A variable degree of intracranial hypertension (ICH) is a common affliction amongst patients with myalgic encephalomyelitis / chronic fatigue syndrome (Higgins 2013, 2015, 2017; Hulens 2018), vestibular dysfunction (Higgins 2015, Liu 2019), endolymphatic hydrops (Ranieri 2017), chronic headache or migraine (Digre 2002), and [pulsatile] tinnitus (Chiarella 2012). Yet, the majority of these patients remain undiagnosed and continue to suffer. The main reason for this, is that the body may quite subtly demonstrate intracranial hypertension on imaging studies, despite often obvious clinical symptoms. This article will briefly discuss some common causes of intracranial hypertension, its variants, and potential treatment strategies.
“Idiopathic” intracranial hypertension, especially, is a common but underdiagnosed problem that is postulated to mainly affect obese women in child-bearing age. Empirically, I’ve found that other patients also have ICH, but develop secondary CSF leaks (Osborn’s brain 2nd ed., p1144; Higgins 2014, 2019; Perez 2013; Alkhotani 2019; Bidot 2019; Morki 2002) and therefore do not test positive for papilledema and elevated lumbar punctures. IIH is diagnosed when there is no clear cause for the elevated CSF pressures, yet most patients with IIH are known to demonstrate venous anomalies that reduce cranial venous outflow. Pseudotumor cerebri is another term for ICH which implies that the CSF elevations are secondary to another pathology, for example venous sinus stenosis or thrombosis. As we will discuss in this article, lacking CSF indicators does not rule out intracranial hypertension, as they are unreliable due to frequent secondary leakage, and because they do not cover the important concomitant craniovascular hypertensive aspect (Larsen 2018, 2020) that comes with venous drainage impairment.
Acute versus chronic ICH
Acute variants of ICH are easily diagnosed in hospital settings, as the body does not have time to compensate. Obstructive hydrocephalus (aqueduct stenosis), tumors, subdural hematomae or meningitis are common acute or unbearably expansive pathologies that will almost certainly result in pathological elevation of cerebrospinal fluid pressures and papilledema. Because papilledema and high CSF pressures are the main diagnostic indicators for pathological CSF pressures, these patients tend to be easily diagnosed. Geeraerts (“Non-invasive assessment of intracranial pressure using ocular sonography in neurocritical care patients”; 2008) found that, in intensive care settings, ie., generally acute settings, rapid dilation of the optic nerve sheaths may be noted due to acutely elevated CSF pressures. They found that an optic nerve sheath diameter greater than 5,8 mm Ø correlated with approximately 25 cm H2O CSF pressures, and make it easier and quicker for clinicians to determine when to schedule the patient for shunting or craniectomy.
Fig. Obstructive hydrocephalus
Blood clots in the cerebral venous drainage system, also called dural sinus thrombosis, is a known potential cause of intracranial hypertension and even hydrocephalus. However, in many circumstances, clots in the venous system may not severely affect CSF pressures, but may still greatly impair cerebral blood drainage and thus increase the craniovascular pressures despite the CSF pressures being normal or borderline. A very large venous sinus thrombosis will usually cause a venous infarct, but not always. Actually, up to 50% of clots may occur without secondary venous infarcts (Skalina T, Gaillard F. Cerebral venous thrombosis. Web article. Available from: https://radiopaedia.org/articles/cerebral-venous-thrombosis; Rodallec MH, et al. Cerebral venous thrombosis and multidetector CT angiography: tips and tricks. 2006). A positive MR or CT venogram (demonstrating severe venous obstruction) without a compatible infarct will very frequently be misdiagnosed as a normal venous anomaly (venous variant) despite the patient’s compatible symptoms, and he or she will be sent home.
Mechanical stenosis (narrowing) of the venous sinuses, especially the transverse venous sinuses is yet another phenomenon causing great confusion. Literature has suggested that up to 50% of sinuses may be “idiopathically stenosed”, ie. narrowed. However, how reliable is this? A cranio-venographic study is rarely done unless significant suspicion already forelies with regards to craniovenous pathology. Most of these studies are done due to compatible symptoms, and rarely does there forelie pre-existing venographic images for comparison. In other words, if the scan comes back showing obstruction, but there is no infarct, and no compelling signs of CSF pressure increases (which are unreliable, more on this later), the imaging study will almost definitely be deemed normal. However, the only reliable way to know if the venous obstruction is a normal variant, is either 1. to have pre-existing venograms (prior to symptom onset) that shows similar appearance, or 2. to perform a catheter venography and manometry to ensure that the intradural venous pressures are low and relatively symmetrical, and that the stenotic site can be easily examined with the catheter (ie. without resistance upon catheter entry to stenosed segment). Difficulty entering the stenosed site suggests thrombosis.
Fig. Cerebral venous sinuses
CSF versus craniovascular pressures
Almost all diagnostic measures in the detection of intracranial hypertension are based on CSF pressure markers. Unfortunate, this is very unreliable. First of all, because many if not most of chronic intracranial hypertension sufferers develop secondary CSF leaks through minor (secondary) dural defects or through defects (again, secondary to pressure increase) in the maxillary, ethmoid, frontal, sphenoid or mastoid sinuses. This makes the patient drip CSF and thus the CSF pressures will reduce to where it is borderline high or at the high end of normal ranges.
Fig. Textbook appearance of intracranial hypotension due to CSF leak.
Venous stenosis has been shown to highly associated with intracranial hypertension, as is elevated dural sinus pressures by catheter manometry (De simone, Advancement in idiopathic intracranial hypertension pathogenesis: focus on sinus venous stenosis, 2010). Although not commonly understood, chronic craniovenous drainage insufficiency will result in both elevations of CSF pressures as well as craniovascular pressures. This is why CSF shunting a patient with intracranial hypertension will not have curative effect if it is venogenic, ie. if it is originating from chronic venous insufficiency, but it may be curative if the patient suffers from obstructive hydrocephalus, for example.
If the patient has an underlying venous pathology that is not being detected, the patient may or may not develop significant indicators of elevated CSF. For example, stenosis or thrombosis of the superior sagittal sinus, which is the main drainage pathway for CSF, will almost inevitably result in papilledema and elevated lumbar punctures, as well as possible hydrocephalus. Venography will be indicated unless other causes of hydrocephalus are already seen. However, if one transverse sinus is obstructed, especially the hypoplastic one, this may not be enough of a problem to cause significant CSF drainage impairment, but will certainly reduce blood drainage in that hemisphere and therefore increase the likelihood for ipsilateral migraine, vestibular dysfunction, tinnitus, etc., due to consequent vascular congestion on that side. Venography should still be done. I reiterate; craniovenous drainage deficiency, indicated by stenosed segments identified upon MR or CT venography, will to a variable degree increase the intracranial blood pressures, regardless of whether or not the CSF pressures appear normal. The actual venous pressures can be determined by catheter manometry, if venography revealed stenosed segments. High venous pressures with compatible symptoms, and lacking markers for CSF pressure elevation, should not automatically be rendered as a coincidental finding.
Fig. left-sided transverse sinus thrombosis. A compatible white-vessel sign also seen on axial T1-weighted images.
Another virtually unknown cause of craniovascular hypertension is thoracic outlet syndrome. TOS, with or without symptoms of brachial arterial insufficiency, may induce what I have called a secondary craniovascular hyperperfusion phenomenon (TOS CVH). It is a fancy word that means that the blood that is prevented from entering the arm in TOS, rather reverts to the head through the vertebral and common carotid arteries, resulting in chronic hypersaturation and dilation of the cranial arteries. This is damaging to the brain’s vasculature and also causes autoregulation impairment. Therefore, another protective response is initiated. Just like excessive CSF pressures may narrow the intracranial arteries and cause an ischemic stroke in ICU settings, low or comparatively low CSF pressures will allow hyperdilation of the intracranial arteries in TOS CVH. Increasing the CSF pressures will prevent hyperdilation from TOS CVH, but will, over time, result in “idiopathic” intracranial hypertension (IIH). Studies from the 80s show that patients with increased systemic hypertension also developed, seemingly compensatory, increases in CSF pressures. I strongly believe that this is a protective reaction to prevent arterial hyperdilation. If a patient with significant CVH develops a secondary CSF leak, which are usually asymptomatic, they will develop POTS as the arteries are now allowed to hyperdilate and will be difficult to saturate when being upright. Degree of orthostatic incompetence depends on how impaired the cerebral autoregulation is and how hyperdilated the arteries are. Therefore, I postulate that a scalenectomy may be a better treatment for true IIH (presuming the venous system truly is normal and not merely misinterpreted as normal), than shunting.
Fig. Compression of the distal subclavian artery will increase peripheral resistance in the thoracic outlet, and force increased blood flow towards the head through the vertebral and common carotid arteries. This phenomenon is worse when lying down and better when upright. It may impair cerebro autoregulation, hyperdilate the arteries and induce secondary increased CSF pressures to protect against the hyperperfusion-induced arterial hyperdilation.
“Idiopathic” intracranial hypertension
Idiopathic means “without known cause”. That does not mean that there is no cause. IIH is often misdiagnosed due to improper interpretation of the craniovenous system. For example, if thrombosis of one lateral sinus without adherent venous infarct is deemed a “normal variant”, “normal hypoplasia”, despite elevated CSF pressures and clear signs of IIH, then the patient may be improperly scheduled for CSF shunting rather than being put on anticoagulative treatment (thrombolytic treatment), balloon venoplasty or stenting.
In incidences where the dural sinuses truly appear normal, the jugular outlet should be examined. In many circumstances, severe jugular outlet obstruction will be noted. Jugular outlet obstruction by the styloid process or C1 transverse process is a common problem. Jayaraman et al. 2012 found that up to 30% of patients undergoing neurovascular workups (MRA) demonstrated internal jugular vein stenosis. Ding et al. 2019 found that 70% of patient with cervical spondylosis had some degree of uni- or bilateral jugular vein stenosis. Ahn et al. 2014, interestingly, found that up to 70% of patients with thoracic outlet syndrome also had internal jugular vein stenosis! In my clinical experience, there is a very high prevalence of TOS in ICH patients.
Fig. Internal jugular vein compression by the C1. The degree of compression is often better demonstrated with TOF (time of flight, non-contrast sequences) as the signal will attenuate according to actual flow reduction. In the contrasted scans, normal signal continues post-stenosis and therefore the degree of stenosis will have to be measured while signal strength should be disregarded.
If the jugular outlet appears obstructed on CT venography (A CT is preferable, as the styloid process is difficult to reliably visualize on MRI), a styloidectomy, transversectomy or jugular stenting can be done. Venous stents tend to increase risk of thrombosis (clotting) and this can be lethal in certain circumstances. Therefore, all other options should be done prior to stenting, such as balloon venoplasty and the before-mentioned.
Fig. The dominant internal jugular vein is crushed between the styloid process and C1’s transverse process, clearly demonstrated on this CT venogram. The patient did not demonstrate papilledema on fundus exams, but showed signs of AV nicking and copper wiring, which are early signs in chronic hypertensive retinopathy.
If both the dural sinuses as well as jugular outlets are indeed completely normal, then TOS CVH is the most likely cause of the patient’s IIH (as explained above). Usually along with severe anxiety or whiplash, as both of these co-morbidities cause TOS. TOS is an undiagnosed epidemic amongst patients with chronic pain and its symptomology is all over the spectrum.
Common imaging findings in ICH
Osborn’s brain states, correctly, that you’ll often only find one single element of these findings. Yet, most radiologist will not diagnose ICH unless many and obvious CSF pathology indicators are seen on imaging. This is rarely seen, and ICH is very underdiagnosed! Therefore, it is the clinician’s job to render clinical suspicion and to interpret the images with greater care. It is nearly impossible for the radiologist do to this, as they do not work with the patients and therefore cannot build proper clinical suspicion. Therefore, it is and must be the clinician’s job: He or she must both examine the patient and review the images to render the diagnosis. Preferably, in cooperation with an open-minded radiologist that understands that book-knowledge does not carry over perfectly to clinical settings.
Common diagnostic findings in ICH, suggestive of increased CSF pressures, are lateral ventricular narrowing (“slit ventricles”; suggestive of brain swelling), pituitary concavity or an empty sella, posterior orbital flattening, increased optic nerve sheath diameter => 5,8 mm but preferably greater than 7mm, cerebellar descent through the foramen magnum (often borderline, and not frank Chiari). Dilation of the ventricles generally suggests a large problem with the superior sagittal sinus, the dominant transverse sinus, or aqueductal obstruction. The increased intraventricular pressures often result in periventricular edemae (also known as transependymal edema). The transverse process of the C1 will obstruct the jugular foramen on sagittal images, preferably black-blood sequences with 3mm slice thickness. Moreover, there may or may not be a white-vessel sign in the distal sigmoid sinuses, suggestive of severe flow stasis or thrombosis.
Fig. Distended optic nerve sheaths with orbital flattening and papilledema, empty sella, and concomitant venous sinus stenosis. A “textbook” appearance of pseudotumor cerebri.
Secondary cerebrospinal leaks, common in intracranial HYPERtension
Chronic elevations in cerebrospinal fluid pressures result in CSF leaks. This is not well known, but is still stated black on white in Osborn’s brain 2nd ed (p. 1144). Chronic CSF increases causes small defects in the thin and porous bony plates of the sinuses and thus may result in minor meningoceles or leakages. Unfortunately, because nearly all ICH imaging-indicators are based on CSF pressures, a CSF leak will reverse all or most of these signs. This is why a venography is important also when the plain head MRI appears normal. Moreover, rendering the venography as a normal variant, if it does demonstrate anomalies, may be easy to do if the plain head MRI is normal, but unfortunate and premature if there are compatible symptoms. Rather, a catheter venogram and manometry should be done to measure the venous sinus pressures, presuming that the signal loss is within the dural sinus system. If the obstruction is at the skull base by the C1 or styloid process, this is never a normal anomaly and should not be interpreted as one.
Imaging signs in CSF leak involve subdural effusions which may be “halo like”; surrounding the brain, or more commonly, only involving the anterior aspects of the brain. The pituitary gland may be convex and swollen, and there may be presence of epidural vein dilation in the spinal canal. Significant sagging of the brain is usually not seen unless the leak is very severe.
In clinical practice, I’ve found that most patients suffering from CSF leaks are symptomatic not due to the leak (as the pressures are not low enough to cause real problems), but due to the underlying elevated blood pressure. The patients who become afflicted with orthostatic incompetence tend to concomitantly suffer from strong TOS CVH, usually along with anxiety or a previous bad whiplash injury. Or, they may have a large leak that needs surgical repair, but in such case, the lumbar puncture will be below reference. Higgins et al. have shown that fixing the cause of ELEVATED pressures will render the body able to automatically repair minor leaks that are seen in secondary CSF leaks due to chronic ICH (Higgins 2014, 2019).
Patients who have been diagnosed with primary leaks should be careful to examine whether or not they have underlying venous congestion. Treating the leak in such a case will not help; rather, it may make you worse. If it works, the improvement will usually be very short-lived. They have no, or poor response to blood patches. Again, I am referring to secondary CSF leak. Most CSF “leakers” that I have consulted with, have underlying severe venous congestion, TOS, and also, usually, a history of anxiety or whiplash. In other terms, their leak is secondary to longstanding high pressure. The leak is usually not primary. It is also important to be aware that numerous, repeated lumbar punctures and blood patches may result in adhesive arachnoiditis, a nasty condition that is very hard to treat. Avoid repeated blood patches unless there is no doubt that the condition is primary and does not have underlying factors of venous drainage compromise.
A GP should always exclude other causes first. This is why the patient does not see a specialist before they see a general practitioner. Compatible symptoms, either sudden (to some extent suggestive of aqueduct stenosis or dural sinus thrombosis) or insidious onset of headache, tinnitus, visual impairment without frank ocular pathology, vestibular dysfunction, headache, dizziness or presyncope when bending down, and more, are common symptoms that render suspicion for a potential intracranial hypertension and warranting further diagnostic studies.
First, one would have to identify the presence as well as the most likely cause of the eventual increased pressure. This is difficult and requires knowledge about clinical neurology as well as radiology. A plain head MRI along with a venogram is a good start. Look for narrowing or dilation of the lateral ventricles, depression or swelling of the pituitary, cerebellar tonsillar descent, dilation of the optic nerve sheaths, orbital flattening, or epidural vein dilation in the spinal canal. Some of these signs are for ICH, some are for leaks.
Signs of severe CSF elevation such as brutally distending optic nerve sheaths, papilledema or hydrocephalus warrants a lumbar puncture. However, the lumbar puncture is usually not helpful in circumstances where plain head MRI findings are borderline normal, despite obvious clinical symptoms (suggesting that these are of craniovascular origin rather than CSF).
If there are signs of leak, the most likely cause is underlying ICH, unless, as stated, the lumbar puncture truly is below reference range and this is a trauma case. Generally, large primary leaks will demonstrate a positive myelography, whereas secondary leaks, even when substantial, will not show a positive myelography. With regards to sampling the leak and confirming the fluid as CSF, false negatives are common. CSF rhinorrhea may have to be sampled several times before finally being deemed CSF.
If venous anomalies are detected on MRV or CTV, then where? If the jugular outlet demonstrates signal loss, follow up with a contrasted venous phase CTV (Run CT 45 seconds after contrast infusion). The patient should not be lying on the head wedge, but rather have the head and neck lying flat (this improves sensitivity, as jugular outlet obstruction to great extent is a postural problem). Look for obstruction of the IJV between the styloid process and C1 lateral mass. A lacking papilledema or high lumbar puncture does not mean that the patient is healthy, as the intracranial blood pressures can be very high despite normal or borderline CSF pressures. This can be seen on ultrasound doppler scans (Larsen 2020) either as increased pulsatility (early phase) or systolic dampening (late / severe phase).
If the anomaly is within dural sinuses, it can be hard to know if the lesion is a partially obstructing thrombus, a fully obstructing thrombus, or “mere” stenosis. Perform bloodwork for increased clot risk, and ask the patient whether or not they have any risk factors such as hormonal aberrancy, hormonal supplementation, dehydration at time of onset, stroke risk in family, history of malignancy, smoking, etc. A follow up with catheter venography and manometry allows the clinician to estimate the likelihood of the anomaly being normal anatomy or pathology. If it is truly a normal variant, the manometric pressures will be low (ref. range 2-6 mm Hg; Cheyuo et al. Venous Manometry as an Adjunct for Diagnosis and Multimodal Management of Intracranial Hypertension due to Meningioma Compressing Sigmoid Sinus. 2019) and there will be no significant pressure gradients (according to the literature, less then 10 mm Hg, but probably even less if the stenosis is truly a natural variant). It should be relatively easy to pull the catheter through the stenosed segment. Difficulty pulling it through suggests thrombosis, especially if the patient had acute onset with no compatible history or additional risk factors for thrombogenicity.
Fig. Patient with sudden onset of severe headache reminiscent of thunderclap headache. No compatible history. Was dehydrated and had known hormonal aberrancies. Was diagnosed with left-sided transverse sinus stenosis, but it was not possible to pull the catheter through the stenosed segment. Catheter venography and manometry showed a completely occluded left-distal TS with collateral filling, suggestive of thrombosis. Manometry showed clearly abnormal pressures. This was a sudden thrombosis of the left transverse sinus, misdiagnosed for three years.
If the venous system is normal, suspect a thoracic outlet syndrome-induced craniovascular hyperperfusion phenomenon. Again, it implies that the blood restricted from entering the brachium, reverts to the head through the vertebral and common carotid arteries, causing hypersaturation of the intracranial arterial system. This problem may cause severe headache, fatigue, dizziness, bradycardia especially when supine, tinnitus, etc. (Larsen 2020). The patient may also have pain between their shoulderblades, chest pain, brachialgia or shoulder pain. They will usually demonstrate some degree of myotomal weakness when doing upper extremity strength neurological workups. Halstead’s test can be held for 60 seconds, look for tingling or a pain in the brachial plexus-innervated areas. Roos’ test will be positive within 30 seconds, usually. Morley’s test is usually positive. Raising the arms may improve the patient’s POTS when they stand up but worsen their headache or induce syncope when lying down.
Conservative treatment for intracranial hypertension
Treatment with acetazolamide or beta-blockers may be used to reduce the CSF and blood pressures. I don’t recommend usage of diuretics, especially in patients with concurrent venous pathology as this may increase risk of thrombosis due to preexisting slow outflow. If the patient has thrombosis, early detection and treatment with thrombolytics is important, before the clot fibroses (hardens), which may happen within six weeks in some circumstances. However, one may still respond to anticoagulative treatment after six weeks. Be aware that anticoagulation, especially with concurrent ICH will increase the risk for brain bleeds.
If the atlas is obstructing the jugular outlet, this may be treated conservatively as seen in my Myalgic Encephalomyelitis article or atlas misalignment article. It is a postural and muscular dysfunction, in most circumstances, which can be ameliorated or even cured with conservative treatment, especially in mild/moderate cases. If the patient suffers from TOS CVH, this may also be treated conservatively (but carefully), especially in mild to moderate incidences. Surgery is more viable in advanced cases. Patients with TOS CVH should avoid lying flat more than necessary, and preferably sleep on a bed wedge. Preferably on their sides. For treatment strategies, read my thoracic outlet syndrome article. Patients with anxiety as a significant comorbidity should also read my muscle-bracing article, as chronic somatic tension increases both vascular and CSF pressures. Anxiety is very, very common amongst these patients and is an amplifying factor in its intensity, development and progression.
Patients with symptomatic leaks due to underlying high pressures (lumbar puncture will not be below or at the low end of the reference range) should, in absolute contrast to common belief, not be lying flat. This worsens CVH and thus, slowly but surely, worsens the hyperdilation, damages the brain, and its autoregulative mechanisms. Patients with POTS or similar, again in incidences where the lumbar puncture is NOT below or at the low end of reference, without large traumatic leaks, should lie elevated on a bed wedge. Cerebral blood flow reduces when upright, thus the CVH reduces, preventing progression of the disorder. The patient should sleep and rest on a bed wedge or in a comfortable, inclined chair. The underlying ICH problem, whatever caused it (usually CVH and anxiety, with or without concurrent venous drainage impairment), should be treated simultaneously. Gradually, the pressures will decrease and this will allow the body to repair minor leaks automatically (Higgins 2014,2019). Only very large leaks with obvious imaging findings should warrant surgical repair, usually of traumatic origins.
Surgical options for intracranial hypertension
For jugular outlet obstruction, transversectomy or styloidectomy may be beneficial (Dashti 2012, Higgins 2015, 2017). It may also be done by performing atlantoaxial traction, facet joint alignment and fixation, cf. the work of the renowned neurosurgeon Atul Goel (Goel 2015). If the pathology is intradural, stenosis, balloon venoplasty may be attempted. Stenting can also be attempted, but once again, it increases clotting risk. Scalenectomy with pectoralis minor botox injections may be done for TOS CVH.
Fig. Improvement of venous congestion as well as neurological comorbidities after jugular outlet decompression by styloidectomy, in an ME patient.