Transient Neurologic Symptoms after Spinal Anesthesia : A Lower Incidence with Prilocaine and Bupivacaine than with Lidocaine

With institutional review board approval of the University of Basel and written informed consent, we studied 90 patients classified as American Society of Anesthesiologists physical status I or II who were scheduled for short gynecologic procedures during spinal anesthesia. The number of patients studied was based on power analysis assuming a 35% incidence of TNSs after lidocaine spinal anesthesia (as we have previously observed [5–7]) and using the ability to detect a 95% reduction in TNSs with beta set at 0.2 and alpha set at 0.05. Patients with a history of chronic back pain or preexisting neurologic abnormalities were excluded from the study.

Patients were assigned using a computer-generated randomization scheme to receive 2.5 ml 2% lidocaine in 7.5% glucose, 2% prilocaine in 7.5% glucose, and 0.5% bupivacaine in 7.5% glucose. The study solutions were provided by the hospital pharmacy of the University of Basel and were provided in blinded vials. The solutions were prepared by dissolving crystalline glucose (Siegfried AG, Zofingen, Switzerland) and either lidocaine hydrochloride (Pfannenschmidt AG, Hamburg, Germany), prilocaine hydrochloride (Synopharm AG, Barsbuettel, Germany), or bupivacaine hydrochloride (Orgamol SA, Evionnaz, Switzerland) in sterile water. Table 1shows the physicochemical characteristics of the study solutions. Glucose concentration was measured using electronic polarometry (Polartronic, Schmidt-Haensch, Berlin, Germany); osmolarity was determined by measuring depression of freezing point (Micro-Osmometer 3MO, Needham Heights, MA); pH was measured using an electronic pH meter (691 pH Meter; Metrohm AG, Herisau, Switzerland); specific gravity was measured using a digital densitometer (DMA45, A. Paar AG, Graz, Austria).

Premedication consisted of midazolam, 7.5 mg orally, approximately 60 min before surgery. Intraoperative monitoring included electrocardiography, oscillometry, and continuous pulse oximetry. After initiation of a peripheral intravenous infusion, patients were placed in the left lateral decubitus position, and the skin was prepared with a 10% solution of povidone-iodine in isopropyl alcohol (Betaseptic[registered sign]; Mundipharma, Bard Pharmaceuticals, Hamilton, Bermuda). Dural puncture was performed at the L3-L4 interspace with a 25-gauge pencil-point spinal needle (Pencan[registered sign]; B. Braun, Melsungen, Germany) using a midline approach. After free flow of cerebrospinal fluid was obtained, 2.5 ml of the study solution was injected with the bevel of the needle directed cranially; thereafter, the patients were turned supine and immediately placed in the lithotomy position.

Problems related to spinal puncture, such as multiple attempts, bleeding, or paresthesias, were noted. Segmental level of anesthesia was assessed bilaterally using cold swabs at 5-min intervals for 15 min and, thereafter, at 15-min intervals until complete recovery from sensory block. Motor block was assessed using a modified Bromage scale [8](0 = able to move hip, knee, ankle, toes; 1 = unable to move hip, able to move knee, ankle, toes; 2 = unable to move hip and knee, able to move ankle and toes; 3 = unable to move hip, knee, ankle, able to move toes; 4 = unable to move hip, knee, ankle, toes) until the Bromage scale equaled 0. Times to void and to ambulate were noted by the postanesthesia care unit nurse.

Blood pressure was measured before and then at every minute for 15 min after subarachnoid injection; thereafter, blood pressure was recorded in 3-min intervals until complete recovery from spinal anesthesia. Minimal mean arterial pressure was noted, and maximal decline in mean arterial pressure with regard to preinduction values were calculated. Hypotension was treated by intravenous administration of ephedrine if systolic arterial pressure decreased approximately 25% from baseline or if symptoms such as nausea or dizziness were reported.

On postoperative day 1, patients were interviewed to identify TNSs using a standardized symptom checklist (Figure 1). Such symptoms were defined as pain and/or dysesthesia in the legs or buttocks, starting after recovery from spinal anesthesia; localized pain at the site of puncture was, by definition, not considered a TNS. The time from regression of sensory block to S2 (as defined by recovery of normal temperature sensation at the posterior aspects of the thighs) to the onset of TNSs was noted. Symptomatic patients were asked to rate the degree of discomfort using a visual analog scale (0 = no pain, 10 = worst conceivable pain). Sensory and motor examinations were performed on all patients with TNSs and repeated on a daily basis until they were symptom-free.

Results are expressed as means +/- SD, or median and range. Continuous variables among groups were compared using analysis of variance, followed by comparison with the Student-Newman-Keuls test, if appropriate. In case of deviation from normal data distribution, analysis of variance was applied on rank transformed data (the Kruskal-Wallis test), followed by the Student-Newman-Keuls test applied on logarithmically transformed data, if appropriate. Categorical data among study groups were compared using chi squared test with Yates correction. Continuous variables among patients with and without TNSs were compared using the Student's t test or the Mann-Whitney test, as appropriate. P < 0.05 was considered significant.

Demographic data did not differ among study groups (Table 2). The relevant aspects of the anesthetic and surgical procedures and perioperative characteristics are summarized in Table 2. Satisfactory surgical anesthesia was achieved in all patients. The times for regression to S2, for complete recovery from motor block, and to void, were similar after lidocaine and prilocaine but prolonged after bupivacaine (Table 3). Complete recovery from anesthesia was documented in all patients on the evening of the operative day.

The incidence, duration, and characteristics of TNSs are provided in Table 4. Such symptoms were observed in nine patients (30%) receiving lidocaine, in one patient receiving prilocaine (3%), and in none receiving bupivacaine. The difference in the incidence of TNSs after lidocaine and prilocaine was significant (P = 0.036). The TNSs were bilateral in all patients and resolved within 48 h, except for the one patient with pain who received prilocaine whose symptoms lasted for 4 days. No patient had evidence of bowel or bladder dysfunction or abnormal reflexes. There was no difference in the incidence of localized pain at the site of puncture among the study groups (Table 4). Demographics and perioperative characteristics did not differ between patients with and without TNSs.

The results of the present study indicate that TNSs after spinal anesthesia are less likely to occur with prilocaine than with lidocaine. Further, the results confirm that sensory and motor block with prilocaine is similar to that obtained with lidocaine. Taken together, these data suggest that prilocaine might be a suitable alternative to lidocaine for short surgical procedures, particularly in the outpatient setting. However, because of the limited number of patients studied, the 95% confidence interval for TNSs with prilocaine had an upper limit of 17%. Thus larger follow-up studies are required before concluding that, with respect to TNSs, prilocaine is a practical alternative to lidocaine for spinal anesthesia.

Previous studies have shown a lower incidence of TNSs with bupivacaine compared with lidocaine. However, interpretation of previous data is limited by lack of randomization, [5,9,10]variability in the glucose or epinephrine content of the anesthetic solutions, [11]and inequality of anesthetic dosage. [5,6,9,11]In the present study, lidocaine and bupivacaine were administered in solutions of equivalent baricity, glucose content, and at a 4:1 ratio, consistent with potencies determined in most experimental studies. [12–14]Thus, in addition to confirming a low incidence of TNSs with bupivacaine, the present results indicate that differences in incidence are due, at least in part, to an intrinsic property of the anesthetic.

The cause of TNSs remains unknown. In their initial report, Schneider et al. [15]proposed that TNS represented transient neurotoxicity. Based on cadaver preparations demonstrating stretch of the lumbosacral nerve roots in the lithotomy position, the authors also hypothesized that lithotomy positioning reduced blood flow to the lumbosacral nerve roots, predisposing them to local anesthetic-induced neurotoxic effects. Data from a subsequent clinical study demonstrated that patients receiving lidocaine spinal anesthesia for procedures in the lithotomy position appear to be at higher risk for the development of TNSs, [10]which is consistent with this cause but by no means conclusive. Further, in vitro and in vivo experimental data suggest that the neurotoxic potential of lidocaine may exceed that of bupivacaine. [13,16,17]However, we cannot conclude from this that TNSs and neurotoxic injury share a common mechanism.

Myofascial pain syndrome or posterior spinal segment dysfunction may also cause a radicular type of leg pain [18]and recently were proposed as alternative causes of TNSs. [19]However, ascribing these symptoms to a musculoskeletal origin is difficult because of the disproportionally high incidence of TNSs observed with lidocaine in the present (Table 4) and previous studies. [5,6,9,11]One possible explanation might be that the local anesthetics differ with respect to an intervening factor; for example, spinal lidocaine might produce a more profound motor block with the resultant differences in TNSs due to greater musculoskeletal relaxation. However, we did not observe a difference in motor blockade among study groups (Table 3). This finding is also consistent with a previous study comparing hyperbaric solutions of lidocaine and bupivacaine. [20] 

Although the cause of TNSs remains to be determined, several studies have succeeded in eliminating factors such as hyperosmolarity, concentration, or addition of epinephrine or glucose that do not promote their occurrence after lidocaine spinal anesthesia. [6,7,11]Conversely, available data identify patient positioning as a major contributing factor to TNSs after lidocaine spinal anesthesia. [10,11]This may account for some of the variability in incidences reported in previous studies. [5,9–11]However, that TNSs were observed in one patient receiving prilocaine, together with reports on their occurrence after bupivacaine, tetracaine, and mepivacaine spinal anesthesia, emphasize that this complication is not solely associated with the use of lidocaine. [21–25] 

In conclusion, in this study prilocaine was associated with a significantly lower incidence of TNSs compared with lidocaine. Furthermore, the duration of action of prilocaine was comparable to that of lidocaine. Taken together, these results suggest that it might be appropriate to use prilocaine in place of lidocaine for spinal anesthesia. However, because of the limited number of patients, large-scale studies are required before concluding that the incidence of TNSs with prilocaine is low enough for it to be a useful alternative to lidocaine. In addition, although it is possible that these transient neurologic symptoms represent the lower end of a spectrum of toxicity, a relation between these symptoms and more serious neurologic deficits occurring at higher doses [26]remains speculative. Thus the present study cannot address the relative safety of these anesthetics with respect to their potential for clinical injury.

The authors thank Winifred von Ehrenburg, M.A., and Joan Etlinger for editorial advice.