Current Concerns with Neuraxial Anesthesia

From the 14th Annual Advances in Physiology and Pharmacology in Anesthesia and Critical Care, sponsored by
Wake Forest University School of Medicine, November 2-5, 2008, White Sulphur Springs, WV

Educational Objectives

The goal of this program is to improve neuraxial anesthesia and its associated problems. After hearing and assimilat­ing this program, the clinician will be better able to:

1.   Discuss the pathophysiology of perioperative neurologic injury.

2.   Consider (and put into perspective) the relative frequency of neurologic injuries associated with regional anes­thesia and pain medicine practice.

3.   Anticipate the diagnosis and management of neurologic injury and reduce the likelihood of complications.

4.   Exonerate “bad bupivacaine” as a cause of failed spinal anesthesia.

5.   Explore true causes of failed spinal anesthesia and choose a safe procedure for cases in which spinal failure oc­curs.

Faculty Disclosure

In adherence to ACCME Standards for Commercial Support, Audio-Digest requires all faculty and planning commit­tee members to disclose relevant financial relationships within the past 12 months that might create any personal con­flicts of interest. Any identified conflicts were resolved to ensure that this educational activity promotes quality in health care and not a proprietary business or commercial interest. For this program, the faculty and planning commit­tee reported nothing to disclose.

Acknowledgments

Drs. Neal and Nelson spoke in White Sulphur Springs, WV, at Fourteenth Annual Advances in Physiology and Phar­macology in Anesthesia and Critical Care, held November 2-5, 2008, and sponsored by Wake Forest University School of Medicine. The Audio-Digest Foundation thanks the speakers and the Wake Forest University School of Medicine for their cooperation in the production of this program.

Anesthesia-related Neuraxial Injury: Beyond Epidural Hematoma

Joseph M. Neal, MD, Clinical Professor of Anesthesiology, University of Washington School of Medicine, and Staff Anesthesiologist, Virginia Mason Medical Center, Seattle

Risk: difficult to obtain incidence or frequency data; American Society of Anesthesiologists’ (ASA) Closed Claims Study shows most complications resulting in lawsuits involve hematoma; needle injury accounts for only 17% of lawsuits; surveillance study conducted over 10-mo period, looking at use of regional anesthetics in France; few complications reported; medicolegal data may overestimate occurrence of injury, while clinical studies potentially underestimate incidence; clinical data from Sweden show only 127 neurologic injuries in 1.8 million anesthetics (only 9 of these needle injuries); incidence of needle injury to spinal cord »1 in 200  000; when findings for both neuraxial and peripheral blocks combined, risk for major permanent complications 0 to 4 in 10  000

Pathophysiology of neuraxial injury: typically linked to needle- or catheter-related mechanical damage (may in­volve spinal cord, spinal nerve roots, or spinal vasculature) or mass lesion competing for limited area within spi­nal canal

Mechanical: frequently, spinal cord injuries due to inaccurate determination of vertebral levels or anatomic varia­tion in terminal portion of spinal cord; further complications include inconsistent and unreliable signals when needles or catheters contact or enter spinal cord (eg, ligamenta flava incompletely fused, progressive [caudad to cephalad] epidural space narrowing); another anatomic problem with spinal column involves intraventricu­lar foramen; large amount of local anesthetic tends to egress through foramina; with age, space becomes more narrow and more difficult for blood to dissipate; risk for injury to neuraxis, especially during interscalene anes­thesia; spinal nerve root can be injured if farther lateral than suspected

Vascular injury: location of spinal cord blood supply can be uncertain, particularly the supply to lumbosacral and lower thoracic segments; damage to arteria radicularis magnus (ie, artery of Adamkiewicz) can potentially compromise spinal cord circulation; if blood pressure within realm of autoregulation, and no reason to suspect spinal cord injury, then anything within autoregulation should not cause spinal cord ischemia; local anesthetics near spinal cord decrease metabolic rate; spinal cord regulates blood supply based on metabolic needs (meta­bolic needs lower with local anesthetics); spinal cord relatively resistant to low flow within realm of autoregu­lation; animal studies indicate epinephrine does not negatively affect blood flow to spinal cord; vascular accidents (eg, anterior spinal artery syndrome) almost always indicate underlying anatomic abnormality, often in elderly patient with atherosclerosis and hypertension; problems may occur with transforaminal particulate steroids (spinal cord blocked by particles); celiac plexus block may potentially cause hematoma and reduce blood flow to spinal cord

Mass lesion: includes, eg, patient with prostate cancer who has epidural metastases; spinal cord crowded by metas­tasis; imaging helps avoid metastasis; abscess and hematoma compete with spinal cord for fixed space; reports of large volumes of local anesthetic, in pregnant women with epidural lipomatosis during course of management of labor, resulted in some degree of injury to spinal column; spinal stenosis involving hypertrophic ligamentum flava and bony outgrowth (particularly with age) does not necessarily cause spinal cord injury, “but it gives you very little wiggle room” if, eg, blood or pus begins competing for space; final common pathway involves mass effect compressing spinal cord; ultimately reaches height where compression of arterial and venous supply to spinal cord occurs; spinal cord becomes ischemic, thus resulting in infarction

Pathophysiology of peripheral injury: “double crush theory” neither confirmed nor refuted in humans; if nerve has sustained relatively small subclinical injury (from, eg, diabetic neuropathy, neurotoxin), another small subclinical injury may potentially cause damage to nerve; difficult for needle to cut nerve; if fascicle penetrated, neurons ex­posed to local anesthetics that can cause injury; epinephrine decreases clearance of local anesthetic, thereby en­hancing time-dependent component of injury

Diagnosis and treatment: among imaging modalities, magnetic resonance imaging (MRI) preferable to computed tomography (CT), although diagnosis should not be delayed if only CT available; likelihood of recovery diminishes as time to decompression approaches 8 hr; diagnosis of suspected peripheral nerve injuries involves nerve conduc­tion studies and electromyelography (EMG); traditional to wait 2 to 3 wk for neurologic consultation; however, with severe injury, studies should be obtained before 2 to 3 wk, and both injured and noninjured side should be in­cluded to establish preexisting disease and help follow progression of lesions; suspected compressive lesions of neuraxis require rapid diagnosis and treatment; decompression of epidural hematoma should occur in £1 hr; for small sensory lesion in periphery that appears to be resolving, wait 2 wk before referring to neurologist; if patient awakens with complete lesion, progressive lesion, or lesion with motor effects, “those are prognostically much more scary,” but may be reversible; peripheral injuries relatively common in first few weeks after anesthetic block­ade, but quickly disappear; neuraxial injury more uncommon and prognosis always guarded

Recommendations from practice advisory: controversial  —whether to perform blocks in patients who are either anesthetized or sedated to point of being unable to recognize and/or report any sensation that anesthesia provider would interpret as atypical during block placement; those who advocate blocks in anesthetized patient argue that if patient not anesthetized or heavily sedated, he or she would never accept regional anesthesia (RA) and accompany­ing benefits; furthermore, anesthetized or heavily sedated patient will not move and therefore potentially impale themselves on needle (particularly cogent argument in pediatric cases); those who do not advocate blocks in anes­thetized patients believe it eliminates patient reporting and provider appreciation of significant paresthesia or pain on injection (may, eg, eliminate warning signs of local anesthetic toxicity); furthermore, proponents note neural in­jury (eg, seizures, cardiac arrest) has been reported in both anesthetized and awake patients; no definitive studies in­dicating patient reporting of atypical sensation prevents nerve injury (randomized clinical trials unlikely because injuries “incredibly rare” and large numbers of patients required for statistical validity); neuraxial injury  —  when placing peripheral nerve block near neuraxis, planned interscalene block may actually become neuraxial block; case reports of peripheral nerve stimulation in anesthetized patient, resulting in unexpected permanent nerve injury; however, even awake patient may not detect needle entering spinal cord (sensory innervation to meninges inconsis­tent); likewise, entry of needle into spinal cord does not necessarily result in injury; nerve damage suspected when needle in spinal cord and then local anesthetic injected (pressure of injection results in large afferent nonspecific in­put); stopping injection means damage does not increase; peripheral injury  —  pain on injection does not always in­dicate damage to peripheral nerve; when needle enters peripheral nerve, likely that needle intraneural but extrafascicular; concern increases only when needle enters fascicle; patient may not have pain on injection; recommendations  —  do not routinely perform neuraxial or peripheral blockade in anesthetized patient (however, risk of child moving and damaging themselves on needle outweighs risk of blockade while child anesthetized); in­terscalene block should not be performed under general anesthesia (GA) or heavy sedation; no evidence that ability to use peripheral nerve stimulator or ultrasonography will affect incidence of nerve injury; avoid pain on injection by stopping if indicated pain greater than expected; data from patients with preexisting neurologic disease neither confirms nor refutes traditional preference for avoiding regional blockade; provider and patient should discuss risk-benefit ratio; if decision made to perform regional block, consider lower concentration local anesthetic and poten­tially avoid epinephrine in patient with diminished blood flow

Why do Spinal Blocks Fail?

Kenneth E. Nelson, MD, Associate Professor of Anesthesiology, Wake Forest University School of Medicine, and Staff Anesthesiologist, Department of Obstetrical Anesthesiology, Wake Forest University Baptist Medical Center, Winston-Salem, NC

Possible causes (brief overview)

Judgment failure: inadequate dose; inappropriate drug; drug error; lack of adjunct when indicated; poor match of anesthetic to surgical procedure; baricity and positioning error

Possible technical failure: free flow of cerebrospinal fluid (CSF), but no aspiration into syringe; movement of nee­dle during injection (eg, patient movement; checking for aspiration halfway [but lack of fluid with combined spi­nal-epidural {CSE} does not portend failure])

Probable technical failure: no free flow of CSF or aspiration; no appearance of CSF at any time (case reports of successful “dry tap” spinal anesthesia)

Technical success but block failure: incomplete block; complete failure (no discernible block); experience of Wake Forest University over 3 yr shows spinal failure incidence 2.7% overall; 47% repeated spinal or CSE; 47% con­verted to GA; 6% converted to epidural anesthesia; series of case reports found clear fluid aspirated before and after injection; no block resulted; several other reports in literature; many anecdotal reports as well

“Bupivacaine gone bad”: perhaps most popular explanation for failed spinal anesthesia; entropy described by sec­ond law of thermodynamics; universal law governing all molecules (including bupivacaine); energy and matter change in one direction, from ordered state to disordered state; rate of change of energy dependent and somewhat predictable; if degradation were due to improper storage, spectrum of failure would be expected; in contrast, re­ports rare for complete failure with success of surrounding blocks in same lot number; for sake of argument, con­sider bupivacaine violates universal laws to decay on L-shaped curve; small doses of spinal fentanyl reliably cause mid-thoracic pin-prick levels; case reports refer to complete absence of block levels (implies that whatever hap­pened to bupivacaine also happened to fentanyl); in vitro testing indicates bupivacaine (and other amides) ex­tremely heat stable (no degradation with storage at 250°F for 15 min; with long-term storage at >100°F, degradation insignificant)

Local anesthetic “resistance”: rachiresistance first attributed to abnormally thick myelin; use of term as synonym for spinal failure now out of favor; tachyphylaxis requires repeat or prolonged exposure (not relevant to single-shot spinal anesthetics); case report  —  failed spinal in cesarean delivery; skin wheals also failed; patient reported previ­ous failed spinal for cesarean delivery and inability to obtain anesthesia for dental procedures; authors postulated sodium channel mutation; speaker skeptical of conclusion

Anatomic abnormalities: Tarlov cyst  —  common (also known as sacral perineural cyst); CSF-containing cystic mass; usually presacral; communication with lumbar CSF variable; almost always incidental finding; other cysts  —  synovial; ganglion; arachnoid; dermoid; cystic neuromas; syrinx (intraparenchymal); CSF cysts  —  in several case reports of imaging conducted after failed spinal, none demonstrated cystic lesion as cause; CSF volume  —  lumbosacral CSF volume primary determinant of extent of sensory block and duration during spinal anesthesia; 10 healthy volunteers received 50 mg spinal lidocaine; axial MRI measurements of lumbar CSF volume compared to clinical parameters (ie, pin-prick level, motor block, and time to regression); higher CSF volume correlated with lower spinal block; quicker regression correlated with higher volume; case report of 37-yr-old woman planning postpartum tubal ligation; patient received hyperbaric 2% mepivacaine, followed by another appropriate dose of mepivacaine, and then addition of large dose of 5% lidocaine; one-sided S1 level and no motor block; uneventful GA and recovery; additive dose of local anesthetic “really pushing the envelope,” especially with restricted spread of toxicity; postoperative MRI revealed extremely large CSF volume; dural ectasia  —  case report of Marfan syn­drome in patient scheduled for elective cesarean delivery; continuous spinal catheter; 21 mg bupivacaine incremen­tally dosed; eventual T10 level; converted to GA; spinal canal completely filled with CSF; physical barriers  —  human spinal arachnoid septa, trabeculae, and “rogue strands”; 62 human cadaver spinal cords dissected; investi­gation not designed to predict spinal failure; findings suggest no barriers to CSF flow

Regional CSF flow velocity: »500 mL CSF produced daily in choroid plexus; absorbed intracranially; net flow intra­cranial; traditionally thought lumbar CSF stagnant pool; now know spinal CSF oscillates vigorously with arterial pulsations; Higuchi and colleagues (2004) found flow velocity affects clinical duration and peak levels of spinal an­esthesia; Eisenach (2003) found CSF mixing primary determinant of early concentration changes; flow rate highly variable among individuals

Maldistribution: injected drug restricted to small volumes of CSF; term popularized by reports of nerve injury with spinal microcatheter; sufficiently describes myriad scenarios; thoroughly discussed in RA literature; concept com­plementary with fact that there is high individual variation in CSF volume and CSF flow rates vary during cardiac cycle; safe care of patient most important consideration; maldistribution must be assumed when spinal block fails; assuming local anesthetic has “gone bad” can be hazardous; maldistribution may cause restricted intense sacral block

Conclusion: no “bad bupivacaine”; anomalous resistance to local anesthetics extremely rare; anomalous anatomy also uncommon, but possible in rare cases; maldistribution most likely; do not jeopardize patient safety by assum­ing local anesthetic bad; instead, assume every spinal injection has delivered drug to subarachnoid space; check sacral dermatomes for restricted block; when reinjecting, choose new interspace, keep total dose within safe maxi­mum, omit morphine (use other adjuncts with caution), be prepared for high block and prolonged block; consider postponing elective case; always consider plan B

Suggested Reading

Bernard CM et al: Regional anesthesia in anesthetized or heavily sedated patients. Reg Anesth Pain Med 33:449, 2008; Carpenter RL et al: Lumbosacral cerebrospinal fluid volume is the primary determinant of sensory block extent and dura­tion during spinal anesthesia. Anesthesiology 89:24, 1998; Eisenach JC et al: Cephalad movement of morphine and fen­tanyl in humans after intrathecal injection. Anesthesiology 99:166, 2003; Higuchi H et al: Influence of lumbosacral cerebrospinal fluid density, velocity, and volume on extent and duration of plain bupivacaine spinal anesthesia. Anesthesiol­ogy 100:106, 2004; Hogan QH: Pathophysiology of peripheral nerve injury during regional anesthesia. Reg Anesth Pain Med 33:435, 2008; Kavlock R, Ting PH: Local anesthetic resistance in a pregnant patient with lumbosacral plexopathy. BMC Anesthesiol 4:1, 2004; Lee LA et al: Complications associated with eye blocks and peripheral nerve blocks: an Amer­ican Society of Anesthesiologists closed claims analysis. Reg Anesth Pain Med 33:416, 2008; Neal JM et al: ASRA Prac­tice Advisory on Neurologic Complications in Regional Anesthesia and Pain Medicine. Reg Anesth Pain Med 33:404, 2008; Neal JM: Anatomy and pathophysiology of spinal cord injury associated with regional anesthesia and pain medicine. Reg Anesth Pain Med. 33:423, 2008; Nelson KE et al: A comparison of intrathecal fentanyl and sufentanil for labor analgesia. Anesthesiology 96:1070, 2002; Pan PH et al: Incidence and characteristics of failures in obstetric neuraxial analgesia and anesthesia: a retrospective analysis of 19,259 deliveries. Int J Obstet Anesth 13:227, 2004; Parkinson D: Human spinal arachnoid septa, trabeculae, and “rogue strands”. Am J Anat 192:498, 1991; Rigler ML et al: Cauda equina syndrome after continuous spinal anesthesia. Anesth Analg 72:275, 1991; Sorenson EJ: Neurological injuries associated with regional an­esthesia. Reg Anesth Pain Med 33:442, 2008.