Early epidural blood patch in spontaneous intracranial hypotension

Methods.

Patients were included if they had a typical SIH defined as a severe purely orthostatic headache in the absence of obvious causes of dural tear. Headache was defined as severe when it interfered with daily activities and as purely orthostatic when it occurred in <15 minutes in the upright position and disappeared in <15 minutes with recumbency.

Brain MRI was performed using a 1.5-T system with unenhanced T1- and T2-weighted imaging and gadolinium-enhanced T1-weighted imaging in the sagittal and coronal planes.

Once SIH diagnosis was established and after full informed consent was obtained, a first EBP was done, followed by a second in case of failure or relapse. After the failure of two to four EBPs, CSF leak was looked for by MRI, CT myelography, and/or radioisotope cisternography.

The same anesthetist performed all EBPs under strict aseptic conditions in an operating room. Up to 40 mL of the patients’ own blood was slowly injected in L3-L4 or L4-L5 spaces and only was stopped in case of severe lumbar pain. The patient remained supine for 2 hours and was asked to refrain from strenuous exercise for 3 weeks. Follow-up evaluation was performed at 1 month and yearly thereafter or more frequently if necessary. The duration of the follow-up period was 1 to 4 years.

Results.

Treatment and outcome.

Three patients did not receive EBP because their headache changed rapidly during evaluation. One improved spontaneously in a few days, and two others had sinus thrombosis and were treated with heparin.6 Among the 30 patients who underwent a lumbar EBP, 27 patients (90%) had immediate relief (>90% on a verbal analog scale 0 to 10), and 3 did not improve. No complication was observed. Among the 27 with immediate relief, 17 remained pain free at the end of follow-up period, 1 was lost to follow-up evaluation, and 9 had a relapse within a few days to several weeks. In these 9 patients, the second EBP was followed with immediate and sustained relief in 6 (20%); therefore, 23 (77%) patients were pain free after one or two EBPs (figure).

• Download figure
• Open in new tab
• Download powerpoint

Figure. Flow chart of 30 patients with spontaneous intracranial hypotension treated using epidural blood patch.

Three patients improved after the second EBP but relapsed. Investigations showed a leak at the T9, C7, and T8 levels. One patient was cured after surgery, and another one was cured after five EBPs, including two at the site of the leak. The third patient has a disc herniation at the T8 level for which surgery is still debated.

Discussion.

Thirty patients with SIH and severe purely orthostatic headache received early lumbar EBP, performed whenever the typical headache persisted after the end of the clinical and MRI evaluation period; 23 (77%) were cured after one (57%) or two (20%) EBPs, with a follow-up period of 1 to 4 years.

The fact that these 33 patients were recruited during a 3-year period suggests that SIH occurs more frequently than classically thought, particularly because the present series, based on a severe purely orthostatic headache, excluded unusual presentations of SIH, such as nonpositional,7 exertional,8 or even absent headache.9

Baseline characteristics of our patients are essentially similar to those reported in other large series:1,9⇓ female preponderance, mean age of ∼40 years, physical effort as triggering factor, and frequent associated nausea, neck pain, tinnitus, or hearing disturbances. The lower rate of typical MRI changes (68%) compared with other series (>80%1,3,10⇓⇓) may be because of the greater number of recent cases.

There is no consensus regarding the management of SIH. In mild forms, conservative measures are usually sufficient. In severe cases, such as ours, there is little debate about the indication of EBP2,3,5⇓⇓ when PMGE is present on MRI,5 but when MRI is normal, it is usually recommended to proceed with additional diagnostic studies.1,5⇓ However, because these studies imply a dural puncture that may worsen the patient’s condition, we choose to first perform one or two EBPs even in patients with normal MRI and to postpone additional investigations.

The overall success rate after one or two lumbar EBPs (77%) is less than the 90% observed in post-lumbar puncture headache, probably because the leaks, when present, are mostly thoracic and thus distant from the EBP level. Our 77% success rate is higher than the 56% observed in a Mayo Clinic series of 25 patients, possibly because their patients were more severe cases, had a documented CSF leak (implying a dural puncture), and received a smaller quantity of blood (10 to 20 mL) in contrast to 20 to 40 mL in our series.