CSF Hypotension and CSF Leaks

Imaging and Therapy
  • Joanna BladowskaEmail author
  • Daniel J. Warren
  • Mario Muto
  • Charles Anthony Józef Romanowski
Living reference work entry

Latest version View entry history


Intracranial hypotension (IH) is a rare, still not fully understood, and usually a self-limited condition due to low cerebrospinal fluid (CSF) pressure. The clinical symptoms typical for IH syndrome are described. An orthostatic headache is the main clinical finding. This chapter reviews the radiological signs, which may be associated with this disorder, including the brain, as well as spinal findings, such as intracranial pachymeningeal enhancement, subdural effusions/collections or hemorrhage, rostral–caudal brain displacement, enlargement of the pituitary gland, spinal epidural fluid collections, and distension of the spinal epidural venous plexus. These radiological findings may be highly characteristic allowing the neuroradiologist to suggest the specific diagnosis. The imaging methods used for the diagnosis of IH and CSF leak sites, such as brain and spine MRI as well as CT and MR myelography, are discussed. The recommended radiological protocol is also presented.

After the failure of conservative and medical therapy, epidural blood patch (EPB) is the treatment of choice for moderate to severe intracranial hypotension. It consists of the injection of 10–20 ml of autologous blood into the epidural space. It is effective through two mechanisms of action: (a) elevating the pressure in the subarachnoid space by compressing the dura instantaneously and (b) forming a fibrinous clot that seals the dural hole.

Since intracranial hypotension may present with a variety of clinical symptoms and therefore could be often misdiagnosed, it is very important to consider this entity in the differential diagnosis when reporting the clinical neuroradiology cases. Patients misdiagnosed may be exposed to the unnecessary risk of starting treatment for diseases that mimic intracranial hypotension, including aseptic meningitis or pituitary disorders; therefore, every specialist in radiology should know and properly recognize this condition.


Intracranial hypotension Headache CSF (cerebrospinal fluid) leak Pachymeningeal enhancement Sagging brain Epidural blood patch Postdural puncture hypotension 



Constructive interference in steady state


Cerebrospinal fluid


Driven equilibrium


Diffusion-weighted imaging


Epidural blood patch


Fast imaging employing steady-state acquisition


Fluid attenuation inversion recovery


International Classification of Headache Disorders


Intracranial hypotension


New daily persistent headache


Postdural puncture hypotension


Postural orthostatic tachycardia syndrome


Rebound intracranial hypertension


Subdural hematoma


Spontaneous intracranial hypotension


Intracranial hypotension (IH) is a pathological condition caused by spinal leakage of cerebrospinal fluid (CSF) and typically characterized by an orthostatic headache. IH can be primary (spontaneous) or secondary (Limaye et al. 2016). Spontaneous intracranial hypotension (SIH) is presumed to occur from spontaneous CSF leak due to dural dehiscence or dural tears associated with nerve root sleeve cysts or disk herniations and osteophytes in the course of degenerative spine disease. Connective tissue disorders such as Marfan syndrome, Ehlers–Danlos syndrome Type II or autosomal dominant polycystic kidney disease, as well as neurofibromatosis and Lehman syndrome are also considered to be significant predisposing factors (Pattichis and Slee 2016; Lin et al. 2017). Secondary IH can be postoperative, associated with cranial or spinal surgery, spinal anesthesia or lumbar puncture, as well as following craniospinal trauma. IH as a consequence of dural sac puncture in both anesthesiology and interventional pain medicine practices has been named postdural puncture hypotension (PDPH). IH may occur as secondary entity to other pathological condition, such as dehydration or cases of decreased cerebral blood flow (Limaye et al. 2016).

SIH is a relatively new diagnosis that has only really been recognized by the majority of neuroradiologists and neurologists within the last 10–15 years. However, the clinical syndrome was first described by a German physician, Georg Schaltenbrand, in 1938, who reported “aliquarrhoea” connected to headache (Pattichis and Slee 2016).


The incidence of SIH is estimated to be around 2–5 per 100,000 per year; however, it has been believed that this disorder may be greatly underdiagnosed as misdiagnosis of SIH is quite common. Women are more often affected than men (2–5:1) (Lin et al. 2017; Limaye et al. 2016), and the majority of patients are in their third or fourth decade of life, although this entity may also occur in children and older people. Moreover, it has been reported that more females and persons under 40 years of age present with acute onset and more severe headache, while more males and older patients over 40 demonstrate subdural fluid collections on imaging and report longer duration of symptoms (Kranz et al. 2017)

Clinical Symptoms

The clinical syndrome consists of an orthostatic headache, generally occurring within seconds or minutes of standing up, and improves quickly in the supine position. Factors precipitating onset of symptoms include coughing, sneezing, choking, sex, straining, exercising, sport activities, positional changes, picking up objects, trivial falls, or chiropractic neck manipulation (Limaye et al. 2016). The headache may also worsen on coughing, laughing, and the Valsalva maneuver. It is noteworthy that headache, although considered to be the dominant clinical manifestation, is not omnipresent, and there have been reported SIH cases without headache or lacking typical orthostatic features (Pattichis and Slee 2016).

In general the clinical picture can be explained by nerve traction and mass effect on intracranial structures resulting from cranial changes due to the low CSF pressure. The reduced CSF pressure leads to postural headache sometimes associated with nausea and vomiting. Other presenting symptoms have included cranial nerve palsies such as unilateral or bilateral sixth nerve palsies, diplopia, transient visual obscuration, field defects, photophobia, dysgeusia, auditory symptoms (tinnitus, hearing loss, labyrinthine dysfunction), unilateral facial numbness and weakness, as well as radicular syndromes. Moreover, cough and dysphonia, coma or encephalopathy, as well as bulbar dysfunction and ataxia have been reported. Intracranial hypotension may also be associated with hormonal abnormalities, including hyperprolactinemia with subsequent galactorrhea (Pattichis and Slee 2016).

Clinical Evaluation

In patients presenting with headache that exacerbates in the upright position, intracranial hypotension syndrome should be considered in the differential diagnosis, particularly if the onset of symptoms is abrupt. Other disorders that should be taken into account include postural orthostatic tachycardia syndrome (POTS), cervicogenic headache, as well as other primary headache diseases such as new daily persistent headache (NDPH) (Kranz et al. 2017). Moreover, the pathologies that can cause positional headache need to be excluded such as postpartum period, venous sinus thrombosis, and subdural hematoma.

Headache is classified in the International Classification of Headache Disorders, 3rd Edition (ICHD-3), that established the diagnostic criteria for SIH (Table 1).
Table 1

International Classification of Headache Disorders, 3rd Edition (ICHD-3), criteria for the diagnosis of spontaneous intracranial hypotension (SIH). (Adopted from Lin et al. (2017))

ICHD-3 diagnostic criteria for SIH

A. Any headache fulfilling criterion B

B. Headache has developed in temporal relation to low CSF pressure or CSF leakage or has led to its discovery

C. Low CSF pressure (<60 mm H2O) and/or evidence of CSF leakage on imaging

D. Not better accounted for by another ICHD-3 diagnosis

Obviously, the final diagnosis of SIH has to be based on both clinical and radiological findings in conjunction with the medical history. Apart from that, CSF pressure measurement is a useful diagnostic tool to confirm suspected SIH. The opening pressure is characteristically low (less than 60 mm H2O, normal 65–195 mm H2O). Because the CSF in intracranial hypotension sometimes contains an abnormal leukocytosis (up to 200 cells/mm3) or elevated protein content (up to 1000 mg/dl), or both, suspicion of meningeal infection or tumor may lead to further evaluation. The CSF changes are thought to be due to meningeal hyperemia resulting from the low CSF pressure with diapedesis of cells into the subarachnoid space due to increased permeability (Limaye et al. 2016).

On the other hand, it should be underlined that a normal CSF opening pressure does not exclude a CSF leakage. Moreover, it has been reported that approximately 25% of patients with SIH, particularly those who are obese or those who have a long duration of clinical symptoms, may present a normal CSF pressure (Yao and Hu 2017).


It is commonly known that intracranial hypotension syndrome is strongly associated with CSF leakage, located in the cervical or thoracic spine in the majority of patients. However, the true pathogenesis of underlying spontaneous CSF leak unfortunately still remains uncertain.

It is believed that even trivial trauma can be responsible for SIH. Minor trauma, mostly related to downfall, has been reported in 80% of subjects. The predisposing factors include connective tissue diseases, malnutrition, or short stature (Lin et al. 2017).

Moreover, spontaneous intracranial hypotension is in fact a misleading term since a significant part of patients exhibit a normal CSF opening pressure as mentioned above. Therefore, it has been suggested that low intracranial CSF volume rather than CSF hypotension is the dominant cause of the intracranial hypotension syndrome (Kranz et al. 2017; Yao and Hu 2017). The decrease in CSF volume results from CSF leakage from the spine, which is caused by one of the following mechanism: leaks occurring due to dural weakness affecting nerve roots sleeves, ventral dural tears connected to disk herniations, or CSF venous fistulas (Kranz et al. 2017).

It should be stressed that SIH is not caused by a CSF leak at the level of the skull base. According to Schievink et al.’s experience with more than 200 patients with proven skull base leaks, none of them exhibited the clinical or imaging features of intracranial hypotension (Schievink et al. 2012). Furthermore, if patients who have a documented skull base CSF leak present with an associated sudden onset headache, a spinal source of CSF leak should be strongly suspected.

The underlying mechanism can be explained by the entity of gravity that causes a gradient of increasing CSF pressure along the spinal axis in the vertical position. As a result of that phenomenon, the intracranial pressure is slightly lower than atmospheric pressure when standing up. However, this gradient does not exist in the supine position. According to this physiology, spinal CSF leaks provoke orthostatic headaches, while CSF leaks at the level of the skull base do not cause such symptoms (Kranz et al. 2017).

Radiological Findings

The characteristic imaging findings in SIH may be explained by the Monro–Kellie hypothesis. According to this doctrine, the calvarium forms an enclosed space consisted of three compartments, including CSF, intracranial blood, and brain tissue. Since the total volume of the intracranial space is constant, any decrease of CSF volume would be followed by an increase in the volume of other compartment, such as intracranial blood volume in order to maintain the equilibrium state (Limaye et al. 2016; Holbrook and Saindane 2017). This statement precisely explains the typical radiological symptoms found on imaging in the course of SIH, such as dural thickening; subdural fluid collections, thought to result from engorged veins and dural interstitium; as well as enlargement of the venous sinuses and pituitary gland, followed by the sagging of the brain.


Pachymeningeal Enhancement

The most common finding in patients with intracranial hypotension is diffuse pachymeningeal enhancement following intravenous gadolinium administration (Kranz et al. 2017). This can be seen over the convexities, along the tentorium and clivus, and also within the cervical spine; it is usually florid and smooth and does not involve the depths of the sulci (Limaye et al. 2016). There is generally no enhancement around the brainstem either. The dural involvement is continuous with no segmentation or skip areas. The enhancement is linear, but areas of thickening consistent with localized fluid collections can be seen. The reason for the increased enhancement is thought to be due to a greater concentration of gadolinium chelate in the dural vasculature and in the interstitial fluid of the dura since the dural microvessels lack tight junctions, unlike the microvessels of the arachnoid membranes that are the components of the blood–brain barrier. They are therefore inherently “leaky” (Kranz et al. 2017).

T1-weighted contrast-enhanced images are the best technique to detect pachymeningeal enhancement (Fig. 1). However, FLAIR sequence is also worth mentioning as a sensitive method to delineate the thickened pachymeninges that appear hyperintense on FLAIR (Fig. 2). Moreover, FLAIR images can be very useful for the initial diagnosis as well as for follow-up examinations when there are contraindications to gadolinium injection (Forghani and Farb 2008).
Fig. 1

Brain MRI obtained in a 31-year-old female patient with SIH. Axial T2-weighted image (a) demonstrates bilateral slight widening of subdural spaces. Axial postcontrast T1-weighted image (b) reveals diffuse pachymeningeal enhancement

Fig. 2

Brain MRI, axial FLAIR image (a), and axial postcontrast T1-weighted image (b) performed in a 33-year-old woman with SIH. Observe the hyperintense slightly thickened pachymeninges (arrows) on FLAIR image (a) as well as typical smooth pachymeningeal enhancement after contrast administration (b)

It should be stressed that in the course of the disease in patients with chronic IH symptoms, pachymeningeal enhancement can disappear and thus hinder the correct diagnosis of IH.

Subdural Effusions/Collections/Hemorrhage

Subdural fluid collections have been reported in 50% of patients in the published papers (Limaye et al. 2016). Subdural exudate is seen overlying the cerebral hemispheres (Fig. 3) and characteristically beneath the tentorium cerebelli (Fig. 4). Additional locations of subdural collections have been reported, including parafalcine (Fig. 3), clival locations (Fig. 5), and also overlying the cerebellar hemispheres (Forghani and Farb 2008).
Fig. 3

Brain MRI examination performed in a 46-year-old male patient with SIH, axial T2-weighted (a), and axial FLAIR image (b). Bilateral subdural fluid collections are visible, presenting high signal intensity on FLAIR sequence (b). Observe also a very thin parafalcine collection

Fig. 4

In the course of SIH, subdural fluid collections can be appreciated beneath the tentorium cerebelli (arrows), particularly on sagittal (a) as well as coronal (b) T2-weighted images. Note also the subdural exudates overlying both the cerebral hemispheres (b)

Fig. 5

Brain MRI performed in a 46-year-old woman with SIH, axial T2-weighted (a), axial FLAIR image (b), and sagittal T2-weighted (c) images. The subdural fluid collection is located just behind the clivus (arrows). Observe the typical high signal intensity of the exudate on FLAIR image (b). Note also the decreased CSF in the optic nerve sheaths

Subdural fluid collections are usually bilateral and thin, most commonly located on the cerebral convexities, without exerting any mass effect on the gyri. The subdural effusions are due to proteinaceous oozing of fluid from the hyperemic dural, and this also explains why these are brighter than CSF on all pulse sequences. Alternatively the intense meningeal enhancement suggests leaky blood vessels, which can be leaky enough to produce effusions. The effusions are crescentic in configuration and located either below or in between enhancing membranes. Acute deterioration in patients can be due to large subdural hematomas which rarely occur in the syndrome. There have been reports that the headache disappears when such a subdural hematoma occurs due to normalization of the CSF pressure (Pattichis and Slee 2016). These larger acute spontaneous collections are thought to occur due to spontaneous rupture of the bridging cortical veins with resultant hemorrhage into the subdural space (Limaye et al. 2016). The subdural hygromas/collections rarely are large enough to require surgical evacuation.

As already mentioned above, FLAIR as well as PD sequence can be very useful in depicting even thin subdural effusions/hematomas as they demonstrate higher signal intensity compared to hypointense CSF (Fig. 4).

Rostral–Caudal Brain Displacement

The decreased CSF volume leads to rostral–caudal brain displacement called as brainstem “sagging” associated with crowding of the basal cisterns, low cerebellar tonsils location, as well as decreased CSF in the optic nerve sheath (Fig. 5). Midbrain “sagging” may also be observed consisting of flattening of the contour of the midbrain which goes hand in hand with the displacement of the iter, and the pons may be flattened against the clivus (Fig. 6a).
Fig. 6

Sagittal T2-weighted images showing the brainstem location in a patient with SIH (a) with comparison to a healthy person (b). There is a characteristic “sagging brain” appearance appreciated in a subject with SIH (a) with decreased mamillopontine distance as well as the pontomesencephalic angle. Observe the low-lying cerebellar tonsils presenting a normal morphology (a)

The low-lying cerebellar tonsils may create the “pseudo-Chiari” appearance and lead to misdiagnosis of Chiari I malformation (Holbrook and Saindane 2017). It should be pointed out that the low-lying cerebellar tonsils in the course of SIH typically demonstrate a normal morphology (Fig. 6), while in Chiari I malformation, the diagnostic criteria include not only the low location but also so-called peg-like configuration of the cerebellar tonsils.

There are some other quantitative signs on imaging helpful in the diagnosis of SIH, indicating the sagging of the brain, including the mamillopontine distance as well as the pontomesencephalic angle (Fig. 6). The mamillopontine distance means the distance measured from the inferior aspect of the mammillary bodies to superior aspect of the pons, which in normal people should be >5.5 mm.

The pontomesencephalic angle indicates the angle between two lines drawn along the anterior margin of the midbrain and the anterosuperior margin of the pons. A normal value of the pontomesencephalic angle is 65 ± 10°, and a value less than 50° is suggestive of intracranial hypotension (Shah et al. 2013).

Flattening of the Optic Chiasm and Pituitary Gland Enlargement

Downward displacement of the optic chiasm and hypothalamus is seen often with the chiasm being “draped” over the dorsum sellae. Reactive hyperemia of dural and epidural venous sinuses that enlarges the densely vascularized pituitary gland may be seen mimicking a pituitary tumor (Fig. 7) that can lead to misdiagnosis and even subsequent pituitary surgery. Hyperprolactinemia and galactorrhea have also been reported. Usually the pituitary enlargement results in pituitary height between 8 and 11 mm, and it has been observed to reverse earlier than meningeal thickening on follow-up MR examinations (Spero et al. 2011). Similarly, hyperprolactinemia also resolves following the effective treatment of SIH.
Fig. 7

MRI of the pituitary gland, sagittal (a) and coronal (b) contrast-enhanced T1-weithed images performed in a 36-year-old female patient who was admitted to the Department of Neurosurgery because of the pituitary adenoma reported in the previous MR examination. There is an enlargement of the pituitary gland mimicking the pituitary tumor. Note also the pachymeningeal enhancement visible along the clivus (a) as well as engorgement of both cavernous sinuses (b)

On the other hand, it should be stressed that although the pituitary enlargement is integrant, this sign is not always appreciated (Spero et al. 2011).

Small Pontine Hemorrhage

Petechial hemorrhages at the pontomesencephalic junction have been reported and represent Duret hemorrhages resulting from brainstem descent.

Slit Ventricles

Slit ventricles or frank ventricular collapse as well as indiscernible basal cisterns has been reported in patients with intracranial hypotension (Limaye et al. 2016).


Spinal Fluid Collections

The most common spinal findings in the course of SIH are fluid collections, reported in 67–100% of patients according to published case series (Medina et al. 2010). Spinal hygromas are noted in the epidural space over a long course, usually extended over five or more spinal segments (Fig. 8). They may be located either anterior or posterior to the dural sac. The presence of increased CSF protein in the collection raises the possibility that it is caused by a transudate due to spinal meningeal hyperemia (Yousry et al. 2001).
Fig. 8

MRI of the cervical spine, sagittal T2-weighted (a), fat-saturated sagittal (b), and axial (c) T2-weighted images obtained in a female patient with SIH. The epidural fluid collection can be appreciated in the anterior part of the cervical canal (arrows), extending over several spinal segments

Enlarged Epidural Veins

Engorged spinal epidural veins have been reported in the cervical region in patients with intracranial hypotension (Yousry et al. 2001). These giant epidural veins may be misinterpreted as jugular venous thrombosis or arteriovenous fistula; however, on spinal angiography no arteriovenous malformation can be identified. The spinal epidural veins provide collateral venous drainage from the intracranial compartment, and anastomoses between the intracranial and extracranial compartments are seen at the vertebral body. Therefore, the enlarged cervical veins mimic that engorgement intracranially. Engorgement of the anterior internal vertebral venous plexus can extend as far cranially as extending over the clivus (Yousry et al. 2001) and is best assessed on axial T2-weighted fat-suppressed or enhanced T1-weighted images (Medina et al. 2010). On axial images, the enlarged epidural venous plexus presents very characteristic morphology described as the festooned or light bulb appearance of the engorged plexus (Fig. 9). The festooning shape arises from the drop of the lateral aspects of the spinal sac with preserved midline attachment of the thecal sac to the posterior longitudinal ligament (Medina et al. 2010).
Fig. 9

Spine MRI obtained in a female patient with SIH, axial contrast-enhanced T1-weighted image shows the enlarged epidural venous plexus in the cervical spine that presents the characteristic festooned appearance

Dural Enhancement

Similarly to intracranial findings typical for SIH, spinal dural enhancement can be also observed after intravenous contrast administration, and it is supposed to be a consequence of dural vasodilatation and engorgement (Medina et al. 2010). The spinal enhancement is usually smooth and circumferential (Fig. 10). This symptom can be often seen with simultaneous intracranial dural enhancement; however, it is noteworthy that both findings are not always present in any one patient.
Fig. 10

Axial postcontrast T1-weighted images of the cervical spine performed in a woman with SIH showing subtle smooth and circumferential enhancement of the spinal dura

Extra-thecal CSF Collections

Fluid collections have been described on MR outside the spine, usually at the level of C1 and C2, often called as the C1–C2 sign or the C1–C2 false-localizing sign (Fig. 11). The CSF collection is usually situated between the spinous processes of C1 and C2. It is believed that it may well be a transudate related to the rich venous plexus in this region rather than a direct result of CSF leak, which means that the C1–C2 sign does not indicate the site of CSF leak. Changes are indistinguishable between SIH and postlumbar puncture headache, which further confirms this statement (Medina et al. 2010).
Fig. 11

Sagittal fat-saturated T2-weighted image in a female patient with SIH shows the CSF collection located between the spinous processes of C1 and C2 which is consistent with the C1–C2 false-localizing sign as there was no CSF leak confirmed at this level

This fluid collection may be difficult to appreciate as it has the same signal characteristic as fat on a T2-weighted and muscle on the T1-weighted image. Fat-suppressed T2-weighted images should be used (Fig. 11).

Other Findings

It has been reported that some abnormalities, such as nerve root cysts, pseudomeningocele, meningeal diverticula, as well as disk herniation or transdural osteophytes, can be associated with SIH. Furthermore, it has been suggested that patients with those abnormalities are prone to fail to respond to conservative therapy, and they may need reparative surgery. On the other hand, it should be emphasized that if a patient does not have symptoms, the presence of these structural abnormalities alone has not been confirmed to be a risk factor of development of CSF leak (Medina et al. 2010).

There are also other findings described in the course of SIH. Meningeal ectasia has been associated with the dilated epidural veins. Occasionally fluid enhancement can be seen around these areas consistent with CSF leak. Collapse of the dural sac has also been observed and ascribed to the low CSF pressure. Finally, clumping of the cauda equina is thought to result from paucity of CSF within the theca and can erroneously simulate arachnoiditis.

Imaging Methods

Brain CT

In an emergency condition as well as in the outpatients, head CT without contrast administration can be a useful initial imaging method (Lin et al. 2017; Limaye et al. 2016). This technique is capable of showing typical CSF leakage features, such as subdural fluid collections or hematomas (Fig. 12), effacement of subarachnoid spaces, and ventricular collapse. Moreover, performing brain CT with very thin slices (0.6 mm) on a multirow CT scanner, it is possible to do sagittal reconstruction in order to report brain sagging or even pituitary enlargement (Fig. 13).
Fig. 12

Emergency CT obtained in a 66-year-old man with severe headache, he denied any cranial trauma. There are bilateral hematomas in frontal regions (arrows). Note that those subdural fluid collections present different density, showing higher density on the right side (a)

Fig. 13

Brain CT performed in a 66-year-old male patient, sagittal image reveals the enlargement of the pituitary gland as well as the typical sagging brain configuration

Brain MRI

Cerebral MRI has become the method of choice in the diagnosis as well as treatment monitoring in patients with intracranial hypotension (Pattichis and Slee 2016). It should be the first-line examination in patients with suspected SIH. The recommended MRI protocol is presented in Table 2. It should be underlined that in suspected cases of SIH, intravenous contrast administration is essential in order to detect the most common sign of SIH that is pachymeningeal enhancement. Furthermore, 3D heavily T2-weighted steady-state images (DRIVE/CISS/FIESTA) should be incorporated in the MR protocol as this sequence can be helpful in demonstrating fluid collection, especially in the infratentorial region (Fig. 14). As already mentioned above, FLAIR as well as PD images are very useful sequences in delineating even thin subdural fluid collections due to higher signal intensity compared to hypointense CSF.
Table 2

Recommended brain and spinal MRI protocol for diagnosis of SIH

Brain MRI

Axial T1 weighted

Axial and coronal PD and T2 weighted

Sagittal T2 weighted

Axial FLAIR images

3D heavily T2-weighted steady-state sequence


3D postcontrast T1-weighted images

Spine MRI

Sagittal T1 weighted

Sagittal T2 weighted

Sagittal T2-weighted fat-suppressed

3D heavily T2-weighted steady-state sequence

Sagittal and axial postcontrast T1-weighted images

Fig. 14

Brain axial heavily T2-weighted steady-state image (3D-FIESTA) obtained in a 46-year-old woman with SIH clearly shows thin fluid collections within pontocerebellar angles (arrows). Note that the collections involve also both internal auditory canals, which could presumably explain the bilateral worsening of hearing experienced by the patients admitted to the Department of Otolaryngology in order to find the underlying disease

It is noteworthy that about 20–30% of patients with clinically confirmed diagnosis of intracranial hypotension syndrome may present a normal brain MR examination without any typical findings of SIH (Lin et al. 2017).

Spine MRI

It has been recommended that MRI of the whole spine without and with intravenous contrast administration should be performed in patients who have failed to respond to conservative therapy (Holbrook and Saindane 2017). MR examination is able to identify the characteristic features of intracranial hypotension and assess the presence of coexisting anatomical abnormalities, and also sometimes it is even possible to find the site of CSF leak. Spine MRI may also facilitate the further planning of patients’ management, especially surgical procedures. The recommended MR protocol for spine imaging is shown in Table 2. Similarly to brain imaging, it is noteworthy that axial heavily T2-weighted steady-state images (DRIVE/CISS/FIESTA) should be included in the MR protocol (Pattichis and Slee 2016) due to the special ability of this sequence to delineate even thin fluid collection (Fig. 15).
Fig. 15

Spine axial heavily T2-weighted steady-state image (3D-FIESTA) obtained in a patient with SIH accurately shows thin epidural fluid collection in the anterior part of the cervical canal (arrow)


CT myelography with iodinated contrast administration is the diagnostic method of choice for evaluating the location as well as extent of the CSF leak (Fig. 16). However, myelography is best avoided unless absolutely necessary as this requires a lumbar puncture and hence a further dural “tear” with potential for exacerbating symptoms.
Fig. 16

CT myelography performed in a 37-year-old male patient with recurrent orthostatic headache and strong clinical suspicion of SIH. Imaging confirms a left-sided T11/T12 meningeal diverticulum – this was an isolated spinal abnormality and was therefore assumed to represent the site of CSF leak. Targeted blood patch was undertaken with good outcome

Contrast medium leaking into the peridural space can be an indicator of the site of CSF leakage. The leaking CSF will accumulate in the peridural space and cases a CSF-like signal on MR in this space. The spinal hygroma is usually walled off by the peridural membrane and may represent a tubular pseudocyst on CT myelography.

Conventional myelography (digital subtraction myelography) may be performed in cases with suspicion of rapid leaks that are difficult to assess on CT myelography due to large volume of extravasated contrast obscuring the exact leak site. If a fast leak is suspected or routine CT myelography results are negative, dynamic CT myelography is recommended. However, in subjects with a presumed slow CSF leak, delayed imaging may be successful in detecting the site of leakage (Holbrook and Saindane 2017).

Magnetic resonance imaging myelography may be achieved either noninvasively or invasively.

Noninvasive MR myelography utilizes heavily T2-weighted 3D sequences (SPACE or HASTE with fat saturation) (Fig. 17).
Fig. 17

MR myelography in a 53-year-old woman with 18-month history of disabling postural headache. Sag T2-weighted image (a) demonstrates prominent extra-arachnoid fluid collections extending throughout the cervical and visualized thoracic spinal canal. MR HASTE myelography (b) reveals extensive cervicothoracic spinal diverticula

Gadolinium-enhanced MR myelography has been increasingly used recently. Moreover, it has been proved to reveal more leaks than CT myelography (Holbrook and Saindane 2017). This method has been shown to be able to detect the site of leak in approximately 20% of cases where CT myelography was negative (Pattichis and Slee 2016). MR myelography is an invasive method which requires intrathecal administration of gadolinium and thus depends on the radiologist’s experience. It should be stressed that this is an “off-license” use of gadolinium and demands local agreement with hospital medicines safety committee. It is noteworthy that gadolinium stays in the spinal subarachnoid space for 24 h after initial injection, thus enabling the delayed imaging acquisitions, which can be useful in detecting intermittent CSF leaks (Limaye et al. 2016; Holbrook and Saindane 2017). Apart from that MR myelography avoids ionizing radiation, which is obviously another advantage of this method.

Radionuclide Cisternography

Radionuclide cisternography requires intrathecal injection of radioisotope by lumbar puncture, followed by serial scans over 24–48 h. This method can identify the presence of extradural leak of CSF (Fig. 18), especially in patients with negative myelography results (Limaye et al. 2016). Indirect features of intracranial hypotension on radionucleotide scanning include the characteristic early appearance of isotope within the bladder indicating early systemic absorption, probably related to CSF leak; however, this does not necessarily distinguish between actual CSF leak and CSF hyperabsorption. Poor passage of radionucleotide from the basal cisterns to the Sylvian fissures and interhemispheric cistern and over the convexities is seen with intracranial hypotension (Lin et al. 2017). Care must be taken not to overinterpret the findings at the site of injection as there may be an artifactual leak.
Fig. 18

Indium radioisotope cisternography (a) and CT myelography – orientated to replicate cisternogram (b) performed in the same patient as shown in Fig. 16 – a 37-year-old male patient with recurrent orthostatic headache and strong clinical suspicion of SIH. Focal radioisotope accumulation on the left at T11/12 – concordant with the CT demonstration of a left-sided T11/T12 meningeal diverticulum

Nowadays, radionuclide cisternography is infrequently performed in the everyday clinical practice because of limitations of spatial resolution as well as progress in CT and MR methods that are regarded to be techniques of choice in patients with suspicion of SIH (Kranz et al. 2017).

Color Doppler of the Superior Ophthalmic Vein

Color Doppler flow imaging of the superior ophthalmic vein has also been used with the finding that the diameter of the superior ophthalmic vein was increased in patients with intracranial hypotension (3.9 ± 0.2 mm) as opposed to patients with other causes for headache (2.6 ± 0.4 mm). There was also an increase in the maximal flow velocity on spectral Doppler (17.0 cm/s ±3.4 vs. 7.3 cm/s ±1.7).

Treatment and Prognosis

The primary goal of therapy is to stop the CSF leak and increase CSF volume. PDPH and SIH typically are self-limited conditions, with the majority of cases resolving within a week via administration of conservative/medical treatment. This resolution most commonly occurs with or without the most effective, yet more invasive, treatment: an epidural blood patch (EPB).

Although most cases are self-limiting and last for several days, in severe cases, IH can persist for several weeks to months and can be extremely debilitating with associated morbidity: in these cases EPB is the gold standard of treatment (Rettenmaier et al. 2017) (Table 3).
Table 3

Treatment of IH

Grade of IH


Mild to moderate IH


Bed rest


Abdominal binder

Mild to moderate IH (failure of conservative treatment)




Methylxanthine derivatives

Severe IH


Epidural blood patch

Severe IH (failure of repeated EPB)



Conservative treatment consists of complete bed rest with the head of the bed flat, hydration, or the application of an abdominal binder; this strategy clearly improves the symptoms; however, there is no evidence for prevention from PDPH or faster recovery after IH.

Medical therapy is based on methylxanthine derivatives, including caffeine or theophylline, and has been used and has shown some benefit in the treatment of IH. The mechanism of action may involve blocking adenosine receptors, which results in cerebral vasoconstriction, counteracting the cerebral vasodilation occurring from CSF leakage and intrathecal hypotension.

The current clinical literature shows some continued support for caffeine use, particularly in mild to moderate cases; however, most of the studies suggest that the improvement from caffeine administration is temporary, and there is no reduction in the rate of further epidural blood patch administration (Sachs and Smiley 2014).

After the failure of conservative and/or medical treatments, EBP is the modality of choice. Today EBP has become the preferred treatment for moderate to severe IH.

It is able to relieve symptoms in 90% of cases, often providing instantaneous relief of symptoms regardless of the site of the leak; furthermore, it can also be used in SIH patients without identifying the site of the CSF leak. In case of failure, it can be repeated because of its low risk of severe complications.

Originally developed in 1960 by a general surgeon, Dr. James Gormley, the procedure consists of injecting autologous, unclotted, and sterile blood (typically 10–20 ml) into the epidural space shortly after a diagnosis of IH has been determined (Shaparin et al. 2014); the injection is typically performed under computed tomography (CT) guidance, and the blood is mixed with heparin and iodinated contrast agent (10% of the blood volume injected) (Figs. 19, 20, and 21).
Fig. 19

Axial CT scan at level of D12–L1 (a) and L3–L4 (b), patient standing prone, showing iodinated contrast injection through the 22 G thin needle positioned into the epidural space with left interlaminar approach: opacification of the epidural space is appreciable (black arrow)

Fig. 20

Same patient of Fig. 19: Axial CT scan at level of D12–L1 (a) and L3–L4 (b), patient standing prone, showing iodinated contrast injection and blood patch after injection: opacification of the epidural space is appreciable (black arrow)

Fig. 21

Same patient of Figs. 19 and 20: Axial CT scan at level of L2–L3 (a) and sagittal multiplanar reconstruction of the lumbar segment (b), patient standing prone, showing iodinated contrast injection and blood patch after injection: opacification of the epidural space is appreciable (black arrows) along all the lumbar segment

The most effective technique reports involve Trendelenburg positioning for 60 min to 24 h after the procedure, with or without acetazolamide premedication.

EBP can be targeted to the site of the CSF leak on imaging or can be delivered blindly into the lumbar region.

The distribution of blood into the epidural space tends to cephalad migration; blood migrates, on average, 3.5 intervertebral spaces above and 1 intervertebral space below the site of injection, after 20 ml of lumbar epidural blood is administered. This suggests that EBP should be performed below the level of puncture and not above if possible, although the interspace that appears technically easiest to access the epidural space is often used (Sachs and Smiley 2014).

Success rates increase with the volume of autologous blood used, with an 80% success rate with 10–15 ml and greater than 95% success rate with 20 ml. Success rate also increases with repeat EBP.

Targeted EBP may have a greater efficacy.

Two main theories have been postulated about the action of EBP.

The “Plug” theory proposes that blood injected into the epidural space forms a fibrinous clot that adheres to and seals the dural hole, thereby exerting a tamponade effect and preventing further leakage of CSF from the site and raising the brain back onto its normal fluid cushion. By preventing further CSF leakage and allowing new CSF to fill the subarachnoid space, CSF pressure is restored, providing relief.

This theory however fails to explain how many patients feel immediate relief after the procedure; indeed the immediate relief cannot be explained by fistula closure because CSF is produced at a rate of 0.5 ml/min, which is inadequate to quickly replenish the amount of extravasated CSF which can be as much as 200 ml/day.

On the other hand, the “Pressure Patch” theory suggests that blood injected into the epidural space elevates the pressure in the subarachnoid space by compressing the dura instantaneously. CSF in the spinal canal migrates into the cranium and immediately restores the intracranial CSF volume and pressure, thereby alleviating the headache.

Actually, relief of symptoms is likely a combination of these two proposed theories: initial headache alleviation is accomplished through restoration of intracranial pressure, and CSF regeneration contributes to the sustained relief, all the while the reparative processes of the body repair the dural defect (Sachs and Smiley 2014).

The complications associated with EBP are related with risk of lower limb paresthesia, epidural infection, and backache (secondary to nerve root and perhaps muscular irritation from the blood), which can last up to 5 days. In addition, if blood is accidentally injected intrathecally, instead of epidurally, it can cause arachnoiditis, meningitis, cauda equina syndrome, and permanent nerve damage.

Although EBP is largely effective, up to 30% of patients will require a second EBP due to return of symptoms, especially if large-bore needles have been the cause of PDPH. This is likely a result of dislodgement of the clot or failure of clot formation at the defect. If two blood patches have been completed and the patient’s headache still persists, then imaging is probably warranted in most situations to confirm that the proper diagnosis has been made.

If symptoms can be improved using EBP treatment, then even a thick SDH can be resolved spontaneously without the need for surgical evacuation.

However, large SDHs might cause uncal herniation and lead to neurological deterioration; surgical evacuation is therefore necessary for those patients having acute changes in consciousness.

Finally, patients who do not respond to an epidural blood patch may be treated with percutaneous placement of fibrin sealant only if there is a known site of CSF leak. When all the abovementioned therapies fail or, in cases with really severe symptoms, require immediate intervention, surgical management should be considered (Limaye et al. 2016; Lin et al. 2017).

In general, the majority of patients show good outcomes with spontaneous recovery of SIH. However, it has been reported that approximately 10% of patients may be affected by the recurrence of headache (Limaye et al. 2016). Moreover, despite successful initial therapy, some patients may present a different type of headaches that improve in the upright position and worsen during sleep. These symptoms may suggest the development of rebound intracranial hypertension (RIH) that is an important complication of therapy caused by an increase in CSF pressure (Kranz et al. 2017; Holbrook and Saindane 2017). Since the clinical symptoms of RIH including mainly headache may be misinterpreted as associated with persistent low CSF pressure, clinicians are responsible for the special awareness of diagnosis and appropriate management of patients with intracranial hypotension.

Checklist for Reporting

  • Assess the midbrain location on sagittal view for a rostral–caudal brain displacement (called as brainstem “sagging”).

  • Assess the subdural spaces for any subdural collections/hemorrhage.

  • Check if there are signs of venous engorgement.

  • Evaluate the sellar region for features of pituitary gland enlargement and flattening of the optic chiasm.

  • Look for dural thickening and enhancement after contrast administration.

  • Assess the spinal canal for any epidural fluid collections, enlarged epidural veins, dural enhancement, or extra-thecal CSF collections.

Positive Sample Report

Clinical Indication

A 56-year-old man with headache since a few weeks. He denied any cranial trauma. The neurological evaluation did not reveal any significant abnormalities.

Imaging Technique

Brain 1.5T MRI scan: axial T1, axial FLAIR, axial, sagittal T2 images, DWI, 3D heavily T2-weighted steady-state sequence, and 3D postcontrast T1-weighted images


There are bilateral subdural fluid collections in frontoparietal regions, mainly on the cerebral convexities (width up to 13 mm in the left parietal area). Subdural fluid collections are also visible in other locations, including parafalcine (width up to 4 mm) and beneath the tentorium cerebelli (width up to 6 mm) (see Figs. 3 and 4). The collections present slightly increased signal intensity on T1-weighted and FLAIR images that could suggest previous hemorrhage.

On sagittal view, the rostral–caudal brain displacement (so-called sagging brain configuration) can be appreciated with associated flattening of the optic chiasm as well as decreased mamillopontine distance and the pontomesencephalic angle.

On postcontrast T1-weighted image, there is diffuse pachymeningeal enhancement.


The MRI findings are consistent with the clinical symptoms indicating the diagnosis of intracranial hypotension syndrome.

Negative Sample Report

Clinical Indication

A 49-year-old man presented with suspicion of CSF hypotension. He reported headache for a month, unilateral facial numbness, and hearing loss since a few weeks.

Imaging Technique

Brain 1.5T MRI scan: axial T1, axial FLAIR, axial, sagittal T2 images, DWI, 3D heavily T2-weighted steady-state sequence, and 3D postcontrast T1-weighted images


The T2-weighted and FLAIR images reveal single hyperintensities in deep white matter of both frontal lobes.

The location of the midbrain is normal. There are no features of subdural collections or venous engorgement. The pituitary gland presents normal size.

On postcontrast T1-weighted image, there are no sign of pachymeningeal enhancement.


There is no sign of CSF hypotension or other explanation for the clinical findings.


  1. Forghani R, Farb RI. Diagnosis and temporal evolution of signs of intracranial hypotension on MRI of the brain. Neuroradiology. 2008;50(12):1025–34. Scholar
  2. Holbrook J, Saindane AM. Imaging of intracranial pressure disorders. Neurosurgery. 2017;80:341–54. Scholar
  3. Kranz PG, Malinzak MD, Amrhein TJ, Gray L. Update on the diagnosis and treatment of spontaneous intracranial hypotension. Curr Pain Headache Rep. 2017;21(8):37. Scholar
  4. Limaye K, Samant R, Lee RW. Spontaneous intracranial hypotension: diagnosis to management. Acta Neurol Belg. 2016;116(2):119–25. Scholar
  5. Lin JP, Zhang SD, He FF, et al. The status of diagnosis and treatment to intracranial hypotension, including SIH. J Headache Pain. 2017;18(1):4. Scholar
  6. Medina JH, Abrams K, Falcone S, Bhatia RG. Spinal imaging findings in spontaneous intracranial hypotension. AJR Am J Roentgenol. 2010;195(2):459–64. Scholar
  7. Pattichis AA, Slee M. CSF hypotension: a review of its manifestations, investigation and management. J Clin Neurosci. 2016;34:39–43. Scholar
  8. Rettenmaier LA, Park BJ, Holland MT, et al. Value of targeted epidural blood patch and management of subdural hematoma in spontaneous intracranial hypotension: case report and review of the literature. World Neurosurg. 2017;97:27–38.CrossRefGoogle Scholar
  9. Sachs A, Smiley R. Post-dural puncture headache: the worst common complication in obstetric anesthesia. Semin Perinatol. 2014;38(6):386–94.CrossRefGoogle Scholar
  10. Schievink WI, Schwartz MS, Maya MM, et al. Lack of causal association between spontaneous intracranial hypotension and cranial cerebrospinal fluid leaks. J Neurosurg. 2012;116(4):749–54. Scholar
  11. Shah LM, McLean LA, Heilbrun ME, Salzman KL. Intracranial hypotension: improved MRI detection with diagnostic intracranial angles. AJR Am J Roentgenol. 2013;200(2):400–7. Scholar
  12. Shaparin N, Gritsenko K, Shapiro D, et al. Timing of neuraxial pain interventions following blood patch for post dural puncture headache. Pain Physician. 2014;17:119–25.PubMedGoogle Scholar
  13. Spero M, Lazibat I, Stojic M, Vavro H. Normal pressure form of the spontaneous intracranial hypotension: a case report with pituitary enlargement and asymptomatic pituitary haemorrhage. Neurol Sci. 2011;32(5):933–5. Scholar
  14. Yao LL, Hu XY. Factors affecting cerebrospinal fluid opening pressure in patients with spontaneous intracranial hypotension. J Zhejiang Univ Sci B. 2017;18(7):577–85. Scholar
  15. Yousry I, Forderreuther S, Moriggl B, et al. Cervical MR imaging in postural headache: MR signs and pathophysiological implications. AJNR Am J Neuroradiol. 2001;22:1239–50.PubMedGoogle Scholar

Suggested Reading

  1. Balani A, Sarjare SS, Dey AK, Kumar AD, Marda SS. Spontaneous intracranial hypotension. J Clin Diagn Res. 2017;11(8):TJ02. Scholar
  2. Bonetto N, Manara R, Citton V, Cagnin A. Spinal subtraction MRI for diagnosis of epidural leakage in SIH. Neurology. 2011;77(21):1873–6. Scholar
  3. Kranz PG, Luetmer PH, Diehn FE, Amrhein TJ, Tanpitukpongse TP, Gray L. Myelographic techniques for the detection of spinal CSF leaks in spontaneous intracranial hypotension. AJR Am J Roentgenol. 2016;206(1):8–19. Scholar
  4. Michali-Stolarska M, Bladowska J, Stolarski M, Sąsiadek MJ. Diagnostic imaging and clinical features of intracranial hypotension – review of literature. Pol J Radiol. 2017;82:842–9. Scholar
  5. Mokri B. Spontaneous low pressure, low CSF volume headaches: spontaneous CSF leaks. Headache. 2013;53(7):1034–53. Scholar
  6. de Noronha RJ, Sharrack B, Hadjivassiliou M, Romanowski CAJ. Subdural haematoma: a potentially serious consequence of spontaneous intracranial hypotension. J Neurol Neurosurg Psychiatry. 2003;74(6):752–5.CrossRefGoogle Scholar
  7. Schievink WI. Spontaneous spinal cerebrospinal fluid leaks. Cephalalgia. 2008;28(12):1345–56. Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Joanna Bladowska
    • 1
    Email author
  • Daniel J. Warren
    • 2
  • Mario Muto
    • 3
  • Charles Anthony Józef Romanowski
    • 4
  1. 1.Department of General Radiology, Interventional Radiology and NeuroradiologyWroclaw Medical UniversityWroclawPoland
  2. 2.Department of NeuroradiologyLeeds General InfirmaryLeedsUK
  3. 3.Neuroradiology DepartmentCardarelli HospitalNaplesItaly
  4. 4.Department of NeuroradiologySheffield Teaching Hospitals NHS Foundation TrustSheffieldUK

Section editors and affiliations

  • Charles Anthony Józef Romanowski
    • 1
  1. 1.Department of NeuroradiologySheffield Teaching Hospitals NHS Foundation TrustSheffieldUK

Personalised recommendations