Abbreviations
CISS:
constructive interference in steady state
cMRI:
cranial magnetic resonance imaging
CSF:
cerebrospinal fluid
CT:
computed tomography
CTM:
CT myelography
DSM:
digital subtraction myelography
EBP:
epidural blood patch
FLAIR:
fluid attenuated inversion recovery
fs:
fat saturation
HASTE:
half acquisition single-shot turbo spin echo
MPRage:
magnetization prepared rapid gradient echo
MRI:
magnetic resonance imaging
POTS:
postural orthostatic tachycardia syndrome
SIH:
spontaneous intracranial hypotension
SLEC:
spinal longitudinal extradural CSF collection
SPACE:
sampling perfection with application optimized contrast using different flip angle
evolution
SWI:
susceptibility weighted imaging
T1w:
T1-weighted
T2w:
T2-weighted
TSE:
turbo spin echo
2D:
two-dimensional
3D:
three-dimensional
Introduction
The diagnosis and treatment of spontaneous intracranial hypotension (SIH), which was
first described in 1938 by a German neurologist, have undergone rapid development
in recent decades [1]. However, in spite of increasing awareness of the disease, it is often initially
misdiagnosed [2]
[3]. Studies show that SIH is more common than initially assumed but there is minimal
data regarding the incidence [4]. Rates of 5/100,000 patients per year have been described, which almost corresponds
to the global incidence of aneurysmal subarachnoid hemorrhage [5]
[6]
[7].
The disease is caused by a spinal CSF leak resulting from a dural tear or a direct
connection of the CSF space into a paravertebral vein (CSF-venous fistula) [8]. In contrast, a CSF leak at the base of the skull with CSF rhinorrhea or otorrhea
usually does not cause any SIH symptoms [9]. Predominantly affected are women (f:m approx. 3:2) in middle age [6]. The resulting symptoms vary greatly and, in addition to orthostatic headache as
the main symptom, can cause a broad spectrum of symptoms, with a comatose state being
the most severe presentation [10]. Studies also show that the symptoms of low CSF pressure can have a significant
effect on quality of life. Depression, anxiety, and suicidal thoughts are common in
this patient population [11]
[12]. An incorrect diagnosis can result in an incorrect therapeutic workup and even unnecessary
surgical interventions [3].
In recent years, the diagnostic and therapeutic options have developed rapidly: even
micro-pathologies can be visualized with high-resolution imaging and can be treated
with various interventional therapies. As a result of the further development of the
diagnostic workup and treatment, the disease can be detected and properly treated
increasingly earlier, thus avoiding serious complications. As a result of the growing
awareness, increased detection of the disease and thus a greater incidence can be
expected. This review provides guidelines for the care of affected patients.
Definition and pathophysiology
In the current third edition of The International Classification of Headache Disorders,
the headache that occurs in association with SIH is described as an orthostatic headache
usually accompanied by neck stiffness and hearing disturbances [13]. The diagnostic criteria for low-pressure headaches are summarized in [Table 1]. Due to the different origins and prognoses, it is necessary to differentiate between
spontaneous hypotension and other types of low-pressure headache. For example, there
are postdural puncture headaches (PDPH) as well as headaches caused by a CSF fistula
following surgery or trauma [13].
Table 1 Modified diagnostic criteria of spontaneous intracranial hypotension (SIH) based on
the International Classification of Headache Disorders 3rd edition [10]
[13]. These criteria cannot be applied for patients in whom a lumbar puncture or previous
trauma has led to the development of CSF loss.
|
Modified diagnostic criteria of spontaneous intracranial hypotension
|
|
A.
|
Any headache fulfilling criterion C
|
|
B
|
CSF opening pressure < 6cm H2O and/or evidence of CSF leakage on imaging
|
|
C
|
Headache that has developed in temporal relation to low CSF pressure or CSF leakage
or has led to its discovery
|
|
D
|
Not better accounted for by any other ICHD-3 diagnosis
|
Even if SIH was defined as a reduced cerebrospinal fluid opening pressure (< 6 cm
H2O) intially, recent studies have shown that the majority of patients have normal intrathecal
pressure [14]. The term “spontaneous intracranial hypotension” is actually misleading since it
is a condition of hypovolemia rather than true hypotension [15]
[16]. Due to this new information and the high sensitivity of MRI in the primary diagnostic
workup of SIH (see below), intrathecal pressure measurement is no longer necessary
as a diagnostic intervention [13].
According to the Monro-Kellie doctrine, a loss of volume in a compartment of a closed
system results in an increase in volume in another compartment. In the case of CSF
loss in SIH, the venous sinuses become dilated with characteristic morphological changes
on imaging [10]
[17]. It is assumed that brain sagging and the resulting traction on the meninges cause
the usually occipital headache [3].
Modified according to Schievink et al. and Farb et al., the following three main types
of CSF leak have become established in clinical practice [8]
[18]:
-
Type 1 (approx. 50% of cases) [18]: ventral dural tear usually caused by a bone spur or rarely by a soft disc herniation
[19]. ⅔ of type 1 leaks are located in the upper thoracic region [20].
-
Type 2 (approx. 20% of cases) [18]: lateral dural tear in the region of the nerve root sheath, with the exact cause
of the dural tear being unclear to date. Dural weakness and micro-traumas are being
discussed [21]. Type 2 leaks are usually located between the mid-thoracic spine and upper lumbar
spine.
-
Type 3 (approx. 25% of cases) [18]: CSF-venous fistulas, i.e., a direct connection from the CSF space to the internal
or external spinal venous plexus. Tears of spinal arachnoid granulations may be involved
in the pathogenesis [22]. Type 3 leaks have a right-sided predilection and are primarily located in the
mid and lower thoracic spine [23].
In addition to these three types, there are indications of a fourth group in the sacral
region (approx. 6% of cases), with this being almost exclusively seen in women [24]. Further information about sacral leaks and etiology is not yet available.
Clinical manifestation
The main symptom of SIH is position-dependent, orthostatic headache with improvement
of symptoms in a lying position [3]. Accordingly, patients often report an increase in headache primarily in the second
half of the day (so-called “second half of the day headache” [25]), while the symptoms are only mild or absent upon waking up in the morning [25]
[26]
[27]. It is not uncommon for patients to be able to name the exact onset of the symptoms
[3]. Typical symptoms are mental impairment (“brain fog”) as well as a feeling of water
in the ears (aural fullness) [10]
[26]
[27]
[28]. Moreover, there is significant variability in potential associated symptoms ([Table 2]).
Table 2 Frequency and variance of associated symptoms of SIH. Table adapted to a meta-analysis
by D’Antona et al. [6].
|
Symptoms of SIH
|
Frequency
|
|
Headache
-
orthostatic headache
-
nonorthostatic headache
|
94–99% [6]
87–96% [6]
4–13% [6]
|
|
Nausea, vomiting
|
46–62% [6]
|
|
Neck stiffness
|
32–53% [6]
|
|
Dizziness
|
13–42% [6]
|
|
Hearing disturbances
|
18–38% [6]
|
|
Tinnitus
|
14–26% [6]
|
|
Photophobia
|
5–16% [6]
|
|
Diplopia
|
3–10% [6]
|
|
Other visual symptoms
|
7–21% [6]
|
|
Changes in consciousness
|
8–22% [6]
|
|
Cognitive impairments
|
2–11% [6]
|
Important differential diagnoses of SIH include postural orthostatic tachycardia syndrome
(POTS), migraine, and cervicogenic headache [29]
[30]. However, differentiation from other diseases like Chiari malformation or frontotemporal
dementia can also be difficult (see below).
As a rule, the symptoms of SIH can change over the course of the disease, with the
headache or the postural dependence of the headache becoming less prominent and other
symptoms becoming more significant [3]
[31].
Step-by-step diagnosis
Indications of SIH (MRI and SIH score)
The first step in the diagnostic workup is to evaluate the probability of a loss of
CSF. Noninvasive MRI examination with suitable sequences is sufficient for this purpose
[32].
MRI of the head
A standard protocol should include at least one sagittal T1w sequence after contrast
administration, ideally a 3D sequence (e.g. MPRage, “magnetization prepared rapid
acquisition with gradient echoes”), and a non-contrast FLAIR sequence (“fluid attenuated
inversion recovery”) of the head [30].
The “Bern SIH score”, which has become established in the literature for evaluating
the presence of a CSF leak ([Fig. 1]), is calculated based on these sequences [17]
[33]. Different SIH signs are weighted differently in this score. Thus, pachymeningeal
enhancement, dilation of the venous sinus, and a reduced suprasellar distance are
very sensitive signs of SIH and are therefore each given 2 points [17]. Knowledge of the ubiquitous pachymeningeal contrast enhancement in SIH is decisive
for the differentiation from inflammatory or neoplastic conditions (usually leptomeningeal
enhancement) [29]. Minor criteria include a reduced prepontine and mamillopontine distance and the
presence of hygromas or subdural hematomas (each given 1 point) [17]. Adding the points results in a point scale between 0 and 9. A score of ≤ 2 has
a low probability of a CSF leak, a score of 3–4 a moderate probability, and a score
of ≥ 5 a high probability. The scores help with decisions about the further approach
[17].
Fig. 1 SIH score assessing the probability of the presence of SIH. Sagittal, fat-saturated
T1w sequence after contrast shows the measurement of the suprasellar distance (open
white arrow in A; ≤ 4mm gives 2 points), mamillopontine distance (asterisk in A; ≤ 6.5mm gives 1 point), and prepontine distance (diamond in A; ≤ 5mm gives 1 point). Same sequence in B shows a dilated, convexly configured transverse sinus (solid white arrows in B; yields 2 points); the normal, non-dilated sagittal sinus is triangularly concave
in configuration. The coronal fat-saturated T1w sequence after contrast shows pachymeningeal
enhancement (solid white arrows in C; yields 2 points). Below this, a hypointense rim in the sense of a hygroma (dashed
white arrows in C; results in 1 point) can be seen, which appears as a narrow, hyperintense band in
the coronal FLAIR sequence (white dashed arrows in D). 0–2 indicates low, 3–4 intermediate, and ≥5 high probability for the presence of
SIH.
MRI of the spine
The importance of an MRI examination of not only the head but also the spine must
be emphasized. As the disease progresses, the signs in the head and thus the SIH score
can decrease and epidural fluid collections along the spine can be the only sign of
SIH on MRI [18]
[34]
[35].
Heavily T2w 3D sequences of the spine, e.g., T2 SPACE (“sampling perfection with application
optimized contrast using different flip angle evolution”) with fat saturation ([Fig. 2]
C–F) are suitable to detect or rule out this epidural fluid (as a result of a dural tear)
with high sensitivity [30]
[36]. Complete visualization of the sacrum should be ensured [30]. Fat saturation on T2w sequences is essential for differentiating between epidural
fat and epidural fluid. Other T2w 3D sequences like fat-saturated T2 TSE (“turbo spin
echo”) or T2 CISS (“constructive interference in steady state”) are also possible
but can sometimes have disturbing flow artifacts [30]
[37]. At least isotropic layers should be acquired (slice thickness of ≤ 1.0 mm; duration
approx. 5 minutes per examination block) to allow axis-corrected, three-dimensional
visualization of the spinal canal [30]
[36]. Coronal 2D T2w HASTE (“half acquisition single-shot turbo spin echo”) myelograms
are acquired quickly (approx. 20 s per examination block) and are helpful for visualizing
epidural fluid or prominent meningeal diverticula ([Fig. 2]A, B) [24].
Fig. 2 Spinal MRI sequences in the SIH workup. Coronal, 2D T2w HASTE myelogram of the middle
thoracic spine down to and including the sacrum shows a conspicuous, prominent meningeal
diverticulum at the level of thoracic vertebra (Th) 10/11 on the right (which in this
case was the origin of a CSF-venous fistula, see also [Fig. 5]C, D) and example of a streaky sacral fluid collection (dashed oval white ring in B), which was later confirmed as a sacral dural leak. 0.74 mm 3D isotropic, fat-saturated
T2 SPACE sequence of the cervical/upper thoracic spine shows a “spinal longitudinal
extradural fluid collection” (SLEC) ventrally in the sagittal slice (open white arrows
in C) and in the axial reconstruction (open white arrows in D), which often indicates a ventral CSF leak. The same sequence in another case in
the region of the middle and lower thoracic spine, shows a fluid collection with ventral,
but also lateral and dorsal components (open white arrows in E and F), which often indicates a lateral cerebrospinal fluid leak. In this case, the dura
can be clearly distinguished from the epidural fluid collection as a hypointense line
(D and F).
The significance of meningeal diverticula is a frequent point of discussion, especially
because they are regularly found in healthy subjects. In addition, the available data
regarding their significance in the case of low CSF pressure is minimal. At present,
it can only be stated that CSF-venous fistulas originate from nerve root diverticulum
in approximately 80% of cases [38]
[39] (which are often prominent in our experience; see [Fig. 2]A). Type 2 leaks seem to be more frequently associated with broad-based cysts adjacent
to the dura (as a result of an arachnoid herniation through the lateral dural tear)
[40]. However, the presence of meningeal diverticula alone does not indicate SIH [41].
Sequences of the spine after intravenous contrast administration are not needed and
do not provide any additional information [30]
[36]. The intrathecal administration of gadolinium is also not superior to non-contrast,
heavily T2-weighted MRI of the spine with respect to the detection of epidural fluid.
The same applies to the conventional CT myelography with intrathecal iodine-containing
contrast agent administered with the patient in a supine position [32]
[36]
[42].
Identification of the leak
When targeted treatment is intended, the second step in the diagnostic workup is the
exact identification and localization of the CSF leak.
A special myelography or angiography suite (ideally with a tilting table) is needed
for digital subtraction myelography (DSM). Any conventional CT unit can generally
be used for CT myelography (CTM), with certain equipment being helpful for good patient
positioning. A customized tiltable table placed on the CT patient table is used for
this purpose in our hospital in order to achieve a head-down position (which can also
be achieved by placing pillows under the stomach or pelvis). By acquiring multiple
consecutive scans in the suspicious section of the spine, a dynamic examination can
be achieved (necessary in the case of type 1 or type 2).
Examination in prone position
Correct positioning is decisive for sufficient diagnostic workup: If there is suspicion
of a ventral leak (type 1) based on the MRI images, patients should be positioned
in a prone position with the head sufficiently low (approx. 10–20°).
This significant tilting in the prone position is necessary for the contrast agent
to be able to overcome the natural kyphosis of the spine and distribute uniformly
to the base of the skull [20]. In our experience, ventral leaks usually have a high flow rate so that a dynamic
DSM with very high temporal resolution is essential. Since the contrast agent leakage
can often be seen within a couple of seconds, the site of the CSF leak can be easily
missed in the case of later visualization. Alternatively, dynamic CT myelography can
be performed.
Examination in lateral decubitus position
In the case of suspicion of a lateral leak (type 2) or a CSF-venous fistula (type
3), examination in a lateral decubitus position (DSM or CTM) is necessary. A slight
head-down position (approx. 6–7°) is usually sufficient to distribute the contrast
agent along the spine and particularly along the diverticula. Analogous to ventral
leaks, a dynamic examination is needed (DSM or CTM) for lateral leaks.
Visualization of CSF-venous fistulas (type 3) is often challenging. In general, both
modalities are suitable. However, CTM was shown to be more sensitive than DSM for
this purpose in a current study [43]. A dynamic examination is not necessary. The scan should be performed immediately
after intrathecal contrast administration.
Examples of the visualization of a type 1, 2, and 3 CSF leak by DSM, CTM, and cone
beam CT are shown in [Fig. 3], [Fig. 4], [Fig. 5].
Fig. 3 Ventral leak (type 1), visualized with DSM and CTM. Digital subtraction myelography
(DSM; A) in prone position and unsubtracted images of the same series (B) showing a ventral CSF leak at the level of Th 3/4 (white open arrow in A and B), with the contrast agent spreading cranially in the epidural space. This was caused
by two paramedian bone spurs (solid white arrows in C), which can be seen in the subsequent CT myelography (CTM) in the supine position.
Dynamic CTM in the prone position (several CT scans): In this case, the first native
CT scan shows a bone spur Th 1/2 (solid white arrow in D). In the second CT scan, a faint contrast egress can be seen leaking into the ventral
epidural space at the level of Th 1/2, which becomes clearer on the third CT scan
(large solid white arrow in E and F). Cranial distribution of the contrast in the epidural space (small solid white arrows
in E, F).
Fig. 4 Lateral leak (type 2), visualized with DSM and CTM. Digital subtraction myelography
(DSM) in the right decubitus position showing a contrast egress at the level of Th
10/11 on the right (open white arrow in A) and linear spread of the contrast agent in the epidural space cranially (small solid
white arrows in A). Dynamic CT myelography (CTM) in the right decubitus position to visualize a lateral
CSF leak in another example: In the first CT scan at the level of Th 12/L1 on the
right, an epidural contrast egress is seen coronal (open white arrow in B), which spreads cranially in the epidural space (solid white arrow in B and in the axial slice in C). On the immediately following second CT scan, the epidural contrast agent leakage
increases (solid white arrows in D and E).
Fig. 5 CSF-venous fistula (type 3), visualized with DSM, CTM and ultra-high resolution cone-beam
CT. Digital subtraction myelography (DSM) in the right decubitus position shows a
small meningeal diverticulum (asterisk in A), from which paravertebral venous vessels contrast (solid black arrows in A) in the sense of a CSF-venous fistula Th 12/L 1 on the right. Another example of
a CSF-venous fistula, shown in CT myelography (CTM), which was performed about 40s
after intrathecal injection of contrast agent in the right decubitus position (no
dynamic CTM) and shows a contrast drain into the paravertebral vein Th 10/11 on the
right (open white arrows in B). Ultra-high-resolution and highly zoomed image of another CSF-venous fistula Th
10/11 on the right, which originates from a prominent meningeal diverticulum (see
also [Fig. 2]A), using cone-beam CT at a resolution of 0.14 mm (open white arrows in axial slice
in C and in coronal slice in D). Within the meningeal diverticulum, a fine jet of contrast agent (black solid arrows
in D) is visible between the nerve root sheath and the fundus of the cyst (black asterisk
in D), which fills the cyst with contrast medium (not part of the actual CSF-venous fistula).
DSM and CTM have different advantages and disadvantages ([Table 3]) that should be adjusted to the different types of leaks depending on the suspicion
on the basis of the MRI finding and depending on the experience and options on-site
(available examination equipment). While high doses can occur in some cases particularly
during dynamic CT (multiple consecutive scans) [20], DSM has a general advantage due to the relatively lower radiation exposure [44]. Recent developments at the fluoroscopy suite include conebeam CT (rotation of one
tube) to check questionable findings or to acquire ultra-high-resolution 3D images
([Fig. 5]
C, D) [21]
[45]
[46].
Table 3 Characteristics of the different investigation techniques in SIH workup.
|
Modality
|
Advantages
|
Disadvantages
|
|
DSM
|
-
High temporal resolution (dynamic investigation)
-
Especially suitable for high-flow leaks (type 1 and 2)
-
Relatively lower radiation exposure than CTM
-
Additional option to use cone-beam CT
|
-
Limited area can be visualized (usually 49cm detector field)
-
Complex investigation
-
Susceptible to artifacts (breathing, movement)
-
Projection imaging (superposition phenomena)
-
Small lesions can be missed
-
Major learning curve for the examiner is to be expected
|
|
CTM
|
-
Temporal resolution/dynamic investigation possible
-
High local resolution due to 3D imaging (thus also displaying small findings, e.g.
in type 3 leaks)
-
CT is widely used
|
-
Relatively higher radiation exposure
-
Lower temporal resolution than DSM
-
Major learning curve for the examiner is to be expected
|
The flowchart in [Fig. 6] shows the diagnostic workflow. A certain term that effectively describes the workflow
has become established in scientific and clinical discourse. Patients with epidural
fluid on MRI are characterized as SLEC + (spinal longitudinal extradural CSF collection),
with this fluid indicating a dural tear (type 1 and 2 or sacral), while a high Bern
score (Head +) with SLEC neg. spine indicates a type 3 CSF-venous fistula [18]
[26].
Fig. 6 Two-stage diagnostic concept in the SIH workup.
As a result of pitfalls regarding correct positioning and the timing of contrast administration,
both DSM and CTM are technically challenging and require a specific routine to achieve
a high-quality result. Otherwise, there is a risk of subjecting patients to repeated
examinations and unnecessary radiation exposure.
Future possibilities
Technical advances like the development of photon counting CT and the use of hybrid
systems (angiography suite + CT) will potentially be important in the future: The
high spatial resolution of photon counting CT with a relatively low radiation dose
and spectral analysis can help to improve diagnosis and to reduce radiation exposure
[47]
[48]. Hybrid devices have the advantage of the ability to directly couple the two modalities
DSM and CTM on-site [49].
Therapeutic approach
Conservative measures and blood patch
In the early phase of SIH, conservative treatment methods like bed rest, hydration,
and caffeine (even without preceding localization of the leak) can generally achieve
a temporary reduction of symptoms [4]
[35]. However, there is only minimal evidence for the use of these measures for lasting
treatment. After two weeks without significant improvement at the latest, a non-targeted
epidural blood patch (EBP) should be considered [30]. The patch works by increasing epidural resistance and thus providing temporary
compensation of the loss of CSF (immediate effect). Potential closure of the defect
as a result of granulation processes is being discussed [50]
[51]. It is unclear if the EBP has a lasting effect due to the heterogeneous nature of
the available data [52]. Smaller studies were able to show a closure rate of approx. ⅓ of patients [53]
[54]. A volume of approx. 20 ml seems to have a good effect [55].
10–14 days after EBP, follow-up should be performed to evaluate the subjective clinical
status [30]. If symptoms persist, a second EBP can be offered, or further diagnostic workup
(step 2, see above) can be initiated in order to provide definitive treatment [30]. This approach is important particularly with respect to avoiding chronic manifestations
and long-term damage [31]
[56].
Even in the case of severe symptoms and complicated courses (e.g., evidence of subdural
hematoma on imaging), prompt introduction of step two for leak identification is recommended
[30]. In this case, early referral to a CSF center can be helpful [30].
Definitive treatment methods
Definitive treatment (surgical or interventional) requires precise localization of
the CSF leak on myelography.
In the case of a targeted blood or fibrin patch, a puncture needle is placed as close
to the exact location of the CSF leak as possible under CT guidance in order to inject
blood or fibrin (approx. 2–4 ml) [57]
[58]
[59]. This method also has good results in patients who have not benefited from non-targeted
EBP [60]. The effect has been shown in larger studies, particularly for type 1 and 3 [58]
[59].
Alternatively, surgical closure is available as an option for all described leak types.
A microsurgical approach, which has a high success rate and very low rate of complications
(neurological complications < 2%) if performed at a center with significant expertise,
is suitable for this purpose [61]. The success rate after surgical leak closure is very high (over 90%). However,
patients with a CSF leak that has been present > 3 months can expect full recovery
to take approx. 3–6 months in many cases [28]. Moreover, in the case of successful surgical closure, residual symptoms can persist
in approx. one fourth of patients, which highlights the chronic nature of the disease
[28].
The endovascular option, which was first described in 2021, is available in addition
to surgery for the treatment of CSF-venous fistulas. The fistula can be permanently
closed by means of transvenous embolization of the draining, paraspinal vein using
a liquid embolic agent (onyx) [62]. Given the high success rate, this method also provides a safe and minimally invasive
alternative [63]. An example of MRI scans before and after transvenous embolization is shown in [Fig. 7].
Fig. 7 Treatment response after CSF-venous fistula embolization. Preinterventional sagittal
fat-saturated T1w sequence after contrast administration in an 88-year-old female
patient with two synchronous CSF-venous fistulas shows pronounced pachymeningeal enhancement
(open white arrows in A). After transvenous embolization of the CSF-venous fistulas at the level of Th 2/3
on the right and Th 10/11 on the left (coronal and axial hyperdense Onyx-cast in native
cone-beam CT in C and D), the pachymeningeal enhancement is no longer detectable 4 months after treatment
(B).
Long-term consequences and complications of diagnostic workup and symptoms
Complications are an important aspect of SIH that should not be underestimated ([Fig. 8]) [10]
[64]. Associated subdural hematomas ([Fig. 8]A) occur relatively frequently. Particularly in younger patients with chronic subdural
hematoma, CSF leaks seem to make up a relevant percentage of cases [64]. A cerebellar tonsillar herniation morphologically mimicking a Chiari malformation
on imaging ([Fig. 8]B) can result in the development of a syringomyelia [3]
[65]. Sinus vein thrombosis or a clinical presentation similar to frontotemporal dementia
(brain sagging dementia) is less commonly associated with SIH. In contrast to true
frontotemporal dementia, the symptoms can completely regress after treatment provided
if the leak was found [10]
[66]. In the case of a persistent dural tear, irreversible long-term damage, like paresis
and muscle atrophy in bibrachial amyotrophy or gait ataxia and hearing loss due to
infratentorial superficial siderosis ([Fig. 8]C), can occur [67]
[68]. To visualize these hemosiderin deposits that presumably occur due to the chronic
dural violation, the cMRI examination can be supplemented by a T2* or SWI (susceptibility
weighted imaging) sequence [30]
[67].
Fig. 8 Complications of SIH. Coronal FLAIR sequence showing bilateral, space-occupying chronic
subdural hematomas (open white arrows in A) in a 58-year-old patient with already impaired consciousness; the cause was a ventral
CSF leak (type 1). Axial SWI sequence shows a streaky, infratentorial siderosis (open
white arrows in B) in a 78-year-old patient with severe chronic progressive ataxia with tinnitus; the
cause was a ventral CSF leak (type 1). Sagittal T1w sequence after contrast shows
prominent “brain sagging” with no longer visible suprasellar (open white arrow in
C) and mamillopontine distance (asterisk) as well as a tonsillar descent (“secondary
Chiari”, dashed white line). The cause of the 56-year-old female patient with symptoms
of frontotemporal dementia, as well as gait disturbances, dysphagia, and dysarthria
over many years, was a CSF-venous fistula (type 3).
A known complication after definitive treatment of the leak is called rebound hypertension,
which occurs in up to one fourth of patients [69]. This is associated with headache when in a lying position (and improvement when
standing) and can usually be effectively treated with acetazolamide (lowers CSF production)
[28]
[70].
Conclusion
Spontaneous intracranial hypotension is a severe, underdiagnosed disease with a broad
spectrum of symptoms that far exceed headaches. Knowledge of the quick (< 14 d) and
step-by-step diagnostic workup and treatment is extremely important for improving
long-term treatment success and avoiding long-term damage and complications for patients.
Based on the complex and potentially radiation-intense diagnostic workup, collaboration
with a center with appropriate expertise is helpful for ensuring targeted and successful
treatment (surgical or interventional).
