Clinical Neurology and Neurosurgery 140 (2016) 4346

Contents lists available at ScienceDirect

Clinical Neurology and Neurosurgery

journal homepage: www.elsevier.com/locate/clineuro

Subarachnoideal blood spread following epidural blood patch given to

treat spontaneous intracranial hypotension: Can it cause neurological
complications?
E. Ferrante a, , F. Rubino b , M. Mongelli c , I. Arpino b
a

Headache Centre, Department of Neurological Science, Azienda Ospedaliera Niguarda Ca Granda, Milano, Italy
1st Unit of Anaesthesia and Intensive Care, Azienda Ospedaliera Niguarda Ca Granda, Milano, Italy
c
School of Specialization in Anaesthesiology and Intensive Care, University of MilanBicocca, Milan, Italy
b

a r t i c l e

i n f o

Article history:
Received 14 July 2014
Received in revised form 3 November 2015
Accepted 12 November 2015
Available online 26 November 2015
Keywords:
Headache
Subarachnoideal blood
Spontaneous intracranial hypotension
Epidural blood patch

a b s t r a c t
Objective: (1) To determine the frequency of subarachnoid blood spread following epidural blood patch
(EBP) in a cohort of subjects with spontaneous intracranial hypotension (SIH). (2) To describe the outcome
of these patients.
Patients and methods: In a cohort of 106 patients exhibiting SIH, spiral spinal CT scans were obtained
post-lumbar EBP and neuroradiological data was reviewed for evidence of subarachnoideal bleeding.
Results: Subarachnoideal blood spread was detected on spinal CT scans following EBP in 9 of 106 patients
with SIH. All patients exhibited a complete recovery and no neurological complications were observed.
Conclusions: A low incidence of subarachnoideal blood spread was observed following EBP given to treat
SIH. Instances of subarachnoideal blood spread were not associated with neurological complications or
altered efcacy of the EBP procedure.
2015 Elsevier B.V. All rights reserved.

1. Introduction
Spontaneous intracranial hypotension (SIH) is a syndrome
resultant from cerebrospinal uid (CSF) leakage and a subsequent
loss of CSF pressure. SIH produces a range of symptoms, but most
notably manifests with orthostatic headache (other symptoms
listed in Table 1) [1]. Lumbar autologous epidural blood patch (EBP)
is a procedure that has been primarily used for the treatment of
post-lumbar puncture headache [6], but in recent years EBP has
become a reliable treatment (and often the rst line treatment) for
orthostatic headache due to SIH [3].
Complications from EBP are quite rare but can include
pneumocephalus [7], spinal subdural haematoma or iatrogenic
subarachnoid haemorrhage [8,9], with secondary communicating
hydrocephalus and arachnoiditis [10]. To date, it is not known
whether neurological complications can arise from subrarachnoidal blood spread following EBP. Here we report the frequency
of subarachnoideal blood spread in a cohort of SIH patients and

Corresponding author. Tel.: +39 0264442388/+39 0264447007;

fax: +39 0264442819/+39 0264442154.
E-mail address: enricoferrante@libero.it (E. Ferrante).
http://dx.doi.org/10.1016/j.clineuro.2015.11.007
0303-8467/ 2015 Elsevier B.V. All rights reserved.

describe the outcome of 9 patients with intrathecal blood spread

following lumbar autologous EBP.
2. Patients and methods
We evaluated a population of 106 patients over a 20-year period
between April 1992 and May 2012. Patients were referred to us for
orthostatic headache from SIH (according to the International Classication of Headache Disorder 2nd ed, 2004 criteria for headache
attributed to spontaneous or idiopathic low CSF pressure [11]) and
were treated with lumbar EBP. For 23 cases observed between the
years of 19922004 (prior to the development of the 2004 diagnostic criteria), lumbar puncture was performed to measure the CSF
opening pressure, which was low in 18 of those cases (78%) and
normal in 5 cases (22%). The later also demonstrated orthostatic
headache and diffuse pachymeningeal enhancement on brain MRI
typical of SIH.
Prior to EBP, all patients had failed to recover from SIH during
a conservative non-invasive treatment period that ranged from 9
days to 13 months (median 62 days), where treatment consisted
of one or more of the following: bed rest, over-hydration, caffeine,
non-steroidal anti-inammatory drugs, steroids, antidepressants,
and analgesic administration. All patients gave full informed consent before treatment, and patient data (case series, clinical,

44

E. Ferrante et al. / Clinical Neurology and Neurosurgery 140 (2016) 4346

Table 1
Symptoms associated to SIH other than headache [1].
Spinal pain
Nausea with or without emesis
Diplopia, horizontal and due to unilateral or bilateral 6th cranial nerve palsy
Diplopia due to 3rd or 4th cranial nerve palsy or both(much less common) [2]
Cochleovestibular manifestations
Photophobia, visual blurring
Upper limb numbness, paresthesias
Gait unsteadiness
Facial numbness, vague paresthesias, or weakness
Change in level of consciousness (from lethargy to coma) [3,4]
Personality change, memory decline, apathy, frontotemporal dementia-like
presentation
Movement disorders: choreiform, parkinsonism, torticollis, tremor
Bibrachialamyotrophy mimicking motor neuron disease [5]
Galactorrhea
Meniere-like syndrome
Upper limb radiculopathy
Trouble with bowel or bladder control

diagnostic, therapeutic and follow-up information) was catalogued

in an Excel database.
A standardized protocol for lumbar autologous EBP was
observed as follows: patients maintained a 30 Trendelenburg position from one hour prior to the EBP procedure through the duration
of the procedure and for 24 h after the procedure. The EBP was
administered under aseptic operatory conditions by two experienced anaesthesiologists. With the patient in a prone position, local
aesthetic was administered followed by placement of an 18G Tuohy
needle at the L2L3, L3L4, or L4L5 level. Loss of resistance to
air injection technique was used to localize the epidural space
and appropriate needle placement was conrmed using uoroscopic guidance: injection of 3 mL of contrast medium (Iopamidol)
into epidural space (Fig. 1). Autologous blood taken from a cannulated big calibres vein mixed with 5 mL contrast medium was
injected slowly into the L2L3, L3L4, or L4L5 epidural space until

the development of patient discomfort (low back pain, headache,

or nausea). Injection volumes ranged from 26 mL to 40 mL (high
volume EBP). At 30 min post-procedure, a spiral spinal CT scan
was performed to observe the spread of the blood patch into the
epidural space. Spiral spinal CT conrmed correct execution of the
procedure. Patients were asked to refrain from strenuous exercise
for 2 weeks post-procedure. Follow-up observations were conducted through clinical evaluation at 1 and 3 months post-EBP and
via telephone evaluation after 3 months. Mean follow-up time was
8 months (range 624 months).
3. Results
In total, 100% (106/106) patients indicated relief of orthostatic
headache at 24 h post-EBP. Complete recovery was obtained in all
patients following 1 (90%), 2 (4%) or 3 (6%) EBP procedures. The
actual dural leak was localized in 55% of patients using spinal MRI
(37 patients), MRI myelography (16 patients), CT myelography (3
patients) or radioisotope cisternography (2 patients). Spinal MRI
and MRI myelography was unsuccessful in localizing the spinal leak
in 48 patients (45%).
Neuroradiological evidence of subarachnoideal blood spread
following EBP was observed in 9 of 106 patients (8%), including 5
women and 4 men (mean age 44 years, range 35-58 years). For 1 of 9
patients with subarachnoideal blood spread, accidental dural puncture was reported during the EBP procedure and conrmed by CSF
leak from the Tuohy needle; the other 8 EBP procedures reported
no dural puncture and correct needle placement as conrmed by
epidurography (Fig. 1).
Of 9 patients with subarachnoideal blood spread, 3 patients
reported headache triggers from mild trauma (n = 1), sneezing
(n = 1), and cervical manipulation (n = 1). Other manifestations
included changes in hearing and tinnitus (n = 6) and diplopia (n = 3).
All patients appeared normal on brain CT scan and showed diffuse pachymeningeal enhancement on brain MRI. Spinal MRI (6
patients) showed CSF collection at the cervical level (n = 1) and
at the cervicothoracic junction (n = 2). Spinal taps performed in 4
patients demonstrated a low CSF opening pressure. The level of the
CSF leak was determined in 3 patients however the actual site of
the leak was undetermined.
None of the 9 patients with subarachnoideal bleeding (indicated
by the presence of blood mixed with Iopamidol in subarachnoideal
space, Fig. 2) developed neurological complications and orthostatic
headache was completely resolved following 24 h of the Trendelemburg position post-EBP.
4. Discussion

Fig. 1. Epidurography showing right needle placament (black arrow) and spread of
contrast medium in the epidural space (arrow heads).

In recent years, spontaneous intracranial hypotension (SIH) has

become recognized as a substantial contributor to headache in
patients. Intracranial hypotension typically results from a CSF leakthat, in turn, leads to a decrease in CSF volume. Older theories of
CSF over-absorption or under-production have not been substantiated. A decrease in CSF volume rather than decrease in CSF pressure
is the core pathogenetic factor (independent variable), while CSF
pressures and clinical or imaging changes are variables dependent
on CSF volume. CSF opening pressure is normal in a substantial
minority of patients. Essentially, all cases of spontaneous intracranial hypotension result from spontaneous CSF leaks, typically at the
spinal level, particularly the thoracic spine or cervicothoracic junction. Preexisting dural weakness (meningeal diverticula, dilated
nerve root sleeves, ectasia of dural sac), likely related to a heritable disorder of the connective tissue matrix (Marfan syndrome
or marfanoid features, joint hypermobility, retinal detachment
at young age, personal or family history of arterial dissections,

E. Ferrante et al. / Clinical Neurology and Neurosurgery 140 (2016) 4346

45

Fig. 2. From left to right: (a) sagittal, (b) coronal and (c) axial spinal after EBP showing blood mixed with iopamidol in the subarachnoid space (white arrow) and air in the
epidural space (white star).

aneurysms), has gained momentum as a signicant contributory

factor. In the context of pre-existing dural weakness, an otherwise
trivial trauma may be an etiologic factor in some patients. One
consequence of a decrease in CSF volume is sinking of the brain.
This leads to traction or distortion of the anchoring or supporting
pain-sensitive structures of the brain and therefore to orthostatic
headache. Dilatation of intracranial venous structures likely also
plays a role. Traction, distortion, or compression of some of the cranial nerves, the lobes of the brain, brainstem, mesencephalon, and
diencephalon are thought to be responsible for the non-headache
manifestations of SIH. Cochleovestibular manifestations (e.g., tinnitus, hearing change, dizziness) may be related to traction on the
eighth cranial nerve, but more plausible is an alteration in the pressure of the perilymphatic/endolymphatic uid of the inner ear.
In most cases, SIH cannot be localized to an actual CSF leak site
using traditional neuroradiological investigation (radioisotope cisternography, spinal MRI, MRI myelography and CT myelography).
Rather, diagnosis is conducted through clinical assessment of the
patient using a triad of factors: orthostatic headache, diffuse pachymeningeal enhancement (uninterrupted and not nodular) with no
leptomeningeal abnormalities on brain MRI, and low CSF opening pressure. However, we have observed a qualitative increase in
referrals of atypical cases with variability in one or more of these
hallmarks.
Proposed treatments for SIH range from conservative palliative
care (bed rest, over-hydration, caffeine, theophylline, steroids, and
NSAIDs) to more invasive procedures including EBP. The efcacy
conservative therapies is often unpredictable in patients: administration of caffeine or theophylline does not produce reliable relief
in SIH patients. The improvement of SIH symptoms with corticosteroids is often variable, incomplete, and not durable: few patients
report denitive improvement. In our group, we observe a protocol that proceeds to therapeutic intervention with a lumbar EBP
following the failure of conservative treatment for 7 days. Lumbar
EBP was initially used only for treatment of post-lumbar puncture
headache [6], but in the last few years EBP has become a reliable
treatment and even a rst line therapy for orthostatic headache
due to SIH. The reported success rate of lumbar EBP ranges from
30 to 100% in studies of SIH [3], >3040% in studies on accidental dural puncture in obstetric anaesthesia, and is 99% in cases of
cerebrospinal uid leak secondary to diagnostic lumbar puncture.
EBP appears to ameliorate SIH symptoms by two primary mechanisms: the immediate effect is related to volume replacement and
compression of the dural sac, which restores the original buoyancy position of the brain. EBP secondly produces sealing of the
CSF stula and prevents additional CSF loss.

A contributor to the efcacy of the EBP procedure in our hands

may involve (1) the maintenance of patients in the Trendelenburg position before, during and for a prolonged period after EBP
and (2) the use of a high injection volume. Other centres have
reported administering EBP to patients in a horizontal position and
maintaining patients in a recumbent position for only a few hours
following EBP. We believe that the use of the Trendelenburg position potentially reduces ow of spinal CSF stula and promotes
collapse of the dura mater to seal the dural leak. Accordingly, use
of the Trendelenburg position may enhance the efcacy of the EBP
procedure in our centres; this hypothesis requires further study.
The safety of EBP is well established and complications are rare
[810], however we sought to determine whether the apparent
complication of subarachnoideal blood spread could be associated
with neurological complications or a change in treatment efcacy.
Importantly, subarachnoideal haemorrhage can cause spinal arachnoiditis, characterized by stiff neck, low back pain with or without
leg weakness, paraesthesia, leg pain, often bilateral, hyporeexia,
sensory and motor radiculopathy, urinary and bowel sphincter
dysfunction [14], and sexual dysfunction. Studies have evaluated
the response of arachnoidal villi to the presence of blood in subarachnoid space, and have suggested that blood clots can obstruct
the arachnoideal villi and inhibit cerebrospinal uid reabsorption; furthermore, bilirubin-associated inammation can lead to
the irritation of cerebral membranes and cause communicating
hydrocephalus [15].
In 9 patients exhibiting subarachnoideal blood spread following
EBP in the treatment of SIH, we observed no neurological complications or alterations in treatment efcacy. In 8 of 9 patients where a
successful EBP resulted in subarachnoid blood spread, we suppose
that blood passage through the dura mater was due to augmentation of hydrostatic pressure in epidural space caused by a high
volume blood patch. It is possible that bleeding acted to promote
clotting and easier closure of the dural leak.
Our work is the rst to report on the frequency of neuroradiological evidence of subarachnoideal blood spread following lumbar
EBP for SIH. Importantly, this retrospective single-centred study
would benet from follow-up investigations involving multiple
centres and a larger cohort. Additionally, our study does not control for the application of EBP in SIH (and indeed this control is not
historically in the literature due to the fact that most studies report
clinical observations of EBP effectiveness) and does not control for
spontaneous recovery. Additional work is required to validate our
nding that the apparent complication of subarachnoideal blood
spread following EBP does not cause neurological symptoms or
reduce the efcacy of the procedure in the treatment of SIH.

46

E. Ferrante et al. / Clinical Neurology and Neurosurgery 140 (2016) 4346

Conict of interest statement

Authors declare no conict of interest.

[5]
[6]

Acknowledgements

[7]

Scientic editor Ashley Symons, Ph.D. provided professional

English-language editing for this article.

[8]
[9]

References
[10]
[1] B. Mokri, Spontaneous low pressure, low CSF volume headaches: spontaneous
CSF leaks, Headache 53 (2013) 10341053.
[2] E. Ferrante, A. Savino, A. Brioschi, R. Marazzi, M.F. Donato, M. Riva, Transient
oculomotor cranial nerves palsy in spontaneous intracranial hypotension, J.
Neurosurg. Sci. 42 (Sep. (3)) (1998) 117119, discussion 180.
[3] E. Ferrante, I. Arpino, A. Citterio, R. Wetzl, A. Savino, Epidural blood patch in
Trendelenburg position pre-medicated with acetazolamide to treat spontaneous intracrania lhypotension, Eur. J. Neurol. 17 (2010) 715719.
[4] E. Ferrante, I. Arpino, A. Citterio, A. Savino, Coma resulting from spontaneous intracranial hypotension treated with the epidural blood patch in the

[11]

[14]

[15]

Trendelenburg position pre-medicated with acetazolamide, Clin. Neurol. Neurosurg. 111 (2009) 699702.
E. Ferrante, D. Facchetti, I. Arpino, P. Tiraboschi, A. Citterio, Intracranial
hypotension mimicking cervical radiculopathy, Eur. Neurol. J. 3 (2011) 4647.
J.B. Gormley, Treatrment of post-spinal headache, Anesthesiology 21 (1960)
565566.
E. Ferrante, F. Rubino, L. Porrinis, Pneumocephalus: a rare complication of
epidural catheter placement during epidural blood patch, Headache 54 (2013)
539540.
J.A. Aldrete, T.L. Brown, Intrathecal hematoma and arachnoiditis after prophylactic blood patch through a catheter, Anesth. Analg. 84 (1997) 233234.
Z.I. Hasiloglu, S. Albayram, H. Ozer, D.C. Olgun, H. Selcuk, M.Y. Kaynar, Cranial
subarachnoid hemorrhage as anunusual complication of epidural blood patch,
Clin. Neurol. Neurosurg. 113 (2011) 689692.
N. Maach Al, O.J.M. Vogels, T.L. Bollen, P.H. Wessels, Arachnoiditis and communicating hydrocephalus as a complication of epidural blood patch, J. Neurol.
257 (2009) 672673.
Headache Classication Subcommittee of the International Headache Society,
The International Classication of Headache Disorders: 2nd edition, Cephalalgia 24 (Suppl. 1) (2004) 8152.
E.M. Massicotte, M.R. Del Bigio, Human arachnoid villi response to subarachnoid hemorrhage: possible relationship to chronic hydrocephalus, J. Neurosurg.
91 (1999) 8084.
I. Rice, Obstetric epidurals and chronic adhesive arachnoiditis, Br. J. Anaesth.
92 (2004) 109120.