physiological role of VVC has not been demonstrated before. Results of this study revealed that VVC is capable of down-regulating the elevated ICP and exerted no effect on normal ICP (Fig. 1A). A negative correlation between VVC and ICP was identi ed after adjusted for the severity of BTSS (Table 2). Under the condition of BTSS, more venous blood drains into vertebral venous plexus. VVC allows more intracranial venous blood to pass through, thus guarantees the venous return. The presence of VVC compensated the obstruction of transverse sinus-sigmoid sinus-internal jugular vein pathway. However, the severity of BTSS was proofed to be irrelevant with ICP after adjusted for VVC (Table 2). This nding was evidenced by Bono et al [22]. This lack of association may be due to the compensatory effect of vein of Labbe and occipital sinus [23,24].
In agreement with the regulatory effect on ICP, VVC alleviated IH-related clinical manifestations. Tinnitus, papilledema and headaches were the three major symptoms of IH [3]. VVC improved papilledema and tinnitus. However, headache was not relieved (Table 3). Since NPRS and HIT-6 is subjective, scores given by patients with distinctive pain tolerance may be inaccurate. It is indicated that NPRS and HIT-6 are more suitable in the application of self-control studies.

Conclusions
The present study demonstrated BTSS is correlated with IH. The presence of VVC, which is postulated to be congenitally formed, normalizes the elevated ICP and alleviated IH-related symptoms in BTSS patients. The presence of VVC may be used as an indicator of relatively low ICP in BTSS patients.

Consent for publication
All patients reported in this case series signed a letter of consent for publication.

Availability of data and materials
The data that support the ndings of this study are available from the corresponding author via E-mail upon reasonable request.

Competing interests
All authors declare that they have no con icts of interest.  The presence of VVC reduced the elevated ICP in BTSS patients. The BTSS patients revealed a higher ICP than normal controls. and BTSS patients with VVC exhibited a lower ICP than those without VVC (A). The incidence of VVC was signi cantly higher in BTSS patients with normal ICP or mild IH than those with severe IH (B). *p < 0.05, **p < 0.01, ***p < 0.001. BTSS, bilateral transverse sinus stenosis; ICP, intracranial pressure; IH, intracranial hypertension; VVC, vertebral venous collaterals.

Figure 2
There was no correlation between the ITSS score and incidence of VVC. ITSS, index of transverse sinus stenosis; VVC, vertebral venous collaterals.