Journal of Clinical Anesthesia (2015) 27, 6872

Case Report

Persistent paralysis after spinal anesthesia for

cesarean delivery
Valerie Zaphiratos MSc, MD, FRCPC (Obstetric Anesthesia fellow),
Dolores M. McKeen MD, MSc, FRCPC (Staff Anesthesiologist),
Bruce Macaulay MD, FRCPC (Staff Anesthesiologist),
Ronald B. George MD, FRCPC (Staff Anesthesiologist)
IWK Health Centre, Dalhousie University, Halifax, NS, Canada
Received 12 July 2013; revised 14 August 2014; accepted 22 August 2014

Keywords:
Cesarean delivery;
Spinal anesthesia;
Neuraxial anesthesia;
Neurologic complication;
Paralysis/paraplegia;
Adverse outcome;
Anterior spinal artery
syndrome

Abstract Anterior spinal artery syndrome has rarely been reported as a cause of permanent neurologic
complications after neuraxial anesthesia in obstetric patients. A parturient developed anterior spinal
artery syndrome after spinal anesthesia for cesarean delivery. A healthy 32-year-old parturient presented
at 412/7 weeks for primary elective caesarean delivery for breech presentation. Spinal anesthesia was
easily performed with clear cerebrospinal fluid, and block height was T4 at 5 minutes. Intraoperative
course was uneventful except for symptomatic bradycardia (37-40 beats per minute) and hypotension
(88/44 mm Hg) 4 minutes postspinal anesthesia, treated with ephedrine and atropine. Dense motor block
persisted 9 hours after spinal anesthesia, and magnetic resonance imaging of the lumbosacral region was
normal, finding no spinal cord compression or lesion. Physical examination revealed deficits consistent
with a spinal cord lesion at T6, impacting the anterior spinal cord while sparing the posterior tracts.
2014 Elsevier Inc. All rights reserved.

1. Introduction
The decrease in anesthesia-related maternal mortality over
the second half of the 20th century is greatly attributed to the
increase in neuraxial anesthesia for cesarean delivery (CD)
and remains the anesthetic of choice for elective CDs [1,2].
Fortunately, serious and permanent neurologic complications related to the use of neuraxial anesthesia in the obstetric
population are rare events [3-6]. The exact incidence of
permanent neurologic complications is difficult to determine

Corresponding author at: IWK Health Centre, Department of Women's &

Obstetric Anesthesia, 5850/5980 University Avenue, PO Box 9700, Halifax,
NS, Canada, B3K 6R8. Tel.: +1 902 470 6627; fax: +1 902 470 6626.
E-mail address: dolores.mckeen@iwk.nshealth.ca (D.M. McKeen).
http://dx.doi.org/10.1016/j.jclinane.2014.08.003
0952-8180/ 2014 Elsevier Inc. All rights reserved.

but has been quoted to be between 0.3 and 1.2 per 100,000 [4].
The etiology of injury includes direct trauma, injection of toxic
substances, injection into unintended places, epidural abscess
or hematoma, transient neurologic syndrome, and vascular
phenomenon [3-5].
The complication of anterior spinal artery syndrome after
neuraxial anesthesia in the obstetrical population is especially rare [3-5]. The exact incidence is difficult to determine,
although in a retrospective series of more than 500,000
obstetric patients in the United Kingdom, 1 patient was
found to have had anterior spinal artery syndrome after
neuraxial anesthesia; and it is uncertain whether epidural
anesthesia was the direct cause [6]. Anterior spinal artery
syndrome presents as a predominant motor deficit, with or
without loss of pain and temperature sensation, and with
intact vibration sense and proprioception [4]. Anatomically,

Persistent paralysis after spinal anesthesia for CD

a single anterior artery and 2 posterior arteries supply the spinal
cord. The anterior spinal artery begins superiorly at the fusion
of the vertebral arteries, and radicular arteries support its blood
supply distally. The largest and most important of these is the
great radicular artery of Adamkiewicz, which arises from the
aorta between T9 and L2. Multiple levels of the spinal cord do
not receive radicular branches, thus leaving watershed areas
that are susceptible to ischemic injury. Often, the etiology is
related to severe, prolonged hypotension, arteriosclerosis, or
mechanical interference with the aortic blood flow, caused by
emboli or vasospasm. Certain types of surgical procedures can
predispose or increase the risk of anterior spinal artery
syndrome such as surgery to the thoracolumbar aorta or
instrumented spine surgery [3,5,7].
A 32-year-old woman developed persistent neurologic
findings consistent with anterior spinal artery syndrome after
spinal anesthesia for a primary elective CD.

2. Case report
The patient provided written informed consent for publication of this report. A 32-year-old woman of Ethiopian origin,
gravida 4 para 2 abortus 1, presented at 412/7 weeks for primary
elective CD for breech presentation. Her current pregnancy
had an unremarkable prenatal course. Obstetrical history
revealed 2 previous children both delivered via assisted
vaginal delivery (forceps) with epidural analgesia. She had no
previous medical or anesthetic history. Her preoperative blood
pressure (BP) was 98/70 mm Hg, and hemoglobin (Hb) level
was 10.8 g/dL.
After placing standard monitoring and a 16-gauge intravenous (IV) catheter, spinal anesthesia was performed (T0) in the
sitting position with an aseptic technique at level L3-L4 with a
25-gauge Whitacre needle (BD Medical, Franklin Lakes, NJ).
Clear cerebrospinal fluid was obtained on the first attempt
without trauma or paresthesia and was followed by painless
injection of hyperbaric bupivacaine 0.75% 12 mg, fentanyl 20
g, and preservative-free morphine 150 g. The patient was
placed supine with left lateral tilt, and a phenylephrine infusion
(routinely used) was begun at 50 g/min. Block height
progressed as expected attaining a T4 level at 5 minutes after
spinal anesthesia. Four minutes after spinal anesthesia, the
patient presented symptomatic bradycardia (37-40 beats per
minute) with ventricular escape beats and hypotension (BP, 88/
44 mm Hg). The phenylephrine infusion was initially reduced
by half (25 g/min), and a bolus of ephedrine 10 mg IV and
atropine 0.6 mg was given. Surgery began 8 minutes after this
hemodynamic, event and the remainder of the CD proceeded
uneventfully. Blood loss was 750 mL, and the lowest
intraoperative BP was 78/33 mm Hg at the end of surgery,
which was promptly treated with ephedrine 10 mg IV (Fig. 1).
One hour after spinal anesthesia (T0 + 0:57), the patient
was transferred to the postanesthesia care unit (PACU) and
was stable (BP, 107/65 mm Hg; heart rate, 82 beats per
minute; oxygen saturation as measured by pulse oximetry,

69
97%). Over an hour after arrival at PACU (T0 + 2:10), the
patient had an episode of mild hypotension (BP, 90/49 mm
Hg). She had a soft abdomen, no excessive vaginal bleeding,
denied pain or nausea but still had a complete motor block. A
bolus of 500 mL of crystalloid was given, and the head of the
bed was lowered. Two more boluses of 500 mL of
crystalloid and colloid (Voluven; RxList, San Clemente,
CA) were given within the next 90 minutes to treat mild
hypotension (systolic BP, 85-90 mm Hg). Less than 3 hours
after spinal anesthesia, the patient started feeling abdominal
pain and received oral acetaminophen 1 g and naproxen 500
mg. Four hours after spinal anesthesia, the patient's
abdominal incision pain increased (verbal rating score, 7/
10) and she was given 10 mg of oxycodone. Hemoglobin level
was 8.2 g/dL 5 hours postpartum.
With no further hypotension and return of pain sensation,
PACU discharge was still not possible due to slow motor block
regression. The on-call staff anesthesiologist performed a
neurologic assessment (T0 + 6:25). She had little hip flexor
movement, was able to move her toes bilaterally, decreased
pinprick sensation to T4 but able to feel touch over her
abdomen. Initial clinical impression was slow spinal block
regression, with the plan for continued observation with hourly
neurologic examinations. The patient was reassessed 1 hour
later, and the neurologic examination remained essentially
unchanged. After telephone consult with neurosurgery, a
compressive spinal lesion, such as a hematoma, seemed
clinically unlikely due to the high sensory block, sensation of
pain, and little improvement in the serial motor assessments. A
diffusion-weighted unenhanced magnetic resonance imaging
(MRI) of the lumbosacral spine demonstrated no spinal cord
compression, and the spinal cord and conus medullaris were
normal (T0 + 9:40) (Fig. 2).
Serial neurologic examinations overnight continued to
have some evidence of slow motor block regression with
increased movement in her feet but minimal hip flexion. She
remained afebrile, and her BP was stable. The next morning,
16 hours after spinal anesthesia, neurology was consulted
and found proximal hip/knee strength 2 + to 3 per 5
bilaterally, ankle strength 4 + per 5, and decreased sensation
to soft touch and pinprick to T6 but intact vibration sense
and proprioception. Neurology suggested that her examination was suggestive of an anterior spinal cord lesion at T6
with a probable ischemic etiology due to hypotension,
vasospasm, or emboli or, less likely, an inflammatory
etiology. A second diffusion-weighted MRI (Fig. 2) was
requested to image the whole spine, which concluded no
spinal cord compression, no syrinx, and no filum lipoma;
however, ischemia or inflammation of the spinal cord could
not be ruled out. During the course of post-operative day 1,
motor and sensory function improved.
During the next week, the patient's motor and sensory
function continued to improve with the left side now more
affected than the right. Investigations included a normal
transesophageal echocardiogram, a negative sickle cell
screen, and a negative Hb electrophoresis. The patient's

70

V. Zaphiratos et al.

Fig. 1 Intraoperative blood pressure and heart rate over the course of the cesarean delivery. Certain intraoperative events are identified.
Superior and inferior solid lines: systolic and diastolic blood pressure, respectfully. Dotted line: heart rate.

urinary catheter was removed on POD 5, and it was

discovered that she had no sensation to void. Her bowel
function was normal. The patient was transferred to
rehabilitation medicine 2 weeks postoperatively for ongoing
physiotherapy. On admission to the rehabilitation unit, she
was independent for her activities of daily living and was
ambulating greater than 10 m with a walker. She was
discharged home the next day with arrangements for
outpatient physical therapy. Because of persistent mild
urinary retention, she was taught intermittent catheterization
to continue at home.
During the postoperative course, the patient was further
questioned, and medical records of her previous pregnancies
were reviewed in depth. After her first labor epidural,
discharge to the ward was delayed more than 6 hours for the
epidural motor block to regress, but no permanent neurologic

sequelae ensued. After her second labor epidural, she

complained of difficulty ambulating on the postpartum ward,
secondary to hip pain and weakness. This was assessed by
obstetrics and physiotherapy. She required discharge home
with a walker for ambulatory assistance for several days. Both
anesthesia records for these epidurals were reviewed and
appeared to be unremarkable. Neither of these unusual
recoveries was reported in her prior anesthetic history.
At 1 and 9 months' follow-up, she was able to ambulate
independently without aid and did not need intermittent
bladder catheterization. She was found to have mild
persistent left hip flexor weakness (4 +/5), dysesthesia at
T10-T12, and the development of mild neuropathic pain in
her posterior proximal left leg. At 12-month follow-up, she
had complete resolution of her residual motor and sensory
deficits, but her neuropathic pain persisted. Her final

Fig. 2 Unenhanced diffusion-weighted sagittal magnetic resonance imaging image of low thoracic and lumbar spine (left) and T2 image of
thoracic spine (right) demonstrating absence of spinal cord compression or cord lesion.

Persistent paralysis after spinal anesthesia for CD

diagnosis was neuropathic pain postincomplete spinal cord
injury due to anterior spinal artery syndrome during CD.

3. Discussion
The first case of confirmed anterior spinal artery
syndrome in pregnancy was reported in 1938 by Barre and
D'Andrade [8]. Five case reports of anterior spinal artery
syndrome in the obstetric population have been published
since that time. In 3 reports, epidural neuraxial anesthesia
was used, whereas the other 2 do not report whether
neuraxial anesthesia was used.
Ackerman et al [9] presents the case of a laboring
parturient with a continuous epidural infusion with epinephrine (1:400,000) and a presentation of mild hypotension (BP,
86/60 mm Hg). She underwent emergency CD with epidural
anesthesia and had significant blood loss (N 1500 mL).
Postoperatively, she developed anterior spinal artery syndrome. The etiology in this case was hypothesized to be the
combination of the mild hypotension with the vasoconstrictive properties of the epinephrine.
The hypercoagulable state of pregnancy may also
contribute to anterior spinal artery syndrome as suggested
in 2 postpartum cases of anterior spinal artery syndrome
[10,11]. In the case published by Dunn and Ellison [10], a
woman presented with an abrupt onset of flaccid paraplegia
of the abdomen and legs on postpartum day 20 after her
fourth uncomplicated pregnancy and delivery. There is no
comment on whether neuraxial anesthesia was used. Light
touch, vibratory, and position senses were intact leading to
the clinical diagnosis of anterior spinal artery infarction. The
cause of the infarction was not known; however, the authors
speculate that arterial infarcts may be due to hormonal effects
on cerebral vascular endothelium or to the combined effect of
preexisting atherosclerosis and the normal hypercoagulable
state of pregnancy.
Soda et al. [11] present the case of a 35-year-old woman,
6 days postpartum who presented with acute paraparesis,
sensory disturbance in her lower body, and urinary
disturbance. There is no comment on whether neuraxial
anesthesia was used. She was found to have thrombocytosis
and high fibrinolytic activity with improvement in symptoms
after administration of antithrombotic agents and methylpredonisolone. Coagulation disorders during the peripartum
period were assumed to be a cause of anterior spinal artery
syndrome in this case.
Eastwood presents the case of a diabetic scleroderma
patient [12]. The etiology of spinal artery ischemia was
postulated to be the combination of the volume of local
anesthetic (35 mL) injected and the collagen deposition in
the epidural space. These elements decreased spinal artery
blood flow and caused ischemia due to reduced compliance
of the epidural space. This hypothesis was confirmed by the
epidural venogram, which showed a compressed venous
pattern in the epidural space [12].

71
Ben-David et al [13] describes an unusual case of anterior
spinal artery syndrome. The delivery occurred 5 hours after
uneventful placement of an epidural for labor analgesia and 3
hours after the last dose of local anesthetic. Since the
delivery, the patient was unable to move her legs but retained
tactile sensation. The epidural catheter was promptly
removed; and within 30 minutes, she had complete
resolution of her paralysis without evidence of neurologic
dysfunction. The authors postulate that epidural catheter
irritation caused vasospasm of the anterior spinal artery or a
radicular artery leading to temporary ischemia of the anterior
spinal cord.
The etiology has not been confirmed in this case.
Arteriosclerosis seems highly unlikely to be contributory
given the young age and negative cardiac history of this
patient. Cesarean delivery is not a surgery that is considered
a risk for mechanical compression of the aorta for a
prolonged period. Although aortocaval compression is likely
in the supine term parturient, the patient was placed in left
lateral tilt, and systolic BP never demonstrated a reading
lower than 78 mm Hg.
The normal transesophageal echocardiogram ruled out
patent foramen ovale with possible embolic etiology. A
negative sickle cell screen and normal Hb electrophoresis
refute the possibility of vasoocclusive crisis. A neuraxial
catheter was not used, making catheter-induced vasospasm not
possible. Clinical evidence of posterior cord sparing is not
typical with an inflammatory etiology, making this less likely.
We hypothesize that the combination of mild hypotension with
the vasoconstricting properties of the phenylephrine infusion
and the hypercoagulable state of pregnancy may have all
contributed to spinal artery syndrome.
Although our patient's history of prolonged block with
her previous obstetric epidurals may be incidental, we cannot
rule out the possibility of whether this history is predictive of
an abnormal vascular supply with susceptible watershed
areas or an underlying predisposition to myelopathy that may
have contributed to spinal cord ischemia.
Furthermore, research in rats suggests that opioid agonists
may be neurotoxic in the setting of spinal cord ischemia via Nmethyl-D-aspartate receptor activation [14-17]. The authors
evoke the possibility that a small dose of intrathecal morphine
(30 g) may potentiate motor dysfunction after a short period of
artery ischemia (aortic balloon occlusion for 6 minutes). The
effect of the intrathecal morphine dose used in our case (5 times
what is described in the rat model) combined with a mild episode
of hypotension (less severe than complete aortic occlusion) is
not known. From this literature, it is possible, yet highly
unlikely, that intrathecal morphine or fentanyl may have
contributed to the etiology.
Although MRI imaging demonstrated absence of spinal
cord pathology, such as trauma or compression, it was not
contributory in determining our diagnosis. In fact, diffusionweighted MRI may assist in the diagnosis of cord ischemia,
but resolution is limited making it easy to miss an ischemic
or inflammatory lesion, especially in the early stage, where

72
acute ischemia is often not visible [5]. Our diagnosis of acute
spinal artery syndrome was based solely on the clinical evidence.
In conclusion, anterior spinal artery syndrome is a rare
cause of persistent paralysis after neuraxial anesthesia in the
obstetrical population. We cannot determine the exact
etiology in our present case but suggest multiple etiologies.

References
[1] D'Angelo R. Anesthesia-related maternal mortality: a pat on the back
or a call to arms? Anesthesiology 2007;106:1082-4.
[2] Rollins M, Lucero J. Overview of anesthetic considerations for
cesarean delivery. Br Med Bull 2012;101:105-25.
[3] Loo CC, Dahlgren G, Irestedt L. Neurological complications in
obstetric regional anaesthesia. Int J Obstet Anesth 2000;9:99-124.
[4] Wong CA. Nerve injuries after neuraxial anaesthesia and their
medicolegal implications. Best Pract Res Clin Obstet Gynaecol
2010;24:367-81.
[5] Kowe O, Waters JH. Neurologic complications in the patient receiving
obstetric anesthesia. Neurol Clin 2012;30:823-33.
[6] Scott DB, Hibbard BM. Serious non-fatal complications associated with
extradural block in obstetric practice. Br J Anaesth 1990;64:537-41.
[7] Gonzalez M. Vascular disease. In: Hines RL, Marschall KE, editors.
Stoeltings anesthesia and co-existing disease. 5th ed. Philadelphia:
Churchill Livingstone; 2008. p. 140.
[8] Barre J, D'Andrade C. Paraplegia par ramollissement aigu unisegmentaire
de la moelle, survenue au cours de la grossesse. Rev Neurol 1938;69:133-5.

V. Zaphiratos et al.
[9] Ackerman WE, Juneja MM, Knapp RK. Maternal paraparesis after
epidural anesthesia and cesarean section. South Med J 1990;83:
695-7.
[10] Dunn DW, Ellison J. Anterior spinal artery syndrome during the
postpartum period. Arch Neurol 1981;38:263.
[11] Soda T, Shimizu I, Takagaki M, et al. A case of benign postpartum
anterior spinal artery syndrome with thrombocytosis and high
fibrolytic activity. Brain Nerve 2011;63:1125-9.
[12] Eastwood DW. Anterior spinal artery syndrome after epidural
anesthesia in a pregnant diabetic patient with scleredema. Anesth
Analg 1991;73:90-1.
[13] Ben-David B, Vaida S, Collins G, Naum M, Gaitini L. Transient
paraplegia secondary to an epidural catheter. Anesth Analg 1994;79:
598-600.
[14] Kakinohana M, Nakamura S, Fuchigami T, Davison KJ, Marsala M,
Sugahara K. Mu and delta, but not kappa, opioid agonists induce
spastic paraparesis after a short period of spinal cord ischaemia in rats.
Br J Anaesth 2006;96:88-94.
[15] Kakinohana M, Marsala M, Carter C, Davison JK, Yaksh TL.
Neuraxial morphine may trigger transient motor dysfunction after a
noninjurious interval of spinal cord ischemia: a clinical and
experimental study. Anesthesiology 2003;98:862-70.
[16] Kakinohana M, Kakinohana O, Jun JH, Marsala M, Davison KJ,
Sugahara K. The activation of spinal N-methyl-D-aspartate receptors
may contribute to degeneration of spinal motor neurons induced by
neuraxial morphine after a noninjurious interval of spinal cord
ischemia. Anesth Analg 2005;100:327-34.
[17] Kakinohana M, Fuchigami T, Nakamura S, Sasara T, Kawabata T,
Sugahara K. Intrathecal administration of morphine, but not small
dose, induced spastic paraparesis after a noninjurious interval of aortic
occlusion in rats. Anesth Analg 2003;96:769-75.

Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.