Abstract

A 40-year-old woman without remarkable medical history received epidural anesthesia for uterine cervix conization. Six hours after the operation, cauda equina syndrome occurred. Magnetic resonance imaging of the spine revealed epidural fluid accumulation around L5, as well as L4/5 herniated intervertebral disc found incidentally. Surgical decompression was performed with H-reflex monitoring. Epidural injection could result in cystic accumulation complicated with cauda equina syndrome.

Keywords

anesthesia, epidural; cauda equina; nerve compression syndromes; H-reflex;

1. Introduction

The rate of neurological complications in neuraxial blockades is around 1/20,000−30,000. Cauda equina syndrome (CES) is the second most frequent neurological complication of neuraxial blockades after spinal hematoma.¹ We present here a case of CES that resulted from cystic accumulation of injected epidural local anesthetics.

2. Case Report

A 40-year-old woman without remarkable medical history was scheduled for uterus cervix conization under epidural anesthesia. She had never experienced low back pain or sciatica. Physical examination revealed no spinal abnormality and there were no contraindications to epidural anesthesia.
After setting up standard monitoring, the patient was placed in the decubitus knee-chest position and premedicated with intravenous midazolam 2.5 mg.After disinfection, the skin at the area of needle entry was infiltrated with 3 mL of 2% lidocaine for needle puncture. A Perican epidural needle with Tuohy bevel (18G ˜ 3.25”; B. Braun Melsungen AG, Melsungen, Germany) was used to enter the L3/4 intervertebral space via a paramedian approach.
The loss-of-resistance to air technique was used to signal correct entry into the epidural space. The bevel of the needle was turned to face caudally and the epidural catheter was advanced downward. As the catheter was advanced some 8 cm beyond the bevel, pain and paresthesia of the legs were elicited,which were relieved after catheter withdrawal by 2 cm. The epidural catheter was fixed with 6 cm in the epidural space. The Shah test was employed to exclude intrathecal or intravascular misplacement.²Subdural injection was excluded by injection of a test dose of 2% lidocaine (2 mL). Epidural anesthesia was conducted smoothly with 15 mL of 1.5% lidocaine mixed with 2 mL of 7% sodium bicarbonate. By pinprick test, anesthetic level up to T10 was confirmed.The patient was placed in the supine lithotomy position for the 90-minute operation. The operation and anesthesia passed uneventfully. The epidural catheter was removed in the postanesthesia care unit and the patient was then returned to the ward.

However, 6 hours postoperatively, the patient complained of bilateral lower extremity weakness and voiding difficulty. Physical examination revealed bilateral muscle weakness of knee and ankle flexion.Babinski’s reflex was absent bilaterally. Perception to pinprick and cold over the perianal area and posterior aspect of both thighs and legs was found to be decreased. Under the impression of CES, magnetic resonance imaging of the spine was arranged. L4/5 herniated intervertebral disc with decreased anterior space of the cauda equina was noted (Figure 1). In addition, a space-occupying fluid lesion was found at the same level, which pushed the cauda equina against the herniated disc (Figure 2).Emer gent surgical decompression was carried out 24 hours after the onset of symptoms due to the patient’s initial hesitation.
Midazolam 2.5 mg, lidocaine 20 mg and propofol 100 mg were given intravenously for induction of general anesthesia. Intravenous succinylcholine 60 mg was used to facilitate endotracheal intubation followed by intermittent rocuronium during the operation for surgical relaxation. After patient positioning, H-reflex was recorded bilaterally from the gastrocnemius muscle following 0.5-ms stimulations of the tibial sensory fibers in the popliteal fossa (Figure 3).Anesthesia was maintained with 2% end-tidal sevoflurane air/oxygen mixture at a flow rate of 2 L/min (FiO2 = 0.5) together with intravenous fentanyl. During the critical periods of nerve decompression, intravenous bolus of rocuronium or fentanyl was avoided to minimize any pharmacologic interference of the H-reflex. During the operation, a cyst of about 1.5 cm in diameter with clear watery content was noted around the L4−5 level. The cyst ruptured during surgical manipulation and the watery content it held welled out. The H-reflex from both legs weakened markedly within a minute and took about 20 minutes to return to the previous level (Figure 3).

Download full-size image

Figure 1. T2-weighted magnetic resonance imaging shows herniated intervertebral disc and accumulated epidural fluid impinging the cauda equina.

Download full-size image

Figure 2. T2-weighted magnetic resonance imaging shows: (A) accumulated epidural fluid (up arrow) pushing the cauda equina upward against the herniated disc (down arrow); (B) normal cauda equina distribution at the same level of an age-matched female in which the nerves of the cauda equina “sinks” evenly in the spinal canal.

Download full-size image

Figure 3. (A) Baseline H-reflex. (B) H-reflex under the effect of lidocaine. (C) Postoperative H-reflex.

At the conclusion of the operation, the H-reflexamplitude had increased and the latency had shortened compared with the baseline.
The patient’s lower-extremity muscle power improved markedly after the operation. On the first postoperative day, she could walk with a cane. However, constipation and sacral numbness bothered her for the next 2 months. Voiding difficulties remained unresolved even 5 years after the operation.

3. Discussion

The incidence of CES is 2.7 per 100,000 epidural blocks. More than two-thirds of CES can result in permanent neurological deficits.¹ Favorable neurological outcome tends to occur in patients who undergo surgical decompression within 8 hours after diagnosis.³
No evidence was found comparing the incidence of fluid accumulation following epidural injection via a median or paramedian approach. However, Cardoso and Carvalho⁴ reported that the peak epidural pressure is directly related to the speed of injection, which in turn correlates well with the duration of effect but inversely correlates with the extent of epidural blockade. More rapid epidural injection could result in more limited distribution but much longer duration. For patients with spinal stenosis, slower speed of injection is reasonable to lower the risk of possible mass effect from fluid accumulation. Paresthesia during needle puncture or catheter placement and pain on injection are risk factors of neurological complications.⁵ Among those who experienced paresthesia during needle puncture or pain on injection, 62−66% had persistent paresthesia as an aftermath.^6,7 Therefore, symptoms during conduction of regional anesthesia should be regarded as a warning, which should prompt more careful neurological follow-up or even a change in the anesthesia plan.

Pre-existing spinal stenosis has been reported to be a risk factor of neurological complications after epidural anesthesia.^8,9 Although there is no previous imaging available as solid proof, the general consensus is that a pre-existing asymptomatic herniated for cauda equina. In this case, the cauda equina lay anteriorly to the accumulated epidural fluid, whereas normal cauda equina is grouped at the dependent site (Figure 2). It is conjectured that the cauda equina was pushed by the supposedly iatrogenic epidural cyst against the intruded intervertebral disc. Injection of fluid into the perineural cyst^10,11 (also known as Tarlov’s cyst), although rare, was also a possible cause of cyst formation. The discovery of the mass to be epidural clear watery accumulation excluded the possibility of hematoma or abscess.H-reflex contributed to the determination of whether the watery collection of the cyst was cerebrospinal fluid¹² or injectant.^13,14

The gastrocnemius muscle, innervated by the tibial nerve, was selected for H-reflex monitoring in our patient. Although not every single root of the cauda equina was subjected to testing, the H-reflex of the gastrocnemius muscle reflected the nerve conduction of L4−5 and S1−2.¹⁵ During the operation, the total disappearance of the H-reflex was considered to be due to the effect of lidocaine released from the ruptured cyst that flooded the adjacent nerve roots (Figure 3). Lidocaine, being an amino amide drug, undergoes enzymatic degradation primarily in the liver. Although some degradation of amide-type compounds may take place in tissues other than the liver, the pharmacologic activity of lidocaine appeared to be preserved in the cyst, which is isolated from the bloodstream to some degree.¹⁶ After flooding, the lidocaine was exposed to the circulation and its effect wore off in 20 minutes.

Intraoperative H-reflex is not thoroughly understood.There have been studies on the effects of sevoflurane and propofol on the H-reflex.¹⁷⁻²¹ Sevoflurane resulted in dose-dependent suppression of the H-reflex. Propofol resulted in H-reflex suppression only at high plasma concentrations. In practical terms, propofol does not significantly affect the H-reflex. Also, hyper- and hypoventilation could affect H-reflex amplitude while H-reflex latency could remain unaffected.²² Therefore, during the maintenance of anesthesia, concentration of volatile anesthetic and minute ventilation should be kept as stable as possible.
Paresthesia during epidural anesthesia may signal pre-existing spinal stenosis and warrants careful evaluation and reconsideration of the anesthesia plan. The H-reflex presentation indirectly showed that epidural injection might, although rare, result in cystic accumulation of injectant and nerve compression. However, the image warning of “antigravity” cauda equina due to epidural cyst compression was challenging. Early diagnosis of CES after epidural anesthesia relies greatly on a high index of suspicion.