Intracranial findings in spontaneous intracranial hypotension: Does the severity of abnormalities correspond with certainty and/or multifocality of cerebrospinal fluid leaks?

Abstract

Background and Purpose

Spontaneous intracranial hypotension (SIH) is caused by spinal cerebrospinal fluid (CSF) leaks. This study assessed whether the certainty and/or multifocality of CSF leaks is associated with the severity of intracranial sequelae of SIH.

Materials and Methods

A retrospective review was completed of patients with suspected SIH that underwent digital subtraction myelogram (DSM) preceded by brain MRI. DSMs were evaluated for the presence or absence of a CSF leak, categorized both as positive/negative/indeterminate and single versus multifocal. Brain MRIs were assessed for intracranial sequelae of SIH based on two probabilistic scoring systems (Dobrocky and Mayo methods). For each system, both an absolute “numerical” score (based on tabulation of findings) and “categorized” score (classification of probability) were tabulated.

Results

174 patients were included; 113 (64.9%) were female, average age 52.0 ± 14.3 years. One or more definite leaks were noted in 76 (43.7%) patients; an indeterminate leak was noted in 22 (12.6%) patients. 16 (16.3%) had multiple leaks. There was no significant difference in the severity of intracranial findings between patients with a single versus multiple leaks (p values ranged from .36 to .70 using categorized scores and 0.22–0.99 for numerical scores). Definite leaks were more likely to have both higher categorized intracranial scores (Mayo p = .0008, Dobrocky p = .006) and numerical scores (p = .0002 for Mayo and p = .006 for Dobrocky).

Conclusions

Certainty of a CSF leak on diagnostic imaging is associated with severity of intracranial sequelae of SIH, with definite leaks having significantly more intracranial findings than indeterminate leaks. Multifocal leaks do not cause greater intracranial abnormalities.

Introduction

Spontaneous intracranial hypotension (SIH) is a condition in which a cerebrospinal fluid (CSF) leak leads to lower-than-normal CSF volume. ¹ Patients are typically adult women, and classically present with orthostatic headaches that worsen in an upright position and improve with lying down.^2,3 However, symptoms are highly variable between patients, and include neck pain, tinnitus, cranial nerve palsies, and cognitive impairment. ⁴

Nearly always, identified CSF leaks are located in the spine. ⁵ These leaks may be related to focal defects in the dura—for example, from spiculated osteophytes or related to nerve root diverticula—or secondary to CSF-venous fistulas. ⁶ Digital subtraction myelogram (DSM) and CT myelography (CTM) have become mainstay modalities to identify the presence and location of CSF leaks.^7–9 Interpretation of these examinations, however, is not always straightforward. Some patients have multiple leaks. Others have indeterminate findings, in which it is not certain whether a leak was definitively discovered.

Clinicians commonly use brain MRIs to evaluate for imaging evidence of SIH. Many intracranial abnormalities have been shown to be associated with SIH, including dural thickening and enhancement, subdural fluid collections, pituitary and/or dural venous engorgement, and “brain sag.” ¹⁰ The presence or absence of these findings have been used to create probabilistic scoring systems, in which one can assess the likelihood of a patient having a CSF leak based on brain MRI.^11,12

Anecdotally, patients with indeterminate findings on a DSM have milder intracranial abnormalities, while patients with multiple leaks may have worse intracranial findings. To date, however, no data to support these hypotheses exist. This study set out to assess whether the certainty and/or multifocality of CSF leaks is associated with the severity of intracranial findings in a cohort of patients with suspected SIH.

Materials and methods

Patient cohort

Following approval by the local institutional review board, a retrospective review was performed of consecutive patients that had undergone a digital subtraction myelogram (DSM) preceded by a brain MRI. All patients had clinically suspected SIH, and were imaged with DSM between 1/1/2020 and 30/11/2022. Exclusion criteria were the presence of a spinal longitudinal fluid extradural CSF collection (SLEC) on pre-DSM spinal MRI or images that were considerably degraded by artifact.

Digital subtraction myelography technique

Per our institution’s standard protocol, DSM imaging was performed over two consecutive days given the recommended maximum intrathecal administration of Omnipaque 300 per day. Patients were placed in opposite lateral decubitus positions for each exam, generally in the right lateral decubitus position on the first day and in the left lateral decubitus position on the second day. For each exam, patients were put on a custom-made cushion that positioned their hips superior to their heads, allowing for administered contrast to flow toward the head during imaging. Following placement of a 20-gauge spinal needle into the spinal canal, two sets of DSM imaging were performed with the flat panel detector centered over the upper and lower spine. All patients were then sent to CT for further imaging, while kept in the same lateral decubitus position. DSMs and subsequent CT myelograms were interpreted together as combined examinations.

MRI brain imaging

MRI imaging was performed on either a 1.5 T or 3.0 T scanner. All components of the Mayo score, and the majority of the components on the Dobrocky score were assessed on 3D fat-saturated post-contrast T1 Space images (TR = 600 ms, TE = 7.2 ms, slice thickness = 1 mm, flip angle = 120°, FOV = 250 × 250 mm²). Additionally, for the Dobrocky score, assessment of subdural fluid collections was performed by comparing post-contrast T1 Space sequences to axial FLAIR images (TR = 9000 ms, TE = 149 ms, flip angle = 180°, slice thickness = 4 mm, FOV = 220 × 220 mm²).

Image analysis

All DSMs and MRI brains were reviewed by four neuroradiologists. DSMs were evaluated for the presence or absence of leak, categorized as positive, indeterminate, or negative. Positive examinations were evaluated for the presence of additional/multiple leaks. Leak laterality was also noted.

Brain MRIs were assessed for each scored component of the Mayo score and Dobrocky score. For each scoring system, both an overall “numerical” score (absolute number) and a “categorized” score (e.g., high vs low risk) were tabulated. For the purposes of statistical analysis, any “indeterminate” cases of dural venous sinus engorgement were considered positive. For the Mayo score, assessed findings included: smooth dural enhancement, dural enhancement involving the internal auditory canals (IACs), pituitary engorgement (defined as convexity along the superior surface of the gland), non-Chiari cerebellar tonsillar descent >5 mm below the foramen magnum, effacement of the suprasellar cistern (≤2.5 mm), dural venous sinus engorgement, and “brain sag” defined as the cerebral aqueduct iter located below the incisural line. The presence of any of these findings was given 1 point for the Mayo scoring system; a score of 0–2 was considered to represent a low probability of a CSF leak, while ≥3 was considered intermediate to high probability.

For the Dobrocky score, findings included “major” findings (each of which was given 2 points if positive): dural enhancement, dural venous sinus engorgement, and suprasellar cistern effacement (≤4.0 mm) and “minor” findings (each of which was given 1 point if positive): subdural fluid collection, prepontine cistern effacement (≤5.0 mm), and decreased mamillopontine distance (≤6.5 mm). The Dobrocky scoring system is based on a 3-tier categorization: 0–2 points = low risk of CSF leak, 3–4 points = intermediate risk, and 5+ points = high risk.

Statistical analysis

Statistical analysis was performed with BlueSky Statistics software (Bluesky Statistics LLC, Chicago, IL, USA). Means and standard deviations were calculated for all continuous variables. A chi-squared test was used to evaluate for significant differences between categorical variables. Logistic regression analysis was used to evaluate for correlations between the numerical scores of the Mayo and Dobrocky systems and (1) the presence of single versus multiple leaks and (2) definite versus indeterminate leaks. Area under curve (AUCs) were compared between both scoring systems.

Results

Patient cohort

One patient was excluded due to inability to undergo MRI because of a neurostimulator. 174 patients therefore made up the final cohort, of which 113 (64.9%) were female, and average age was 52.0 ± 14.3 years. One or more definite leaks was discovered in 76 (43.7%) patients; indeterminate findings were observed in 22 (12.6%) patients (Figure 1). Of the 98 patients with a definite or indeterminate leak, 16 (16.3%) had multiple leaks (Figure 2). Of both definite and indeterminate leaks, the majority (61/98; 62.2%) were on the right side, while 29 (29.6%) were on the left side, 6 (6.1%) were bilateral, 1 (1.0%) involved the internal epidural venous plexus, and 1 (1.0%) was ventral.

Figure 1.

Example of an indeterminate CSF leak in a 54-year-old female. Left-side down digital subtraction myelogram (arrows on A) demonstrated faint curvilinear contrast in the left foramen at T8–9. Corresponding CT myelogram images (arrows on B) also showed contrast extending into the soft tissues from this foramen. The brain MRI (C) was essentially normal, without evidence of spontaneous intracranial hypotension.

Figure 2.

Example of multifocal leaks in a 38-year-old female with spontaneous intracranial hypotension (SIH). Digital subtraction myelogram images noted CSF-venous fistulas both on the left at C1–2 (arrows on A), and on the right at T8–9 (arrows on B). The preceding brain MRI (C) showed multiple sequelae of SIH, including brain sag (dashed oval), cerebellar tonsillar descent (solid straight arrow), suprasellar cistern effacement (dashed straight arrow), and pituitary engorgement (curved arrow).

Single versus multiple leaks

Comparisons of categorized Mayo and Dobrocky scores to the presence of multiple leaks are detailed in Table 1. No significant associations were noted between the categorized scores and the multiple leaks.

Table 1.

Comparisons of categorized Mayo and Dobrocky scores to the presence of multiple leaks.

Similarly, using numerical values from each score: no significant association was found between the Mayo score and single versus multiple leaks (OR = 1.2, 95% CI: 0.9–1.5) (p = .22). Nor was there an association between the Dobrocky numerical score and single versus multiple leaks (OR = 1.0, 95% CI: 0.8–1.2) (p = .99). The AUC for the Dobrocky score was 0.50, and the AUC for the Mayo score was 0.60, with no significance difference between the scoring systems (p = .10).

Definite versus indeterminate leaks

Results using categorized Mayo and Dobrocky scores are detailed in Table 2. Both high probability of a leak using the Dobrocky score (p = .006) and intermediate to high probability of a leak using the Mayo score (p = .0008) were associated with the presence of a definite leak.

Table 2.

Comparisons of categorized Mayo and Dobrocky scores to the presence of definite leaks.

Using numerical values from each score: there was a significant correlation between the Mayo overall score and the presence of a definite leak (OR = 1.6, 95% CI: 1.2–2.1) (p = .0002). There was also a significant association between the overall Dobrocky score and the presence of a definite leak (OR = 1.3, 95% CI: 1.1–1.5) (p = .006). AUC for the Dobrocky score was 0.69, and AUC for the Mayo score was 0.74, with no significant difference between the scoring systems (p = .15).

Discussion

This study sought to determine whether the severity of intracranial abnormalities in patients with SIH is associated with the certainty and/or multifocality of CSF leaks. The results indicate that intracranial findings are not different among patients with single versus multiple CSF leaks. However, patients with a “definite” leak have greater intracranial abnormalities than those with “indeterminate” findings.

The lack of association between intracranial abnormalities and multifocality of CSF suggests that the presence of multiple leaks does not impact the degree to which SIH impacts intracranial structures. The reason for these findings is uncertain. It’s possible that multiple leaks do not change the degree of CSF volume and/or pressure to a greater extent than a single leak. It is also possible that intracranial sequelae of SIH are not inherently tied to the severity of the disease process.

Conversely, there are three major hypotheses to explain to the observed association between “definite” leaks and worse intracranial findings. First, it is possible that at least some of the “indeterminate” cases did not represent true leaks. Second, the relative normalcy of some brain MRIs may have biased the reviewers to interpret the results as being “indeterminate” rather than “definite.” Finally, it is also possible that “indeterminate” leaks contribute to slower, less robust leakage of CSF. However, this latter hypothesis seems less likely, given the lack of association between the multifocality of leaks and intracranial findings.

Nevertheless, the results of the current study are not necessarily surprising. It has long been known that CSF leaks from solitary sites (e.g., following a lumbar puncture) can cause orthostatic symptoms. ¹³ Lumbar punctures can also cause intracranial abnormalities on MRI imaging. Mark et al., for example, found that 5.2% of patients had intracranial subdural fluid collections after lumbar punctures, while 3.9% had diffuse dural enhancement. ¹⁴ Therefore, it is possible that any disruption in the dural—either single or multifocal—can alter the pressure hemostasis within the thecal sac enough to cause both considerable symptoms and imaging abnormalities.

Neither leak certainty nor multifocality has been previously studied in terms of their impact on intracranial abnormalities. Instead, prior reports on these subjects have focused on other associations. Schwartz et al., for example, for that 3.8% of patients with a CSF leak had multifocal leaks. The authors of that study also noted that patients with multifocal leaks had a greater average body mass index, were typically women, and were more often African American. ¹⁵

This study has several limitations, including its retrospective methodology. The classification of “indeterminate” versus “definite” leaks was based on visual analysis of DSMs and CTMs, rather than surgical confirmation. Also, as above, it is possible that the observation of intracranial findings biased reviewers in terms of their assessments of whether a leak was definite or indeterminate. Finally, this cohort specifically assessed patients with so-called “slow” CSF leaks, in which the patients lacked a spinal longitudinal extradural fluid collection. As such, it is possible that the results would have been different with a different cohort of patients.

Conclusions

Certainty of a CSF leak on diagnostic DSM and CTM imaging is associated with severity of intracranial sequelae of SIH, with definite leaks having significantly more intracranial findings than indeterminate leaks. Intracranial findings are not significantly different among patients with single versus multifocal leaks.

ORCID iDs

John C Benson https://orcid.org/0000-0002-4038-5422

Ian T Mark https://orcid.org/0000-0002-4036-2992

Ajay A Madhavan https://orcid.org/0000-0003-1794-4502

Jared T Verdoorn https://orcid.org/0000-0002-1592-1182