26 – Pseudotumor cerebri syndrome

26 Pseudotumor cerebri syndrome

Adult Hydrocephalus, ed. Daniele Rigamonti. Published by Cambridge University Press. © Cambridge University Press 2014.

John D. Pickard

Nicholas Higgins

Pseudotumor cerebri syndrome (PTCS) and normal pressure hydrocephalus (NPH) share malabsorption of cerebrospinal fluid (CSF) as part of their respective mechanisms but their clinical pictures are very different. In general, PTCS is a disease of young people that presents as an almost pure form of raised intracranial pressure (ICP) without focal neurological disturbance and without evidence on investigation of structural abnormality. NPH is a disease of the elderly and ICP is not raised. The cortical subarachnoid space is obliterated in NPH (cortical block and ventricular reflux on isotope cisternography and air encephalography) but patent in PTCS. It is likely that the block to CSF absorption in PTCS is at the level of the arachnoid granulations.

History

The first descriptions of PTCS depended upon the development of the ophthalmoscope by Von Helmholtz in 1851 and the technique of lumbar puncture by Quincke in 1891. Although there had been descriptions of PTCS-like cases in the period 1866–1896 [1] such as “recoverable optic neuritis in young women suffering from uterine derangement” in the 1880s, it was Quincke who in 1897 defined cases of “serous meningitis” and Nonne who coined the term “pseudotumor cerebri” in 1904. The association of PTCS with middle ear disease followed in the period 1905–1936. Then came the introduction both of neuroradiology (angiography, ventriculography, and lumbar air encephalography) and of therapies such as steroids, acetazolamide, and antibiotics for otitis media and subtemporal decompression.

Nomenclature

There has been much debate about nomenclature that has led to some unthinking tautology amongst journal editors and referees. The term in vogue at present is idiopathic intracranial hypertension (“I”IH). Strictly speaking, IIH should only include cases where there is no known cause. This begs the question of how to label cases where there is a known cause – “secondary IIH” is really silly. The term IIH is too exclusive and has led to denial by some journals of the importance of venous changes: to quote one editor, “it cannot be IIH because you have found contributory factors.” Many cases labeled as idiopathic have simply not been fully investigated so that case series of IIH will almost certainly include secondary causes.

The term benign intracranial hypertension (BIH) reflects the absence of sinister intracranial pathology. However, many patients deeply resent the term – their lives are often a misery even though their vision has been preserved.

Pseudotumor cerebri is long established (Nonne 1904, 1914) and is inclusive of patients with similar symptoms and signs requiring similar treatments. It highlights the exclusion of an intracranial mass and makes no assumptions about etiology or mechanism. The addition by Ian Johnston of “syndrome” to PTC (= PTCS) is very helpful and is analogous to the distinction between Cushing’s disease and Cushing’s syndrome [1].

Dandy criteria (modified)

The diagnostic criteria for PTCS were eponymously ascribed to Walter Dandy [2] following his detailed description of the clinical features of 22 cases in 1937, and they have subsequently been modified in the light of new knowledge:

(1) Symptoms and signs of raised intracranial pressure in a conscious patient.
(2) Measured increase in CSF pressure.
(3) Absence of focal neurological signs (except sixth cranial nerve).
(4) CSF of normal composition.
(5) Normal imaging studies (including magnetic resonance venography [MRV]) apart from possible small ventricles
- an empty sella is allowed (see below).
(6) Not attributable to another cause.
(7) Benign clinical course except for effects on vision.

Clinical features

The headache of PTCS has no distinguishing characteristics [3] unlike the headache of raised ICP due to a brain tumor (early morning headache made worse by bending down, etc.). PTCS headache may mimic migraine without aura, chronic daily headache, and tension headache. To complicate matters, many patients may have additional headache secondary to overuse of analgesics. It is instructive to assess which components of the headache are helped by CSF withdrawal at lumbar puncture. Many patients whose visual problems have been resolved by treatment still have various forms of headache. A minority of patients learn that hyperventilation helps their headache and may present with carpopedal spasm. Headache often occurs without papilledema and vice versa.

Visual symptoms [4] occur in over half of PTCS patients and include blurring and obscurations of vision and lateral diplopia. Nausea and vomiting is common (20% of cases). Patients may present with tinnitus which is often pulsatile in character [5]. Pressure on the ipsilateral internal jugular vein may stop the tinnitus, probably because it raises the pressure within the transverse sinus and expands the area of reversible stenosis. CSF withdrawal may also reverse such tinnitus. Other ear symptoms may occur and may be attributable to raised pressure being transmitted to the inner ear through the cochlear aqueduct. A minority of patients may complain of root symptoms in the arms and legs, presumably due to dilatation of the root sleeves – CSF drainage may ameliorate such root symptoms.

CSF rhinorrhea may occur either spontaneously through an empty sella or via dehiscence of the tegmen tympani or superior semicircular canal. CSF leaks after trans-sphenoidal pituitary surgery or foramen magnum decompression for Chiari malformations may prove refractory to reoperation until the presence of PTCS is recognized and CSF pressure lowered.

PTCS certainly occurs in men where it is usually not associated with obesity. Men are more likely to lose vision early: “beware the thin man with PTCS.”

PTCS in childhood is uncommon [6] and may be overdiagnosed unless congenital abnormalities of the optic disk (e.g. drusen) are excluded and CSF pressure accurately measured under calm conditions. Children are often not overweight. Where obesity is not an issue, children must be protected against becoming anorexic.

Signs

Visual abnormalities are present in the majority of patients including papilledema, visual field changes, reduction in visual acuity, and diplopia. An expert neuro-ophthalmology review is mandatory not only to confirm the diagnosis but also to exclude congenital abnormalities and provide an objective baseline as the basis for subsequent assessment of progress.

A general physical and neurological examination is performed to identify any associated conditions, whose treatment might help to reverse the PTCS.

Associated conditions

There are many conditions that have been associated with PTCS. However, few studies of such associations pass the fundamental criterion of causality, namely that withdrawal of the condition results in reversal of PTCS; reintroduction causes recurrence of PTCS.

Female-specific factors: obesity, pregnancy, exogenous estrogens, polycystic ovary.
Pregnancy may precipitate or ameliorate PTCS.
Obstructive sleep apnea.
Familial – some families include members with both PTCS and hydrocephalus.
Cranial venous outflow: congenital, compression/obstruction (extracranial and intracranial).
Hematological abnormality: hypercoagulation, iron deficiency anemia, myeloma, leukemia, polycythemia, etc.
Endocrine disorders: thyroid, Cushing’s, parathyroid, acromegaly, growth hormone replacement, miscellaneous.
Infections: middle ear, chronic meningitis, etc.
Head injury including fractures over the major venous sinuses.
Systemic lupus erythematosus, Behçet’s, Pickwickian syndrome, etc.
Drugs and chemicals: vitamin A, steroids (administration and withdrawal), tetracycline, nalidixic acid, etc.
Nutritional disorders: rickets, malnutrition.
Space flight.

Investigations

Neuroradiology

Computed tomography (CT) scanning will exclude a mass lesion and hydrocephalus but will not detect the subtle features of PTCS that a magnetic resonance (MR) scan will display. These include dilatation of the optic nerve sheaths (Figure 26.1, black arrow) with flattening of the posterior part of the globe of the eye, dilatation of the cortical sulci (meningeal hydrops), empty sella, and narrowing of the transverse sinuses (Figure 26.1, white arrows). The MRV changes have often been dismissed as normal variants but a careful controlled study with headache-free control subjects demonstrated that the venous changes should not be dismissed as MR flow artifacts [7].

Figure 26.1 MR scan showing the features of PTCS.

Lumbar puncture

CSF pressure measurement should be performed by an experienced operator, preferably under X-ray control and with electronic recording of the pressure for 20 minutes or more. Children may require sedation or a general anesthetic – hyperventilation will reduce CSF pressure through lowering of arterial carbon dioxide levels. Traumatic lumbar punctures and use of a simple manometer may underestimate the true CSF pressure.

There has been much debate about what should be the threshold for raised CSF pressure, particularly in the presence of morbid obesity and obstructive sleep apnea. It is essential to appreciate that CSF pressure can fluctuate considerably. In various series, single measurements at lumbar puncture have demonstrated pressures of 10–20 mmHg in a minority of patients with all the other features of PTCS – subsequent continuous monitoring has revealed pressures of >20 mmHg. It is therefore misleading to insist on, say, a pressure of over 25 cmH₂O (18 mmHg) in adults and 28 cmH₂O (20 mmHg) in children before making the diagnosis. If there is doubt, continuous monitoring is required [8,9]. The presence of waves of pressure on electronic recordings are also valuable indicators of raised CSF pressure. There are a small number of patients (~2%) who have symptoms and signs of PTCS at CSF pressures of less than 10 mmHg. Where a Chiari malformation is present, ICP monitoring may be required to confirm the diagnosis although characteristic narrowing of the transverse sinuses on CT venography is indicative of raised pressure.

CT venography (CTV) is superior to MRV in defining any changes in the venous sinuses. Partial reversal of the narrowing of the transverse sinuses and other intracranial venous sinuses by CSF withdrawal may be documented by CTV (Figure 26.2).

Figure 26.2 CT venogram in a case of PTCS with compressible veins: the narrowing of the transverse sinuses is immediately improved by CSF withdrawal by lumbar puncture.

Angiography with retrograde venography may be indicated not only to exclude a dural fistula but also to quantify the importance of venous sinus narrowing.

CSF composition

CSF in PTCS has a normal level of protein and glucose. There is a trend to protein levels that are slightly lower than normal. There may prove to be changes in specific peptides but such refined measurements are still at the research stage [10].

Pathophysiology

The absorption of CSF is a passive phenomenon and described by Davson’s equation [11]:

$\text{[math]}$

where P_csf is the CSF pressure, P_sss is the pressure in the superior sagittal sinus, I_f is the CSF formation rate, and R_out is the CSF outflow resistance.

It has been assumed that, in adults, P_sss remains constant in the face of increasing CSF pressure. As will be shown later, this assumption may not be valid in PTCS patients.

A number of mechanisms have been proposed as the basis for PTCS. More than one mechanism may of course be operating.

(1) Excess CSF production. This is unlikely as the sole mechanism: the only cases of raised ICP due to excess CSF production are choroid plexus villous hypertrophy and choroid plexus papilloma where there are gross increases in size of the choroid plexus which is seen in PTCS. A doubling of CSF production rate by itself would only increase ICP by 3 mmHg, according to Davson’s equation.
(2) Cerebral edema. Early studies with brain biopsies were complicated by methodological artifact. Modern diffusion-weighted MR scanning [12] has confirmed that there is no evidence for cytotoxic or vasogenic cerebral edema in PTCS. This was always unlikely given the depth of the subarachnoid space in PTCS at operation and on isotope cisternography and MR.
(3) Abnormal CSF absorption. There is considerable experimental and clinical evidence for impairment of CSF absorption in PTCS but recent findings with measurement of superior sagittal sinus pressure in PTCS have complicated the interpretation of CSF infusion studies (see below) [13]. It appears that CSF absorption via the arachnoid granulations in the intracranial venous sinuses may be inadequate but that spinal arachnoid granulations and other routes of absorption may then be recruited but only at higher baseline pressures [14]. Interestingly, phase-contrast MR has demonstrated increased CSF movement through the foramen magnum in PTCS [15].
(4) Cerebral venous narrowing. The venous sinuses adjacent to infected ears were always thought to be the link to raised ICP in otitic hydrocephalus. Pierre Janny et al. demonstrated narrowing in the superior sagittal sinus of patients with PTCS using direct pressure measurements through a burr hole [16]. The advent of MR scanning with MRV misled the subject – MRV changes were often dismissed as MR flow artifacts. Very careful methodological studies using headache-free control subjects then confirmed that these venous changes on MRV in PTCS were not artifacts [9].
The introduction of retrograde cerebral venography [17,18] then demonstrated that there was often a significant pressure gradient across transverse sinus narrowings. Venous stenting of such narrowing has helped resolve the papilledema in many patients (see below [19,20]) but the issue now is the distinction between fixed stenoses and stenoses that reverse at least in part with reduction in CSF pressure at lumbar puncture. It is now clear that sagittal sinus pressure in typical PTCS (obese young females) is not independent of CSF pressure [13]. In a study of six patients in which P_sss was measured simultaneously with CSF pressure during retrograde venography, P_sss changed pari passu with P_csf (Figure 26.3).

Figure 26.3 Results from a study of six patients in which pressure in the superior sagittal sinus was measured simultaneously with CSF pressure during retrograde venography. Pressure in the superior sagittal sinus changed pari passu with CSF pressure.
One of the many phenomena surrounding PTCS that remains unexplained is why lumbar puncture provides symptomatic relief often for weeks or months despite the fact that CSF is turned over three times each day. It has been suggested that collapsible venous sinuses may introduce a positive feedback loop into the usual linkage between CSF pressure and drainage. If the narrowing of the venous sinuses is relieved by CSF drainage, the sinuses may remain fully open until a critical CSF pressure is again reached but that may take weeks or months.
(5) Steroid metabolism, inflammation, hypercoagulopathy, and obesity. The venous story alone cannot explain why CSF pressure initially increases, or the relationship to obesity. Intriguing possibilities are emerging that link obesity to altered steroid metabolism and inflammation. For example, obesity may be linked to abnormalities of 11β-hydroxysteroid dehydrogenase type 1, which is present in both the choroid plexus and endothelial cap of arachnoid granulations. Sinclair et al. [20] have suggested that there may be dual effects on the choroid plexus with an increase in CSF production coupled with impaired absorption through the arachnoid granulations. Experimentally, administration of steroids and then withdrawal increases CSF outflow resistance. Hypercoagulability and connective tissue disorders may alter the properties of the endothelial cap of the arachnoid granulations.

Management

It is essential that patients with PTCS are managed in a multidisciplinary team setting with clearly defined care pathways.

Medical interventions

If a causative agent can be identified, then clearly it should be withdrawn or treated. Effective weight reduction through dieting has been confirmed in a randomized controlled trial to improve vision and reduce the severity of headaches [21]. Unfortunately, many patients cannot maintain their diets long term.
Drugs. A randomized controlled trial (RCT) has confirmed that acetazolamide will help patients with the less severe forms of PTCS but many cannot tolerate the side effects [22]. Regular blood chemistry measurements should be made. Topiramate has the advantage of a modest carbonic anhydrase blocking effect plus appetite suppression. Again, it has to be used with care because of potential side effects and some patients just feel awful on it. Conventional diuretics are little used. Glucocorticoids are now restricted to cases where there is very rapid visual deterioration in combination with an urgent CSF drainage procedure. Octreotide [23] has been used in a few cases but great care has to be used in supervising its administration in the light of experience with its use in acromegaly – it is not simple to use.
Repeated lumbar punctures. Neurosurgeons are often referred patients who have been psychologically traumatized by multiple lumbar punctures, often poorly performed. Apart from the often increasing difficulty of performing such punctures, there is a risk of inducing arachnoiditis. The poor quality of life of patients managed by prolonged use of ineffectual drugs and multiple lumbar punctures should not be underestimated.

Surgical interventions

Subtemporal decompression was the first procedure to be used prior to the introduction of shunts. It has to be performed with care to avoid the risk of temporal lobe epilepsy. It should not be offered if a prior foramen magnum decompression has been performed for fear of precipitating brain slump in its various manifestations including further tonsillar herniation, syrinx formation, and midbrain distortion.
CSF diversion is not straightforward [24–26]. The ventricles are small or normal in size so that image-guided placement of the ventricular catheter is required. There is a tendency for ventricular catheters to intermittently block, possibly because of intermittent exacerbations of the underlying condition. Placement of the peritoneal catheter into a grossly obese abdomen is not easy and either migration of the catheter or variable absorption of CSF may occur.
- Lumboperitoneal shunts are often followed by problems with kinking and overdrainage with secondary descent of the tonsils unless a valve with antisiphon properties is used.
- Personally, I (JDP) currently use a lumbopleural shunt that incorporates an adjustable valve and siphon control device. This approach has the advantage of connecting two large spaces with the added advantage of negative pleural pressures with each breath to hopefully keep the shunt patent. A minority of patients develop a refractory pleural effusion and a careful watch has to be kept for secondary Chiari formation.
Optic nerve sheath fenestration. This technique continues to be used in expert hands [27]. Many ophthalmic surgeons have stopped using it because they have experienced complications and it cannot be expected to resolve headaches. Complications that have been described include ocular motility problems, optic nerve hematoma, optic disk infarcts, central and branch retinal artery occlusion, sudden postoperative blindness, dacryocystitis, and fibrosis [1].
Bariatric surgery has been used for morbid obesity, particularly where vision is not under immediate threat [28]. An RCT is at the planning stage in the UK.
Venous stenting was the logical development of the common finding of transverse sinus stenoses. The first case report was published in 2002 [19] and attracted considerable skepticism and foreboding about the potential risks. In practice and in expert hands, it has proven safe when used for PTCS including pulsatile tinnitus [29]. Stent migration, venous sinus perforation, in-stent thrombosis, and subdural hemorrhage have certainly been described [30]. Such complications are more likely to occur when more difficult lesions are tackled such as areas of thrombosis. The controversial concept of chronic cerebrospinal venous insufficiency in multiple sclerosis has confused the literature (Table 26.1).

Table 26.1 Current published experience of stenting for PTCS

Conclusions

The etiology of PTCS remains a mystery but new management techniques are being developed that are achieving sufficient maturity for RCTs to be viable. Multidisciplinary team working is essential. Patients need access to a carefully audited care pathway and to have access to and be part of research. The formation of specific charities devoted to “I”IH in the UK and USA is to be applauded.

Acknowledgments

JDP is supported by an NIHR Senior Investigator Award and by the Cambridge NIHR Biomedical Research Centre Neurosciences Theme.

Conflict of interest

JDP receives occasional royalties from sale of the monograph The Pseudotumor Cerebri Syndrome that was co-authored with Ian Johnston and Brian Owler (Cambridge University Press 2007).