Pathogenesis

Spinal CSF leak symptoms are believed to occur as a result of loss of intracranial CSF volume rather than strictly a reduction in pressure. This is supported by the fact that many patients exhibit normal CSF opening pressures (or even occasionally slightly elevated pressures).

In the spontaneous form of this condition, CSF volume loss occurs by three major mechanisms in the spine. These include leaks resulting from

• •
  

  Dural weakness in the nerve root sleeve or a meningeal diverticulum.

  
  
• •
  

  Ventral dural tears associated with degenerative disc disease/osteophytes/calcified disc herniations.

  
  
• •
  

  CSF-venous fistulas, whereby CSF drains directly into a venous structure through abnormal fistulization, usually between a meningeal diverticulum and the venous system.

In the iatrogenic form of this condition, a dural tear occurs at the site of the spinal procedure or surgery, leading to loss of CSF volume. In patients with hydrocephalus and a shunt, overdrainage may occur when CSF is drained more quickly than it is produced. Shunt overdrainage is characterized by the appearance of severe headache in the setting of normal or smaller-than-normal ventricles. This can present in different ways, one of which is symptoms of intracranial hypotension.

Types of Spontaneous Spinal Cerebrospinal Fluid Leaks

Farb and colleagues describe a simple classification system of the different types of spontaneous spinal CSF leaks. Type 1 and Type 2 leaks demonstrate a spinal longitudinal epidural fluid collection (SLEC) on spinal imaging, best seen with heavily T2-weighted MRI of the spine with fat saturation. Type 1 and 2 leaks are therefore called SLEC positive (SLEC-P). Type 3 and 4 leaks are SLEC-negative (SLEC-N).

• •
  

  Type 1—ventral dural tear (due to osteophyte/calcified disc). SLEC-P because their ventral location spills CSF through the dural tear into the epidural space ( Fig. 1 ).

  
  

  
  

  
  

  

  
  

  
  

  Ventral dural leak (Type 1).

  
  

  ( Courtesy of Farnaz Amoozegar, MD, MSc, FRCPC and Bahareh Cloutier.)

  
  
  
  
• •
  

  Type 2—proximal nerve root sleeve tear or tear within a weakened area of a proximal meningeal diverticulum. SLEC-P because their proximal location also spills CSF into the epidural space ( Fig. 2 ).

  
  

  
  

  
  

  

  
  

  
  

  Proximal (Type 2) and distal (Type 4) nerve root sleeve dural tears. ( Courtesy of Farnaz Amoozegar, MD, MSc, FRCPC and Bahareh Cloutier.)

  
  
  
  
• •
  

  Type 3—CSF-venous fistula. CSF within a pseudomeningocele directly communicates with a vein or venous plexus through an abnormal fistula. Leaked CSF does not enter the epidural space (SLEC-N) ( Fig. 3 ).

  
  

  
  

  
  

  

  
  

  
  

  CSF-venous fistula (Type 3).

  
  

  ( Courtesy of Farnaz Amoozegar, MD, MSc, FRCPC and Bahareh Cloutier.)

  
  
  
  
• •
  

  Type 4—Distal nerve root sleeve dural tear. SLEC-N as leaked CSF flows more peripherally (see Fig. 2 ).

The distinction of SLEC-P ( Fig. 4 A–C) or SLEC-N becomes important when considering subsequent imaging modalities for a patient with a spinal CSF leak.

SLEC in a patient with SIH. ( A ) Sagittal, T2-weighted image showing a hyperintense signal in the dorsal epidural space, which is difficult to differentiate from epidural fat. ( B , C ) Sagittal ( B ), and transverse ( C ), T2-weighted fat-saturated images showing extensive CSF collection in the dorsal epidural space in the thoracic spine and ventral epidural CSF ( green arrows ) in the lower cervical spine. The posterior dura is visible on the sagittal and axial images ( orange arrows ).

( Reprinted with permission from Elsevier. The Lancet Neurology, 2022;21(4):369-380.)

Epidemiology

In the literature, the estimated annual incidence of spontaneous spinal leaks presenting to neurologic attention is about 3.8 to 5 per 100,000. However, this is believed to be an underestimation as not all patients present to medical attention, may not be referred to the neurologic specialties, and this condition is significantly underdiagnosed and misdiagnosed.

Clinical Presentation and Assessment

The most common symptom in patients with spinal CSF leaks is headache, though a minority of patients will not present with or develop headache during their illness. The prototypical or classic presentation is an orthostatic headache, which is a headache that worsens or becomes apparent with upright posture and resolves or lessens with recumbency. The headache can occur within seconds to minutes of taking an upright position or can take hours (so called second-half-of-the day headache). Improvement in headache usually (but not always) occurs within 30 to 60 minutes of becoming recumbent. The quality of the headache is often a pressure or pulling sensation. It is classically more posterior in location, such as the base of the skull and neck, but can be holocephalic, or located elsewhere. Typically, the headache will be diffuse and symmetric. ^,

The mechanism leading to the headache is thought to be related to loss of CSF buoyancy, resulting in downward displacement of the brain and traction of pain-sensitive structures, including the meninges and vasculature. Mechanosensitive ion channels in the meninges cause a stretch-induced nociception. In addition, there is compensatory dilation of pain-sensitive venous structures. In the upright position with gravity pulling down, these mechanisms are intensified, leading to the orthostatic nature of the headache. However, it is important to keep in mind that an orthostatic headache can become less prominent or change over time. So, it is always important to take a careful history and ask the patient the qualities of their headache when it first started. ^{, ,}

Some patients may also have an atypical headache presentation, such as no orthostatic features to their headache, exertional headaches, or rarely paradoxical headaches (worse when supine). In patients with atypical presentations, other diagnoses must be ruled out and there should ideally be imaging support before concluding that the symptoms are secondary to a spinal CSF leak. ^,

Headaches related to spinal CSF leaks usually present in a subacute or chronic manner but can occasionally present as a thunderclap headache. So, spinal CSF leaks must be considered on the differential diagnosis of thunderclap headache. In addition, a new headache that starts and persists (a new daily persistent headache [NDPH]) must also include spinal CSF leak on the differential diagnosis. ^,

Other common symptoms of spinal CSF leaks can include posterior neck pain or stiffness, nausea and vomiting, cognitive impairment, fatigue, changes in hearing (described as muffled hearing, like being under water), nonpulsatile tinnitus, and a disturbed sense of balance. These symptoms relate to irritation or traction of the meninges, stretching of the various cranial nerves (such as the eighth nerve complex) or transmission of CSF pressure changes to that in the perilymph (in regards to the auditory symptoms). ^{, ,}

Many other symptoms can occur less commonly. These can include visual symptoms such as visual blurring or diplopia, sensory disturbances, parkinsonism, ataxia, dementia, altered level of consciousness or coma, and bibrachial amyotrophy. ^{, ,}

When taking the history from the patient, it is essential to ask questions regarding headache onset, characteristics, and the orthostatic nature of the headache, in addition to eliciting other symptoms. Patients should also be asked if they have a personal or family history of connective tissue disorders (CTD) as an inherited CTD can be a risk factor for the development of spinal CSF leaks. In particular, Marfan syndrome, Ehlers-Danlos (type II), and autosomal dominant polycystic kidney disease are the most commonly associated conditions. ^{, ,} Familial cases of spinal CSF leak have also been described, though rare. Furthermore, patients should be asked if they have received any spinal surgeries or procedures in the past, including lumbar punctures and epidural anesthesia.

The formal diagnostic criteria from the International Classification of Headache Disorders (ICHD-III) for headache associated with intracranial hypotension indicates that there must be evidence of low CSF pressure (<60 mm CSF) and/or signs of CSF leakage on imaging, and headache must develop in temporal relation to the CSF leakage. Although these criteria have very high specificity, they have low sensitivity as they can miss patients that have normal CSF opening pressures and/or negative MRI scans of the brain and spine. So, one must be aware that not all patients will fulfill these criteria.

Not all orthostatic headaches are due to spinal CSF leaks. The differential diagnosis includes postural orthostatic tachycardia syndrome (POTS), cervicogenic headache, and various forms of cervical spine disease or instability, and conditions that may affect spinal lumbosacral compliance (a baggy/overcompliant lumbosacral compartment, for eg, due to a sacral perineural cyst). ^{, ,}

Investigations

Magnetic resonance imaging brain

Once there is clinical suspicion of a spinal CSF leak, an enhanced magnetic resonance imaging (MRI) of the brain should be performed. In about 80% of patients, there will be signs of a spinal leak on the MRI of the brain; hence, about 20% of patients may have a negative MRI of the brain. ^, Qualitative and quantitative signs of a spinal leak may be seen. The classic qualitative signs can be remembered by the acronym SEEPS:

• •
  

  S ubdural fluid collections

  
  
• •
  

  E nhancement of Pachymeninges

  
  
• •
  

  E ngorgement of Venous Structures

  
  
• •
  

  P ituitary hyperemia

  
  
• •
  

  S agging of the brain

In terms of the subdural fluid collections, these are most commonly hygromas, but hematomas can also occur. These are usually bilateral and thin, without significant mass effect, but large ones with mass effect do occur occasionally. The cerebral convexities are the most common location, but they can also be present over the cerebellar convexities and retroclival region. ^{, ,}

Enhancement of the pachymeninges is usually diffuse, nonnodular, and secondary to dilation of small thin-walled blood vessels in the subdural zone. An MRI of the brain with enhancement is the preferred test to demonstrate this. ^{, ,}

Engorgement of venous structures can be visualized with routine MRI of the brain or venography. The venous distension sign assesses the inferior margin of the midportion of the dominant transverse sinus. Normally, on T1-weighted sagittal views, this margin shows a concave or straight configuration. In patients with spinal CSF leaks, it may assume a distended convex configuration. Sensitivity and specificity of this sign for spinal CSF leaks is quite high at 94%.

Pituitary hyperemia is seen as a plump looking pituitary with it protruding beyond the sella. ^{, ,}

Sagging of the brain can be seen in various ways, including perichiasmatic cistern effacement with bowing of the optic chiasm over the pituitary, cerebellar tonsillar descent (which can mimic a Chiari type I malformation[CM1]), prepontine cistern effacement, flattening of the pons against the clivus, and a general crowding of the posterior fossa. ^{, ,}

When patients demonstrate cerebellar tonsillar descent, it is important to distinguish this from a true CM1, to ensure that posterior fossa decompression surgery is not unnecessarily provided. In spinal CSF leak, there will usually be other signs of brain sag, which will not be the case for a CM1. In addition, in CM1, the tonsils have a peg-like appearance, which is not typically seen in spinal leaks. Furthermore, there are quantitative measures that can help differentiate the two conditions. Houk and colleagues showed that a cutoff value of less than −15° for slope of the third ventricular floor and less than 45° for the pontomesencephalic angle strongly support a diagnosis of spinal CSF leak (sensitivity and specificity = 1.0).

Lastly, the clinical presentation is often helpful to differentiate CM1 from spinal leaks. In CM1, headaches occur with cough or valsalva manoeuvers, are occipital or suboccipital in location, are of short duration (usually less than 5 minutes), and do not have orthostatic features.

In regards to the quantitative signs of spinal CSF leak on MRI of the brain, the mamillopontine distance, the pontomesencephalic angle, the suprasellar cistern, and the prepontine cistern are additional measures. These measurements are reduced in patients with a spinal leak as compared to normal MRIs of the brain, with established cut-off values. ^, A recent study has shown that a lower optic nerve sheath diameter can also be predictive of a CSF-venous fistula in patients with otherwise normal brain MRIs.

The Bern score is a scoring system whereby imaging features (both qualitative and quantitative) with the best predictive value for a spinal CSF leak are assigned points and summated to provide a probability score that a patient may have positive findings of a spinal leak on subsequent -myelography ( Table 1 ). This scoring system is not meant to be diagnostic and does not take into account the clinical history or any other imaging modalities.

Table 1

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Hydrocephalus Pathophysiology and Epidemiology Adult Hydrocephalus Clinical Subtypes Guidelines for Diagnosis and Management of Idiopathic Normal Pressure Hydrocephalus Diagnostic and Prognostic Biomarkers of Idiopathic Normal Pressure Hydrocephalus in Cerebrospinal Fluid and Blood Temporary Drainage of Cerebrospinal Fluid for Diagnosis and Treatment of Hydrocephalus Idiopathic Intracranial Hypertension with Papilledema

Stay updated, free articles. Join our Telegram channel

Join