REFRESHER COURSE OUTLINE R1

Current issues in
Spencer S. Liu MD spinal anesthesia

S
PINAL anesthesia is a venerable and simple hyperbaric lidocaine and isobaric bupivacaine is excel-
technique. Technical competence is achieved lent with CSF accounting for 80% of the variability for
early during training (> 90% technical success peak block height and regression of sensory and motor
rate) after only 40–70 supervised attempts.1 block.3 Unfortunately, volume of lumbosacral CSF
Although the ease and long history of spinal anesthe- does not correlate with external physical measurements
sia may give the impression that it is an unsophisticat- other than weight. Thus, CSF volume cannot be easily
ed technique, much has been recently learned estimated from physical examination and is not easily
regarding the anatomy, physiology, pharmacology, applied to the clinical setting.3
and applications of spinal anesthesia.
Physiology
Anatomy Cardiovascular
The arachnoid membrane is an important structure, as The most common serious side effects from spinal
spinal agents must be delivered within its confines. The anesthesia are hypotension and bradycardia, and
arachnoid membrane is composed of overlapping lay- closed claims surveys of 40,000–550,000 spinal anes-
ers of epithelial cells connected by tight junctions. This thetics indicate an incidence of cardiac arrest from
anatomic arrangement allows the arachnoid mem- 0.04–10/10,000.4,5 Large surveillance studies typical-
brane, not the dura, to function as the principal ly observed incidences of hypotension around 33%
meningeal barrier (90% of resistance) to materials and bradycardia around 13% in non-obstetrical popu-
crossing in and out of the cerebrospinal fluid (CSF). lations.4 Risk factors for hypotension in non-obstetri-
The arachnoid membrane serves not only as a passive cal populations include block height $ T5, age $ 40
container of CSF but also actively processes and trans- yr, baseline systolic blood pressure (SBP) < 120
ports agents attempting to cross the meninges. Recent mmHg, and spinal puncture above L3 $ 4. Risk fac-
studies demonstrate that metabolic enzymes are tors for development of bradycardia in non-obstetrical
expressed in the arachnoid membrane that can affect populations include baseline heart rate (HR) < 60
agents (e.g., epinephrine) and neurotransmitters beats·min–1, ASA I, use of beta blockers, prolonged
important for spinal anesthesia (e.g., acetylcholine).2 PR interval on electrocardiogram, and block height $
Active transport of compounds across the arachnoid T5. Analysis of closed claims for cardiac arrest during
membrane occurs in the area of the neural root cuffs spinal anesthesia indicated that administration of seda-
where unidirectional transport of materials from the tion to produce a sleep-like state without spontaneous
CSF into the epidural space occurs and may contribute verbalization and lack of early administration of epi-
to clearance of spinal anesthesia agents. After injection nephrine were common patterns of management in
of spinal anesthetics, dilution with the CSF occurs cases of cardiac arrest.5
prior to arrival at effector sites in the CNS. Thus, indi- Cardiovascular effects of spinal anesthesia typically
vidual variation in lumbosacral volumes of CSF and include a decrease in arterial blood pressure and central
distribution within this volume will affect spinal anes- venous pressure (CVP) with only minor decreases in
thesia. Recent imaging with magnetic resonance HR, stroke volume, or cardiac output even in patients
demonstrates great variability between individuals in with poor left ventricular function (ejection fraction <
volume of lumbosacral CSF with a range of 28–81 mL. 50%).6 Hypotension occurs from decreases in systemic
Interestingly, obese individuals have substantially less vascular resistance (SVR) and CVP from sympathetic
CSF (~10 mL less) that is partly due to compression of block with vasodilation and redistribution of central
the neural foramina. Clinical correlation between vol- blood volume to lower extremities and splanchnic
ume of lumbosacral CSF and spinal anesthesia with beds.6 This sympathetic block is rarely complete and

From the Departments of Anesthesiology, Virginia Mason Medical Center and the University of Washington, Seattle, USA.
Address correspondence to: Dr. Liu, Department of Anesthesiology, Virginia Mason Medical Center, 1100 Ninth Avenue, Mail Stop B2-
AN, Seattle, Washington, USA 98101. Phone: 206- 223-6980; Fax: 206-223-6982; E-mail: anessl@vmmc.org
Financial support: Funded by the Department of Anesthesiology, Virginia Mason Medical Center.

CAN J ANESTH 2002 / 49: 6 / pp R1–R5

R2 CANADIAN JOURNAL OF ANESTHESIA

some preservation of sympathetic reflexes to stressful There are three main mechanisms for development of
challenge typically occurs. Sudden bradycardia can core hypothermia after spinal anesthesia.10 The first is
occur from shift in cardiac autonomic balance towards redistribution of central heat to the periphery due to
the parasympathetic system as evidenced in spectral vasodilation from sympathetic block. This effect is
analysis of heart rate variability, from activation of left maximal during the first 30–60 min, causes a decrease
ventricular mechanorecptors from a sudden decrease in in core temperature of about 1–2°C, and depends on
left ventricular volume (Bezold Jarisch reflex), or from extent of sensory block and patient age. The second
increases in baroreflex activity.5 mechanism is loss of thermoregulation characterized
Prophylactic measures to prevent hypotension by reduced shivering and vasoconstriction thresholds
include prehydration with crystalloids or colloids or during spinal anesthesia. Finally, with loss of ther-
administration of vasoactive agents. On the whole, pre- moregulatory vasoconstriction below the level of the
hydration with crystalloids (250–2000 mL) appears to sympathetic block, there is increased heat loss from
temporarily increase preload and cardiac output with- vasodilation. If hypothermia develops, patients should
out consistently increasing arterial blood pressure or be rewarmed, and spinal anesthesia accelerates rewarm-
preventing hypotension.4 Pharmacokinetics of crystal- ing compared to general anesthesia due to the residual
loid explain its poor efficacy, as crystalloids are quickly sympathetic block and vasodilation.10
re- distributed from the intravascular to the extravas-
cular space. Administration of large volumes (> 1 L) of Hypnotic effects
a crystalloid does not appear to confer additional ben- There has been a recent convergence in mechanisms
efit over small volumes (250 mL), and may be detri- of general and spinal anesthesia. MAC, a traditional
mental to patients with limited cardiopulmonary measure of inhalation agent potency for depth of anes-
reserve. Prehydration with a colloid ($ 500 mL) thesia appears to have a primary mechanism in the
appears to be more effective than with a crystalloid at spinal cord.11 In contrast, central neuraxial anesthesia
maintaining arterial blood pressure and perhaps may have direct effects on suppression of conscious-
decreasing the incidence of hypotension depending on ness, and multiple studies have observed that patients
definition and population.7 In contrast to prophylaxis, appear drowsy after spinal anesthesia despite the lack
treatment of hypotension during spinal anesthesia will of sedative medications.12 Correspondingly, both
be effective with crystalloid or colloid due to changes spinal and epidural anesthesia reduce the hypnotic
in kinetics induced by spinal anesthesia and intravascu- requirements of midazolam, isoflurane, sevoflurane,
lar hypovolemia.8 Both clinical scenarios alter kinetics and thiopental in surgical patients and laboratory
of cystalloid and colloid to allow retention within the studies.11 Possible mechanisms for hypnosis during
intravascular space. Prophylactic administration of spinal anesthesia include rostral spread of local anes-
pharmacologic agents may be more effective than pre- thetics or decrease in reticular activating system activ-
hydration for prevention of hypotension.9 Alpha- ity due to interruption of afferent input. Animal
adrenergic agonists (e.g., phenylephrine) reliably models support the latter, as spinal anesthesia in rats
increase arterial blood pressure by increasing SVR, decreases hypnotic requirements of thiopental without
however heart rate and cardiac output may decrease detection of local anesthetic in the brain or cervical
due to increased afterload. Mixed alpha and beta- spinal cord. In humans, the degree of sedation due to
adrenergic agents (e.g., ephedrine) are also effective for spinal anesthesia is related to peak block height with
increasing arterial blood pressure and preventing greater sedation observed with greater block
hypotension but act by primarily increasing heart rate heights.12 Clinical relevance for these observations is
and cardiac output with a smaller increase in SVR.6 the decreased need for pharmacologic sedation with
These different physiologic mechanisms for alpha vs the use of spinal anesthesia.
mixed alpha and beta-adrenergic agents also occur dur-
ing treatment of hypotension during spinal anesthesia.6 Clinical applications
Ambulatory anesthesia
Thermoregulation Spinal lidocaine has been a popular choice for ambu-
Mild perioperative hypothermia is associated with an latory spinal anesthesia, and recent studies have exam-
increased incidence of myocardial ischemia, cardiac ined dose response effects of lidocaine on anesthesia
morbidity, wound infection, blood loss and transfusion and recovery (Table).4 Transient neurologic symp-
requirements. Spinal anesthesia will predictably cause toms (TNS) are clearly associated with use of spinal
core hypothermia within 30–60 min, and patients lidocaine with an approximate incidence of 20% in the
should be monitored and actively warmed if needed. ambulatory setting.13 Concern over the potential for
 R3

TABLE Typical dose response effects of spinal local anesthetics for ambulatory anesthesia

Local anesthetic Dose Peak Duration of Duration Time from induction Anesthetic
block sensory block of motor block until discharge success rate
(mg) (min) (min) (min) (%)
Lidocaine 30 0
(isobaric) 40 T4 (T2-10) 130 (26) 93 (24) 178 (34) 90
60 T3 (T2-10) 162 (32) 128 (31) 216 (33) 90
80 T3 (T1-7) 170 (24) 142 (32) 236 (46) 97

Bupivacaine 5 T5 (T4-7) 123 (27) 50 (20) 181 (30) 75

(hyperbaric) 7.5 T8 (T4-11) 144 (25) 75 (24) 202 (28) 100
10 T8 (T6-10) 194 (26) 100 (24) 260 (30) 100

Mepivacaine 30 T9 (T2-L5) 158 (32) 116 (38) 180 (34) 72

(isobaric) 45 T6 (T2-12) 182 (38) 142 (37) 191 (29) 100
60 T5 (T2-L1) 203 (36) 168 (36) 203 (35) 100

Ropivacaine 8 T9 (T4-L1) 130 (27) 107 (25) 165 (45) 63

(isobaric) 10 T8 (T4-L2) 152 (44) 135 (31) 174 (38) 83
12 T8 (T4-L1) 176 (42) 162 (37) 199 (52) 93
14 T9 (T3-L1) 192 (48) 189 (44) 233 (52) 100

Procaine 100 T5 (T1-10) 120 (23) 100 (30) 244 (43) 83

(hyperbaric)
Prilocaine 50 T6 (T1-10) 128 (38) 165 (37) 253 (55) 100
(hyperbaric)

neurologic injury and for patient comfort has led to be less reliable for surgical anesthesia than lidocaine
interest in alternative spinal local anesthetics. while having a slower recovery (Table). Risk of TNS
Bupivacaine has been the most studied alternative to with procaine (~ 6%) is less than with lidocaine but
lidocaine. TNS is virtually absent in all clinical studies probably greater than with bupivacaine.17 Prilocaine is
with spinal bupivacaine (0–1%).14 Recent dose- approximately equipotent to lidocaine within a dose
response data on clinical anesthetic characteristics for range of 40–70 mg4 and thus may have suitable clini-
spinal bupivacaine (Table) indicate that small doses cal characteristics for ambulatory spinal anesthesia
can be used for ambulatory anesthesia.4 It is particu- (Table). Risk of TNS appears to be minimal with
larly important to select small doses of bupivacaine (# spinal prilocaine (0–1%).
10 mg) to avoid prolonged detrusor block, inability to Both anesthetic success and especially time until
void, and excessively prolonged time until discharge as ready for discharge are dependent on dose of local
compared to equipotent doses of lidocaine. anesthetic. There has been recent interest in using
Mepivacaine has anesthetic characteristics similar to analgesic additives to spinal local anesthetics to
lidocaine with an approximate potency of 1.3:1 decrease the dose of local anesthetic for faster recov-
(Table).15 Reported risk of TNS with mepivacaine is ery while maintaining or improving anesthetic success.
highly variable. Small-scale studies (60–75 patients) Addition of vasoconstrictors (epinephrine and
report a low incidence of TNS (0–8 %), whereas larg- phenylephrine) are effective for prolonging and inten-
er studies (200+ patients) report incidences of ~ 30%.4 sifying spinal anesthesia but are ill advised for ambula-
It seems mepivacaine has similar clinical characteristics tory surgery due to delay in patient recovery and
as lidocaine for spinal anesthesia but likely shares the potential increased risk of TNS.4 Numerous clinical
same risk of TNS. Ropivacaine is approximately 50 to studies have demonstrated that addition of 10–25 µg
60% as potent as spinal bupivacaine. Like bupivacaine, of fentanyl improves success of spinal anesthesia,
there is little risk of TNS with the use of spinal ropi- allows use of less local anesthetic, and does not pro-
vacaine (0–1% incidence) and, in equipotent doses long duration until discharge. For example, 10 µg of
(2:1), it will be virtually indistinguishable from bupi- fentanyl added to 5 mg of hyperbaric bupivacaine for
vacaine for clinical anesthesia and risk of TNS without outpatient knee arthroscopy improved anesthetic suc-
any obvious advantages.16 Spinal procaine appears to cess from 75% with plain bupivacaine to 100%.4
R4 CANADIAN JOURNAL OF ANESTHESIA

A dose of 7.5 mg plain bupivacaine is needed to cal use of small gauge pencil point needles. These clin-
achieve similar success with resultant prolongation of ical observations lessen the plausibility of a concentra-
time until discharge of 187 to 202 min when com- tion dependent neurotoxic etiology. Finally, successful
pared to 5 mg + fentanyl. Similar findings have been treatment of TNS with trigger point injections and
observed with addition of 10–25 µg of fentanyl to NSAIDs also fail to substantiate neurologic injury as
spinal lidocaine in patients undergoing ambulatory an etiology. All together, these data may indicate a
laparoscopy, in vitro fertilization, and knee myofascial etiology for the radiating backpain, and
arthroscopy. Addition of 25 µg fentanyl to 20 mg some experts have called for a change in nomenclature
spinal lidocaine in patients having knee arthroscopy from TNS to postspinal musculoskeletal symptoms.23
provides comparable anesthesia to 50 mg lidocaine
while providing shorter discharge times (145 vs 180 Low molecular weight heparin (LMWH)
min) and reduced incidence of TNS (4% vs 33%).18 Anticoagulants are frequently used in the surgical pop-
Indeed, discharge times after such small doses of local ulation as prophylaxis and treatment for thrombotic
anesthetic are comparable to local anesthesia/propo- conditions. Analysis of closed claims for neurologic
fol infusion.19 Dose response data for spinal clonidine injury indicates that anti-coagulation is a major risk
suggests that a dose of 15–45 µg is an optimal dose for factor for spinal cord injury with spinal anesthesia.24
low dose outpatient spinal anesthesia. This dose LMWH is a fractionated component of standard
improved anesthetic success of 8 mg ropivacaine from heparin and has become a popular agent. LMWH has
60% to 100% for ambulatory knee arthroscopy with- much greater bioavailability than unfractionated
out prolonging recovery.20 heparin, primarily affects coagulation factor X, and
cannot be monitored with partial thromboplastin
TNS time.25 Over a decade of European experience sug-
Prospective randomized studies reveal an incidence of gested that perioperative use of LMWH did not add
TNS after lidocaine spinal anesthesia between substantial risk to spinal anesthesia. However, the US
4–33%.13 Risk of TNS is increased with use of lido- experience has been different with > 40 cases of spinal
caine, ambulatory anesthesia, lithotomy and knee hematomas since its introduction in 1993. In contrast
arthroscopy positions and is unaffected by baricity or to European experience of relative safety, estimated
dilution of lidocaine to 0.5%.13 TNS typically occur risk of spinal hematoma with LMWH and spinal anes-
12–36 hr after resolution of spinal anesthesia, last for thesia in the US is 1:41,000 vs the 1:225,000 in the
two to three days, and is typically rated as a 3–4/10 non-anticoagulated patient.25 Larger daily doses and
for pain intensity (0=no pain, 10=worst pain). more frequent administration with US practice may
Discomfort from TNS is self-limited and can be treat- account for this apparent increase in risk. Guidelines
ed with potent non-steroidal anti-inflammatory drugs to safe use of LMWH and spinal anesthesia include
(NSAIDs) and trigger point injections.21 delay of administration of LMWH for 12–24 hr after
Neurotoxic etiology for TNS remains speculative. spinal puncture. If the patient is already using
Patients reporting TNS do not develop sensory or LMWH, then it should be stopped for at least 12–24
motor deficits in contrast to cauda equina syndrome. hr prior to spinal puncture.25
Imaging of the CNS does not show evidence of injury
to spinal cord or nerve roots in patients with TNS. Conclusion
Sensitive measures of neural electrophysiology (SSEP, Spinal anesthesia is an old, simple, and popular anes-
EMG, nerve conduction velocity, H reflex, F waves) thetic technique, yet much remains unknown regarding
do not change during TNS as compared to before pertinent anatomy, physiology, and pharmacology.
spinal anesthesia.22 Laboratory work in both intrathe- Investigations into physiologic effects of spinal anesthe-
cal and desheathed peripheral nerve models indicate sia reveal complex actions on multiple organ systems.
that concentration of lidocaine is a critical factor in the New local anesthetics, analgesic additives, and tech-
neurotoxicity21 of desheathed peripheral nerves. Yet, niques are being investigated for different applications
clinical trials report high incidences of TNS (17%) as the practice of medicine focuses on outpatient care.
with spinal injection of very dilute lidocaine concen- Safety of spinal agents and complications from spinal
trations (0.5%, 1%)13 that touch upon the minimal anesthesia continue to be examined and re-examined to
effective concentration for spinal lidocaine (0.0–07%). improve safety. Further study will be needed to fully
Indeed, further dilution of lidocaine should occur due resolve these issues and to further understand and
to active mixing in spinal CSF after non-preferential improve the clinical use of spinal anesthesia.
distribution of hyperbaric solution with typical, clini-
 R5

References combined spinal epidural technique. Anesth Analg

1 Konrad C, Schupfer G, Wietlisbach M, Gerber H. 1999; 89: 167–71.
Learning manual skills in anesthesiology: is there a rec- 16 Gautier PE, DeKock MF, Van Steenberge A. Intrathecal
ommended number of cases for anesthetic procedures? ropivacaine for ambulatory surgery: a comparison
Anesth Analg 1998; 86: 635–9. between intrathecal bupivacaine and intrathecal ropiva-
2 Ummenhofer WC, Brown SM, Bernards CM. caine for knee arthroscopy. Anesthesiology 1999; 91:
Acetylcholinesterase and butyrylcholinesterase are 1239–45.
expressed in the spinal meninges of monkeys and pigs. 17 Hodgson PS, Liu SS, Batra MS, Gras TW, Pollock JE,
Anesthesiology 1998; 88: 1259–65. Neal JM. Procaine compared to lidocaine for incidence
3 Carpenter RL, Hogan QH, Liu SS, Crane B, Moore J. of transient neurologic symptoms. Reg Anesth Pain
Lumbosacral cerebrospinal fluid volume is the primary Med 2000; 25: 218–22.
determinant of sensory block extent and duration dur- 18 Ben-David B, Maryanovsky M, Gurevitch A, et al. A
ing spinal anesthesia. Anesthesiology 1998; 89: 24–9. comparison of minidose lidocaine-fentanyl and conven-
4 Liu SS, McDonald SB. Current issues in spinal anesthe- tional-dose lidocaine spinal anesthesia. Anesth Analg
sia. Anesthesiology 2001; 94: 888–906. 2000; 91: 865–70.
5 Pollard JB. Cardiac arrest during spinal anesthesia: 19 Ben-David B, DeMeo PJ, Lucyk C, Solosko D. A compar-
common mechanisms and strategies for prevention. ison of minidose lidocaine-fentanyl spinal anesthesia
Anesth Analg 2001; 92: 252–6. and local anesthesia/propofol infusion for outpatient
6 Butterworth J. Physiology of spinal anesthesia: what are knee arthroscopy. Anesth Analg 2001; 93: 319–25.
the implications for management? Reg Anesth Pain 20 De Kock M, Gautier P, Fanard L, Hody JL,
Med 1998; 23: 370–3; Discussion 384–7. Lavand’homme P. Intrathecal ropivacaine and clonidine
7 Ueyama H, Yan-Ling H, Tanigami H, Mashimo T, for ambulatory knee arthroscopy: a dose-response
Yoshiya I. Effects of crystalloid and colloid preload on study. Anesthesiology 2001; 94: 574–8.
blood volume in the parturient undergoing spinal anes- 21 Hodgson PS, Neal JM, Pollock JE, Liu SS. The neuro-
thesia for elective cesarean section. Anesthesiology toxicity of drugs given intrathecally (spinal). Anesth
1999; 91: 1571–6. Analg 1999; 88: 797–809.
8 Hahn RG, Resby M. Volume kinetics of Ringer’s solu- 22 Pollock JE, Burkhead D, Neal JM, et al. Spinal nerve
tion and dextran 3% during induction of spinal anaes- function in five volunteers experiencing neurologic
thesia for caesarean section [published erratum appears symptoms after lidocaine subarachnoid anesthesia.
in Can J Anaesth 1998; 45: 1223–4]. Can J Anaesth Anesth Analg 2000; 90: 658–65.
1998; 45: 443–51. 23 Rowlingson JC. To avoid “transient neurologic symp-
9 Chan WS, Irwin MG, Tong WN, Lam YH. Prevention toms” - The search continues. RAPM 2000; 25:
of hypotension during spinal anaesthesia for caesarean 215–7.
section: ephedrine infusion versus fluid preload. 24 Cheney FW, Domino KB, Caplan RA, Posener K. Nerve
Anaesthesia 1997; 52: 908–13. injury associated with anesthesia. A closed claims analy-
10 Sessler DI. Perioperative heat balance. Anesthesiology sis. Anesthesiology 1999; 90: 1062–9.
2000; 92: 578–96. 25 Horlocker TT, Wedel DJ. Neurologic complications of
11 Hodgson PS, Liu SS, Gras TW. Does epidural anesthesia spinal and epidural anesthesia. Reg Anesth Pain Med
have general anesthetic effects? Anesthesiology 1999; 2000; 25: 83–98.
91: 1687–92.
12 Pollock JE, Neal JM, Liu SS, Burkhead D, Polissar N.
Sedation during spinal anesthesia. Anesthesiology
2000; 93: 728–34.
13 Pollock JE, Liu SS, Neal JM, Stephenson CA. Dilution of
spinal lidocaine does not alter the incidence of tran-
sient neurologic symptoms. Anesthesiology 1999; 90:
445–50.
14 Freedman JM, Li DK, Drasner K, Jaskela MC, Larsen
B, Wi S. Transient neurologic symptoms after spinal
anesthesia: an epidemiologic study of 1,863 patients.
Anesthesiology 1998; 89: 633–41.
15 Zayas VM, Liguori GA, Chisolm MF. Dose response
relationships for isobaric spinal mepivacaine using the