Atypical presentation should not limit diagnosis of spontaneous intracranial hypotension

The absence of orthostatic headache should not exclude a diagnosis of spontaneous intracranial hypotension, according to findings published in JAMA Neurology.

Normal neuroimaging results and normal lumbar puncture opening pressure should also not prohibit the diagnosis of spontaneous intracranial hypotension (SIH), according to the researchers.

“The term SIH defines a clinical condition characterized by debilitating postural headaches secondary to spontaneous spinal cerebrospinal fluid leak and/or CSF hypotension,” Linda D’Antona, MD, MBBS, and colleagues wrote. “According to the International Classification of Headache Disorders (ICHD), third edition, SIH is diagnosed when headache has developed spontaneously and in temporal relation to a CSF leak (evident on imaging) and/or CSF hypotension (lumbar puncture opening pressure <60 mm CSF).”

SIH represents “a highly misdiagnosed and underdiagnosed disorder,” D’Antona and colleagues continued. It does not occur infrequently, with an annual incidence of 5 per 100,000 individuals per year, but the pathogenesis of the condition is unknown. In addition, the ICHD diagnostic criteria for SIH have evolved substantially in the last several decades and alternative diagnostic criteria have been suggested, according to the researchers.

“These factors have probably contributed to the current uncertainty on how to reliably diagnose SIH and effectively treat these patients,” D’Antona and colleagues wrote.

The current study aimed to provide “an objective summary” of available evidence on the presentation, findings from investigations and treatment outcomes of SIH. The researchers performed a literature search on SIH using three databases from the time of the database inception through April 30, 2020. They included original studies published in English that had 10 or more patients with SIH in their analysis.

Predetermined main outcomes included the pooled estimate proportions of symptoms of SIH, brain and spinal imaging findings and treatment outcomes, including conservative treatment, epidural blood patches and surgical interventions.

Selected studies (n = 144 out of 6,878 eligible articles) included an average of 53 patients with SIH (range, 10-568 patients).

The most common symptoms of SIH included orthostatic headache (92%; 95% CI, 87% to 96%), nausea (54%; 95% CI, 46% to 62%) and neck pain/stiffness (43%; 95% CI, 32% to 53%). Brain MRI represented the most sensitive investigation performed, with diffuse pachymeningeal enhancement identified in 73% of patients (95% CI, 67% to 80%), and were normal in 19% of patients (95% CI, 13% to 24%).

Spinal neuroimaging detected extradural cerebrospinal fluid in 48% to 76% of patients. Digital subtraction myelography and magnetic resonance myelography with intrathecal gadolinium demonstrated high sensitivity for determining the exact leak site. Lumbar puncture opening pressures were low (< 60 mm H₂O) in 67% of patients (95% CI, 54% to 80%), normal (60-200 mm H₂O) in 32% of patients (95% CI, 20% to 44%) and high (> 200 mm H₂O) in 3% of patients (95% CI, 0% to 6%). The highest reported opening pressure was 228 mg H₂O, according to the study results.

Conservative treatment, which most frequently involved bed rest and hydration, was tried in 881 patients over a range of 7 to 9 weeks. Studies reported success with conservative treatment — categorized as resolution of symptoms with no need for further treatment — in 28% of patients (95% CI, 18% to 37%).

Researchers used epidural blood patches most often when conservative treatment failed; the first epidural blood patch was successful (ie, resulting in clinical improvement with no further intervention needed) in 64% of patients (95% CI, 56% to 72%). Large epidural blood patches had greater success rates than small epidural blood patches (77% vs. 66%, respectively). D’Antona and colleagues found no reports of serious adverse events following epidural blood patches; minor, transient adverse events included back pain, radicular pain, tinnitus, paraesthesia, numbness, bradycardia and dizziness.

The present study highlighted “a certain variability” in the clinical presentation associated with SIH, according to the researchers.

“As expected, headache is the most common symptom,” D’Antona and colleagues wrote. “However, the orthostatic headache, once believed to be an essential characteristic of SIH, is not invariably present. In this review, 8%of patients had a nonorthostatic headache and 3% did not experience headaches. These percentages are likely to be underestimations as most authors used the ICHD-2 diagnostic criteria that include the presence of orthostatic headache as an essential criterion. Therefore, a diagnosis of SIH should not be excluded based on the absence of orthostatic headache.”

The results also demonstrated that a CSF leak was not visible via radiological methods for a “significant proportion of patients” who had a “clinical convincing history” for SIH, the researchers wrote.

“In view of the availability, safety and the sensitivity (comparable with other spinal investigations), spinal MRI with contrast should probably be preferred as first step spinal imaging to other more invasive spinal investigations involving the need for spinal punctures and/or the exposure to high doses of radiation,” the researchers wrote. “Digital subtraction myelography could instead play an important role in the identification of the exact leak site and guide targeted treatment; however, larger studies confirming the utility of this investigation are required.”

Overall, the study findings demonstrated that the diagnosis of SIH cannot be excluded in patients who do not present with “all the typical features” of this condition, according to D’Antona and colleagues.

“We propose that brain MRI and spine MRI with contrast could be performed as first-line investigations in patients with clinical suspicion of SIH. While a lumbar puncture could be offered to patients with a clinical picture suggestive of SIH but inconclusive first-line imaging, it needs to be undertaken with caution bearing in mind that the sensitivity of this investigation is relatively low (67%) and there is a risk of worsening SIH,” the researchers wrote. “Treatment with [epidural blood patches] could be attempted early, even if the exact leak location is unknown. Second-line spinal imaging (eg, [digital subtraction myelography] or MR myelography with intrathecal gadolinium) could be offered to patients who do not respond to [epidural blood patches] and require targeted treatment ([epidural blood patches] or surgical).”

Large, randomized clinical trials “will be required to define the best management for SIH,” according to D’Antona and colleagues.