SEE RELATED ARTICLE, P. 660 .
Annals Case
A 42 year old woman presents with 2 days of positional, gradual-onset, diffuse headache associated with vomiting. Her headache improves with lying down and worsens with sitting up. Her vital signs, physical examination, and head computed tomography (CT) results are unremarkable. The diagnosis is intracranial hypotension from a cerebrospinal fluid (CSF) leak through a traumatic dural tear.
I know what you are thinking. We all remember from our medical school days and residency that a postural headache equals a brain tumor. In tumors, though, postural means the headache worsens on lying flat and improves on sitting up. Other headache features concerning for brain tumor include gradual onset, progression over time, and worsening in the morning.
For our case, the postural symptoms are the opposite: the headache worsens with sitting up and improves with lying flat! Instead of elevated intracranial pressure from a mass, spontaneous idiopathic intracranial hypotension (SIIH) results in low intracranial pressure or volume. The classic SIIH case starts with minor trauma leading to a dural tear. CSF leakage from the tear then results in low volume or low pressure within the thecal sac. The headache presumably occurs because of traction on central nervous system structures from a less buoyant or downwardly displaced brain. Another theory involves pain from engorged sensing intracranial veins.
When to Think About SIIH
SIIH is rare and there are limited data estimating incidence, which is likely part of the reason this diagnosis often is missed. In one case series, 94% of patients received an incorrect diagnosis on their initial presentation. In one patient it took physicians 13 years to figure out the diagnosis! That’s a long time to have a headache. Rarely, SIIH may result in more severe symptomatology, such as cerebral vein thrombosis, spontaneous subdural hemorrhage, and even coma.
The key feature on history is the positional nature of the headache. Making this diagnosis more difficult, the positional component may become inconsistent over time. Asking about any positional exacerbation when the headaches first started may be the only clue on history for SIIH. The headache descriptions can vary significantly. In fact, some patients present with a typical “thunderclap” (sudden and severe at onset) and others with an exertional headache. In more severe cases, the intracranial hypotension may result in cranial nerve palsies (most commonly unilateral or bilateral horizontal diplopia from abducens nerve palsy), hypoacusis, tinnitus, or vertigo. SIIH also causes intracranial vascular congestion, which, when present in the pituitary gland, can lead to hyperprolactinemia and galactorrhea. Other reported clinical manifestations include quadriparesis, parkinsonism, and personality changes. Lastly, risk factors are important and may help guide the clinician to the diagnosis. These include patients with Marfan’s and Ehler-Danlos’s syndromes, presumably because these conditions cause the dura to be more flexible, as well as autosomal-dominant polycystic kidney disease.
In 35% of patients, minor trauma can be identified but it can be exceedingly minor, such as a positional change, reaching for a toy on the ground, chiropractic neck manipulation, sneezing, ground-level falls, roller coaster rides, or sexual activity. Admittedly, some of those seem more “minor” than others.
When to Think About SIIH
SIIH is rare and there are limited data estimating incidence, which is likely part of the reason this diagnosis often is missed. In one case series, 94% of patients received an incorrect diagnosis on their initial presentation. In one patient it took physicians 13 years to figure out the diagnosis! That’s a long time to have a headache. Rarely, SIIH may result in more severe symptomatology, such as cerebral vein thrombosis, spontaneous subdural hemorrhage, and even coma.
The key feature on history is the positional nature of the headache. Making this diagnosis more difficult, the positional component may become inconsistent over time. Asking about any positional exacerbation when the headaches first started may be the only clue on history for SIIH. The headache descriptions can vary significantly. In fact, some patients present with a typical “thunderclap” (sudden and severe at onset) and others with an exertional headache. In more severe cases, the intracranial hypotension may result in cranial nerve palsies (most commonly unilateral or bilateral horizontal diplopia from abducens nerve palsy), hypoacusis, tinnitus, or vertigo. SIIH also causes intracranial vascular congestion, which, when present in the pituitary gland, can lead to hyperprolactinemia and galactorrhea. Other reported clinical manifestations include quadriparesis, parkinsonism, and personality changes. Lastly, risk factors are important and may help guide the clinician to the diagnosis. These include patients with Marfan’s and Ehler-Danlos’s syndromes, presumably because these conditions cause the dura to be more flexible, as well as autosomal-dominant polycystic kidney disease.
In 35% of patients, minor trauma can be identified but it can be exceedingly minor, such as a positional change, reaching for a toy on the ground, chiropractic neck manipulation, sneezing, ground-level falls, roller coaster rides, or sexual activity. Admittedly, some of those seem more “minor” than others.
Getting the Diagnosis
In most of the published literature, magnetic resonance imaging (MRI) of the brain with and without gadolinium is the primary diagnostic test. MRI of the spine without contrast may also be useful to identify the site of a dural tear. CT can help diagnose SIIH but has variable sensitivity.
MRI results concerning for SIIH can be remembered with the mnemonic SEEPS ( Figure 1 ). As you may have guessed, “sagging of brain” is not normal. Although the sample size was small, a 2016 article in the Journal of Emergency Medicine found CT results indicative of intracranial hypotension in 5 of 7 patients who received a diagnosis of SIIH and 11 of 11 women with intracranial hypotension from an epidural-related CSF leak. Like CT, MRI without gadolinium can diagnose SIIH but may still miss key findings. In one case series of SIIH, 93 patients received MRI with gadolinium and 6 received MRI without gadolinium. Of those 99 patients, 93% had at least 1 of 3 SIIH MRI findings, 33% had 2 of 3, and 43% had all 3. Of the 6 patients who received MRI without gadolinium, 5 showed at least 1 SIIH MRI finding. In this study, 47% of the 98 patients who had CT myelography demonstrated a high-flow leak and 8% had a low-flow leak. No leak was identified in the remaining patients (45%).
Lumbar puncture with removal of CSF sounds as though it would worsen the intracranial hypotension and perhaps even cause another dural tear. Fortunately, lumbar puncture seems to only worsen symptoms in 5% of patients. An opening pressure less than 6 cm H 2 O is consistent with the diagnosis of SIIH, but the opening pressure can also be normal. A prolonged recumbent position can result in normal opening pressures, even in cases of MRI-positive SIIH. Lumbar puncture can result in a “dry tap” if the pressure is quite low (I can hear future interns using this as an excuse already).
Although it seems as though the CSF in SIIH should be normal, engorged veins can become “leaky,” leading to xanthochromia, elevated protein levels, or even WBC pleocytosis! This may lead to confusion with spontaneous subarachnoid hemorrhage and even meningitis in some cases, and both diagnoses may need to be entertained until more definitive diagnostic data are available. The current diagnostic criteria for SIIH are summarized in Figure 2 . The helpful acronym “positional LEAK” may help you remember that SIIH is positional plus one of the following: l ow opening pressure, e pidural blood patch helps, a ctive CSF leak, and ( k ) cranial MRI showing SIIH.