- Email: [email protected]
Effectiveness of cervical sympathetic ganglia block on regeneration of the trigeminal nerve following transection in rats
Effectiveness of Cervical Sympathetic Ganglia Block on Regeneration of the Trigeminal Nerve Following Transection in Rats Naotoshi Hanamatsu, D.D.S., Ph.D., Mikiko Yamashiro, D.D.S., Ph.D., Masahito Sumitomo, D.D.S., Ph.D., and Hideki Furuya, D.D.S., Ph.D. Background and Objectives: Stellate ganglion block (SGB) is one treatment option for human trigeminal nerve injury. The aim of this study was to evaluate the effectiveness of cervical sympathetic ganglia blocks (SB) by comparing the recovery of severed nerves in 2 rat models, treated or not treated by SB. Methods: The infraorbital nerves (ION) were cut in 108 rats. Fifty-four of them were treated daily by SB for 30 days (SB group). The remainder were left untreated (Control group). The stages of recovery were evaluated neurophysiologically by measuring somatosensory evoked potentials (SEPs) and were histologically analyzed via microscopic observation. Results: The neurophysiologic evaluation showed that SEP amplitude was detected 1 month after cutting the ION in the SB group, but not in the Control group. The average recovery after 8 months was almost 100% in the SB group and about 70% in the Control group. The histologic evaluation showed no significant difference in the number of myelinated nerve fibers per unit area between the 2 groups. However, in the SB group, the mean diameter and distribution of diameters of the myelinated fibers were greater, and myelinated fibers of large diameter were observed at an early stage. Conclusions: The findings suggest that cervical sympathetic nerve block may accelerate the recovery and regeneration of severed ION. The clinical correlation in patients with peripheral trigeminal paralysis remains to be established. Reg Anesth Pain Med 2002;27:268-276. Key Words:
Experimental trigeminal transection, Cervical sympathetic ganglia block, Nerve regeneration.
W
e have occasionally encountered patients with trigeminal paralysis due to lesions of the peripheral branches. The trigeminal nerve is the major sensory nerve affecting dental procedures. It may be anesthetized by an unintentional inflow of root canal filling material into the inferior alveolar canal, an inferior alveolar nerve block, tooth extraction, or apicoectomy. Surgical manipulation during dental implantation has also been regarded as a cause of the paralysis.1,2 Stellate ganglion block (SGB) is used empirically to treat peripheral trigeminal paralysis. There are few evidence-based data for treatment for trigeminal paralysis. Therefore, in these experiments, we examine the effect of sym-
From the Department of Anesthesiology, School of Dentistry at Tokyo, The Nippon Dental University, Tokyo, Japan. Accepted for publication November 3, 2001. Reprint requests: Naotoshi Hanamatsu, D.D.S., Ph.D., Department of Anesthesiology, The Nippon Dental University, School of Dentistry at Tokyo, 2-3-16, Fujimi, Chiyoda-ku, Tokyo 1020051, Japan. © 2002 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/02/2703-0021$35.00/0 doi:10.1053/rapm.2002.31206
268
pathetic ganglion block on nerve regeneration in an experimental model. We performed cervical sympathetic ganglia blocks (SB) on rats with experimentally produced paralysis of the infraorbital nerves (ION). The effects of sympathetic ganglion block on the recovery of the cut nerves were examined neurophysiologically and histologically. Since the rat does not possess ganglia corresponding to human stellate ganglia, the middle and inferior cervical sympathetic ganglia were selected as targets for the block. Blocking these ganglia produces effects similar to the SGB in humans, such as increasing the relative surface blood flow and temperature, and producing a Horner-like syndrome. These effects are reversed as the effect of a local anesthetic faded away. In this study, we regarded the middle and inferior cervical ganglia block as a substitute for the SGB in humans.3,4
Methods All experimental procedures and protocols were reviewed and approved by the Animal Care and Use
Regional Anesthesia and Pain Medicine, Vol 27, No 3 (May–June), 2002: pp 268 –276
Sympathetic Block for Nerve Regeneration
Committee of the Nippon Dental University, School of Dentistry at Tokyo. Preparation of the Rats With Cut ION We cut the ION of 108 rats of the Wistar strain, 7 to 8 weeks old and weighing 200 to 250 g. The animals were housed at a constant temperature of 24°C ⫾ 1°C for about a week before the experiments were started. Anesthesia was induced with sevoflurane (Sevofrane, Maruishi Pharmaceutical, Osaka, Japan), and 15 to 17 mg of pentobarbital sodium (Nembutal, Dai-Nippon Pharmaceutical, Osaka, Japan) was administered by an intraperitoneal injection. Spontaneous respiration was preserved. We incised vertically the skin of the left superior labial region, ablated the soft tissue, and exposed the infraorbital nerve. We completely transected the ION using a sharp pointed knife at the midpoint between the infraorbital foramen and the intersection with the facial nerve. The cut ends of individual fasciculi were reopposed using 7.0 nylon thread under a stereomicroscope (1 suture). We closely approximated the skin using a 14-mm metal clip (Dainer Automatic Seamer, Du¨sseldorf, Germany) (hereinafter, the left side of the superior labial region is referred to as the “Cut Side” and the right side as the “Normal Side”). We divided the rats into 1 group of 54 that received cervical sympathetic ganglia blocks (SB group) and another group of 54 that was untreated (Control group). We further subdivided each group into 9 subgroups of 6 each according to the length of time of observation: 1, 2, 3, 4, 5, 6, 8, 10, and 12 months. Implementation of Cervical SB After cutting the infraorbital nerve, both groups of rats were given free access to solid food (MF; Oriental Yeast Co, Tokyo, Japan) and water. During this period, the SB group underwent SB with 0.1 mL of 1% mepivacaine hydrochloride (1% Carbocaine injection, Astra Japan, Osaka, Japan) on the Cut Side once a day, under general anesthesia with sevoflurane, for 30 days starting on the day following nerve division. We treated the Control group in the same manner, but with 0.1 mL of saline. The injection was given from the dorsal side to the left second vertebra. We confirmed the efficacy of SB based on the following criteria. The onset of signs similar to Horner’s syndrome (blepharoptosis, enophthalmos) appeared immediately after the emergence from anesthesia with sevoflurane, in 1 to 2 minutes after the block and an increase in the body surface temperature of the area innervated by the cervical sympathetic nerves on the blocked side.
•
Hanamatsu et al.
269
Physiologic Evaluation by Somatosensory Evoked Potentials We measured somatosensory evoked potentials (SEPs) of the SB and Control groups during an established observation period. We induced anesthesia with sevoflurane and secured the airway using the Teflon outer tube of an intravenous catheter (Quick-Cath; 18 gauge, Baxter, Tokyo, Japan) as an endotracheal tube. An intracoelic injection of 2 mg/kg of pancuronium bromide (Mioblok; Sankyo, Tokyo, Japan) was administered, and ventilation was controlled with pure oxygen using a small animal respirator (SN-480-5; Shinano Manufacturing, Tokyo, Japan). The tidal volume was 10 to 15 mL, and the ventilatory frequency was 55 to 60 breaths per minute. The body temperature was monitored rectally and maintained at 36°C. Systolic and diastolic blood pressure and heart rate were monitored by a noninvasive sphygmomanometer (BP-98; Softron, Tokyo, Japan), according to the tail-cuff method. SEPs were recorded 5 and 10 minutes after the beginning of controlled ventilation on both the Cut and Normal Sides. We measured SEPs using an evoked potential analyzer (Neuropack 2; MEB-7102, Nihon Kohden, Tokyo, Japan) according to the method of Toda et al.5 Stainless steel, stimulatory needle electrodes were placed in bipolar positions in the superior labial region of the SEP elicited side. A reference electrode was placed ipsilaterally, and a recording electrode was placed contralaterally under the periosteum of the temporal calvarium, which corresponds to the somatosensory area of the region governed by the trigeminal nerve. Electric stimuli with an intensity of 1.4 mA and a single square wave pulse of 0.2-ms duration were repeatedly delivered 20 times at 2 Hz.5 We identified the first clearly positive and negative SEP waveforms as P1 and N1 waves, respectively, and measured the peak-to-peak amplitude between P1 and N1. We adopted the mean value of the 2 measurements as an amplitude value for statistics. The amplitude value of the Cut Side was converted into a percentage in relation to that of the Normal Side. The ratio was used as the metric for statistical analysis. Histologic Evaluation After SEP measurement, we removed the ION to compare the histologic changes. We prefixed them with 2.5% glutaraldehyde in 0.05 mol/L cacodylic acid buffer, pH 7.4, postfixed them with 1.0% osmium tetroxide, dehydrated with ethyl alcohol, exposed to propylene oxide, and embedded in epoxy resins. The embedded samples were sliced at a position 2 mm distal to the severed site to form cross-
270
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
sections 1 m thick, then stained with toluidine blue for optical microscopy. Ultrathin sections were prepared from slices taken from the same region, and double stained with uranyl acetate and lead acetate for transmission electron microscopy (JEM2000EX II; JEOL, Tokyo, Japan). Histologic Evaluation of the 1- and 8-Month Subgroups. We examined 24 cross-sections (1 cross section from each animal in each subgroup of the 2 groups) under an optical microscope. Number of Regenerating Myelinated Nerve Fibers per Unit Area. The number of regenerated myelinated nerve fibers was counted, and the area of transection of the fasciculi in the sample was measured under an optical microscope. We compared the number of regenerated myelinated nerve fibers/mm2 between the 2 groups. Mean Diameter and Distribution of the Diameters of Regenerated Myelinated Nerve Fibers. A fasciculus most representative of the average was selected from each sample for evaluation. We measured the diameter of the regenerated myelinated nerve fibers in each selected fasciculus to obtain the average diameter and distribution of diameters of the myelinated nerve fibers and compared the values. Histologic Structure of the ION in 1- and 8-Month Subgroups. After SEP measurement, the ION on the Normal and Cut Sides were removed for histologic evaluation. Data Analysis Differences in mean values were examined using an unpaired t test. The level of significance was set at .05. A probability of P ⬍ .05 was considered significant.
Results Electrophysiologic Recovery of the Cut ION Measured by SEPs The SEP amplitude increased within 1 month after nerve cutting in the SB group, but not in the Control group. After 8 months, recovery of the SEP amplitude in the SB group had reached almost 100%, whereas that of the Control group was about 70% at 8 months and did not reach 100% until 12 months had elapsed (Fig 1). The recovery rate of the SB subgroups was consistently higher, and the standard deviation was consistently smaller than those of the Control subgroups. Differences in the recovery rates after 1, 2, 4, 5, 6, and 8 months between the SB and Control groups were significant.
Fig 1. Changes in SEP amplitude of severed infraorbital nerves in the SB and Control groups. Recovery rate was expressed as a percentage of SEP amplitude recorded from the Normal Side. In the SB group, SEP amplitude was measurable within 1 month after nerve cutting. After 8 months, average recovery rate in the SB group reached almost 100%, whereas that in the Control group reached 70%. Comparison of the recovery rates between the 2 groups showed significant differences after 1, 2, 4, 5, 6, and 8 months.
Histologic Observation by Optical and Electron Microscopy Histologic Evaluation of Infraorbital Nerves 1 and 8 Months After Nerve Cutting. Number of Regenerated Myelinated Nerve Fibers/Unit Area. In the 1-month subgroups, the number of regenerated myelinated nerve fibers was 6,641 ⫾ 2,200.3 in the SB group and 6,367 ⫾ 1,957.3 in the Control group. In the 8-month subgroups, the number was 10,443 ⫾ 3,210.7 in the SB group and 10,715 ⫾ 4,125.7 in the Control group. The number of myelinated nerve fibers on the Normal Side was 10,597 ⫾ 1,117.4. There were significant differences in the number of regenerated fibers between the 1- and 8-month subgroups in both groups. However, there were no significant differences in the number of regenerated fibers between the SB and Control groups either in the 1-month or 8-month subgroups. Mean Diameter and Distribution of the Diameters of Regenerated Myelinated Nerve Fibers. The average diameter was 10 ⫾ 4.53 m on the Normal Side (Fig 2). In the 1-month subgroups, the average was 6 ⫾ 2.75 m in the SB group and 5 ⫾ 2.67 m in the Control group. In the 8-month subgroups, the average was 10 ⫾ 4.83 in the SB group and 9 ⫾ 3.85 m in the Control group. The average diameter of the regenerated myelinated nerve fibers was greater in the SB group in both the 1- and 8-month subgroups though this difference was not significant. The histogram of the myelinated fibers on the Nor-
Sympathetic Block for Nerve Regeneration
Fig 2. Compared recovery rate of infraorbital nerves between the SB and Control groups 1 and 8 months after nerve cutting: average diameter of regenerated myelinated nerve fibers and distribution of the diameters. Distribution of the diameters, average size, and standard deviation of myelinated nerve fibers on the Normal Side and of regenerated myelinated nerve fibers of 1- and 8-month subgroups of the 2 groups. The average diameter of regenerated myelinated nerve fibers in the SB subgroups was greater (not significantly different). Histogram of myelinated fibers on the Normal Side displayed rather double-phased distribution of larger and smaller diameter fibers. Histogram corresponding to 1-month subgroups of the 2 groups showed single-phased distribution with peaks of smaller diameter fibers less than 8 m, and a few had larger diameters. Histogram on the 8-month SB subgroup showed double-phased distribution of smaller and larger diameter fibers.
mal Side displays a double-phased distribution, with the first peak being in the smaller diameter range of less than 8 m and the second being in the larger diameter range of 8 m or greater. Changes in Histologic Structure of the ION 1 and 8 Months After Nerve Cutting 1 Month After Nerve Cutting. Optical microscopy revealed that the myelinated nerve fibers were scattered and scarcer 1 month after nerve
•
Hanamatsu et al.
271
cutting in the SB and Control groups compared with those on the Normal Side (Figs 3 and 4). Immature myelinated nerve fibers that were still undergoing regeneration were stained with toluidine blue. Not all of the regenerated myelinated fibers were inside the fasciculi. Some were located in the connective tissues of the epineurium. Swelling, formation of vacuoles, and shriveled axons in some myelinated nerve fibers were observed even at this stage. The distribution of regenerated myelinated fibers was not equal in all fasciculi. In the regenerating myelinated fibers, some axons, which were surrounded by projections emanating from what seemed to be fibroblasts, formed small fasciculus-like structures (Compartmentation). Some fasciculi contained a few scattered regenerated fibers, while in others, a large number of regenerated fibers was growing uniformly. The vessels of the epineurium, perineurium, and endoneurium seemed to have reassumed their shape and regained their function. Electron microscopy showed that numerous myelinated nerve fibers were in the process of regeneration in both groups. At the same time, a relatively large number of degenerated nerve fibers was evident. We found that some axons formed similar structures in some myelinated fibers that began regeneration outside the fasciculi from which they had originated. In the degenerated nerve fibers, we found the following: gaps between axons and myelin sheaths were formed owing to axon shrinking; axons had totally disappeared and as a result, the surrounding myelin sheaths turned into wavy myelin balls; cells that appeared to be macrophages ingested and phagocytosed the degenerated nerve fibers. Eight Months After Nerve Cutting Judging from the number and size of myelinated nerve fibers and the thickness of myelin sheaths, regeneration had progressed considerably in the 8-month subgroups compared with the findings from the 1-month subgroups (Figs 5 and 6). Under optical microscopy, the myelinated nerve fibers that had almost completed regeneration occupied most of the fasciculi in both groups. In some fasciculi, the myelinated nerve fibers were not mature; myelin sheaths were irregularly shaped and seemed to be in the premature stage of regeneration. Several axons that had regenerated outside the original fasciculi were occasionally surrounded by a structure like the perineuria to form small fasciculi. Electron microscopy showed that a large number of myelinated nerve fibers had almost completed regeneration in the SB group (Fig 7). This finding coincided with the findings of optical microscopy. However, a
272
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
Fig 3. Optical micrograph of infraorbital nerves 1 month after severance in the SB group. Smaller diameter myelinated nerve fibers, seemingly undergoing regeneration, have densely stained myelin sheath (toluidine blue; ⫻60).
large number of insufficiently myelinated nerve fibers were still in the process of regeneration in the Control group (Fig 8). No nerve fibers had degenerated in either group at this stage.
Discussion The function of the sensory afferent pathway of the trigeminal nerve can be objectively measured by stimulating the trigeminal nerve to assess the
Fig 4. Optical micrograph of infraorbital nerves 1 month after severance in the Control group. Regeneration rate of each fasciculus in the same nerve trunk differed (toluidine blue; ⫻60).
functional recovery of the severed nerve. We used SEPs to objectively and neurophysiologically observe the functional recovery of severed nerve. SEPs are elicited from a region on the skull surface corresponding to the somatosensory field by stimulating the sensory nerve at the periphery and the sum of conduction or the compound potentials of a group of nerve fibers, thus reflecting the function of stimulus conduction.6 The P1-N1 amplitude of SEPs
Sympathetic Block for Nerve Regeneration
•
Hanamatsu et al.
273
Fig 5. Optical micrograph of infraorbital nerve from the SB group 8 months after nerve cutting. Many myelinated fasciculi with regeneration almost completed were seen (toluidine blue; ⫻60).
is dependent on the intensity of the stimulus and is often adopted as an index of the sum of sensation to stimulus in humans or animals.5 In the SB group, the SEP amplitude recovered significantly earlier in comparison to the Control group (Fig 1). The variability of the SEP amplitude tended to be smaller in the SB group than the Control group. We compared the degree of regen-
eration 1 and 8 months after nerve cutting by measuring myelinated nerve fibers/unit area, but differences were not significant in any subgroup in the 2 groups. We found that the number of regenerated myelinated fibers of the 1-month subgroups determined by electron microscopy was slightly different from that determined by optical microscopy. We believe that optical microscopy failed to distinguish
Fig 6. Optical micrograph of infraorbital nerve from the Control group 8 months after nerve cutting. Some fasciculi contain numerous myelinated nerve fibers still regenerating. Myelin sheaths are irregularly shaped and seem to be in the premature stage of regeneration process (toluidine blue; ⫻80).
274
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
Fig 7. Electron micrograph of infraorbital nerve from the SB group 8 months after nerve cutting. Myelinated nerve fibers show the regeneration process is almost complete.
between regenerated fibers and myelin sheaths that had escaped phagocytosis by macrophages, and thus included these sheaths in counting of the regenerated fibers and thereby overestimated the number of myelinated fibers. In the 8-month subgroups, we assumed that optical microscopic measurements were not overestimated, because electron microscopy did not reveal anything resembling remnants of myelin sheaths. The number of myelinated nerve fibers/unit area between the Cut and Normal Sides did not significantly differ in the
8-month subgroups of both groups. We concluded that regeneration in terms of number of fibers was almost complete 8 months after nerve cutting. We measured the number of myelinated nerve fibers, not of axons, as a tool with which to assess the degree of nerve regeneration to avoid the influence of excessive sprouting resulting from the disturbance of axonal growth. The results showed that the distribution of myelinated nerve fibers on the Normal Side was 2-phased; smaller and larger diameter fibers of up to 8 m and over 8 m, respec-
Sympathetic Block for Nerve Regeneration
•
Hanamatsu et al.
275
Fig 8. Electron micrograph of infraorbital nerve from the Control group 8 months after nerve cutting. Many myelinated nerve fibers are still immature.
tively (Fig 2). In both 1-month subgroups, the distribution was single-phased, with the myelinated fibers with smaller diameter being more prevalent than in the Normal Side and only a few fibers of the larger diameter. We attributed these findings to immature myelinization and assumed that the myelin sheath of the smaller fibers would thicken over time and that their diameters would accordingly increase. We also presumed that nerve fibers that had not been myelinated sufficiently (resulting from delayed regeneration) to be recognized under
optical microscopy would gradually join the larger diameter fibers. We considered that the SB 8-month subgroup had attained restoration and regeneration that was comparable to those of the Normal Side, although this subgroup contained fewer nerve fibers between 11- and 13-m diameter compared with the Normal Side (Fig 5). Therefore, we concluded that the myelinated nerve fibers reached maturity at 8 months in the SB group. The Control 8-month subgroup seemd to lack a group of large nerve fibers, especially those larger than
276
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
15-m diameter in comparison to the corresponding SB subgroup of the Normal Side. Consequently, it was assumed that the SB group was still undergoing regeneration even after 8 months (Fig 6). Compared with the electrophysiologic degree of recovery measured by SEPs, fibers with the highest diameter among all the regenerated myelinated fibers played a crucial role in the high recovery rate. Wallerian degeneration occurs at the distal portion of the severed region,7 which we observed even 1 month after the nerve transection in the present study. Blood flow in the endoneurium does not necessarily flow in a regular direction owing to widespread anastomosis and changes direction as soon as pressure or a disturbance of blood flow arises. The microcirculation of a fragment of the rabbit tibial nerve separated by cutting at the central and peripheral sides remains unaffected if the blood vessels from outside are retained.8 That report suggests that the vascular network has considerable reserve capacity. The intraneural vascular system contains adrenergic nerve endings regulated by sympathetic nerves, and postsynaptic conduction causes vasoconstriction via alpha receptors.9 The peripheral vessels constantly receive impulses from vasoconstrictors (resting tone). Therefore, interruption of sympathetic activity inevitably causes vasodilation. A cervical SB temporarily interrupts outflow from the cervical sympathetic nerves and, consequently, produces vasodilation of peripheral vessels governed by the cervical sympathetic nerve. We believe that the effects of the SB are further extended from nutrient vessels around the injured area to the vascular network of the nerve fragment both central and peripheral to the severed region. Our results using SEPs and the distribution of fiber
diameters suggest that the SB promotes myelinization following axonal growth and the maturation of myelin sheaths. In conclusion, we demonstrated that SB promoted the recovery and regeneration of severed ION. The clinical relevance of these findings in humans remains indeterminate.
Acknowledgment The authors thank Prof. Shigeo Aiyama for reviewing the manuscript.
References 1. Hegtvedt AK. Evaluation and management of orofacial sensory nerve injuries. Ohio Dent J 1992;66:16-17. 2. Shankland WE. Anatomical considerations for mandibular implant surgery: Two common post-operative problems. Dent Today 1991;10:48-49. 3. Moore DC. Stellate ganglion block. Springfield, IL: Thomas; 1954. 4. Eriksson E. Atlas der Lokalanasthesia. New York, NY: Thieme; 1970. 5. Toda K, Ichioka M, Suda H, Iriki A. Effects of electroacupuncture on the somatosensory evoked response in rat. Exp Neurol 1979;63:652-658. 6. Liveson JA, Spielholz NI. Interpretation of electrodiagnostic data. In: Liveson JA, Spielholz NI, eds. Peripheral Neurology: Case Studies in Electrodiagnosis. Philadelphia, PA: Davis; 1979:3-8. 7. Gilliatt RW, Hjorth RJ. Nerve conduction during Wallerian degeneration in the baboon. J Neurol Neurosurg Psychiatry 1972;35:335-341. 8. Lundborg G. Structure and function of the intraneural microvessels as related to trauma, edema formation and nerve function. J Bone Joint Surg 1975;57A:938948. 9. Lundborg G. Nerve Injury and Repair (Japanese edition). Ikuta Y, Ochi M, Japanese eds. Tokyo, Japan: Hirokawa; 1991:4-57.
Experimental trigeminal transection, Cervical sympathetic ganglia block, Nerve regeneration.
W
e have occasionally encountered patients with trigeminal paralysis due to lesions of the peripheral branches. The trigeminal nerve is the major sensory nerve affecting dental procedures. It may be anesthetized by an unintentional inflow of root canal filling material into the inferior alveolar canal, an inferior alveolar nerve block, tooth extraction, or apicoectomy. Surgical manipulation during dental implantation has also been regarded as a cause of the paralysis.1,2 Stellate ganglion block (SGB) is used empirically to treat peripheral trigeminal paralysis. There are few evidence-based data for treatment for trigeminal paralysis. Therefore, in these experiments, we examine the effect of sym-
From the Department of Anesthesiology, School of Dentistry at Tokyo, The Nippon Dental University, Tokyo, Japan. Accepted for publication November 3, 2001. Reprint requests: Naotoshi Hanamatsu, D.D.S., Ph.D., Department of Anesthesiology, The Nippon Dental University, School of Dentistry at Tokyo, 2-3-16, Fujimi, Chiyoda-ku, Tokyo 1020051, Japan. © 2002 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/02/2703-0021$35.00/0 doi:10.1053/rapm.2002.31206
268
pathetic ganglion block on nerve regeneration in an experimental model. We performed cervical sympathetic ganglia blocks (SB) on rats with experimentally produced paralysis of the infraorbital nerves (ION). The effects of sympathetic ganglion block on the recovery of the cut nerves were examined neurophysiologically and histologically. Since the rat does not possess ganglia corresponding to human stellate ganglia, the middle and inferior cervical sympathetic ganglia were selected as targets for the block. Blocking these ganglia produces effects similar to the SGB in humans, such as increasing the relative surface blood flow and temperature, and producing a Horner-like syndrome. These effects are reversed as the effect of a local anesthetic faded away. In this study, we regarded the middle and inferior cervical ganglia block as a substitute for the SGB in humans.3,4
Methods All experimental procedures and protocols were reviewed and approved by the Animal Care and Use
Regional Anesthesia and Pain Medicine, Vol 27, No 3 (May–June), 2002: pp 268 –276
Sympathetic Block for Nerve Regeneration
Committee of the Nippon Dental University, School of Dentistry at Tokyo. Preparation of the Rats With Cut ION We cut the ION of 108 rats of the Wistar strain, 7 to 8 weeks old and weighing 200 to 250 g. The animals were housed at a constant temperature of 24°C ⫾ 1°C for about a week before the experiments were started. Anesthesia was induced with sevoflurane (Sevofrane, Maruishi Pharmaceutical, Osaka, Japan), and 15 to 17 mg of pentobarbital sodium (Nembutal, Dai-Nippon Pharmaceutical, Osaka, Japan) was administered by an intraperitoneal injection. Spontaneous respiration was preserved. We incised vertically the skin of the left superior labial region, ablated the soft tissue, and exposed the infraorbital nerve. We completely transected the ION using a sharp pointed knife at the midpoint between the infraorbital foramen and the intersection with the facial nerve. The cut ends of individual fasciculi were reopposed using 7.0 nylon thread under a stereomicroscope (1 suture). We closely approximated the skin using a 14-mm metal clip (Dainer Automatic Seamer, Du¨sseldorf, Germany) (hereinafter, the left side of the superior labial region is referred to as the “Cut Side” and the right side as the “Normal Side”). We divided the rats into 1 group of 54 that received cervical sympathetic ganglia blocks (SB group) and another group of 54 that was untreated (Control group). We further subdivided each group into 9 subgroups of 6 each according to the length of time of observation: 1, 2, 3, 4, 5, 6, 8, 10, and 12 months. Implementation of Cervical SB After cutting the infraorbital nerve, both groups of rats were given free access to solid food (MF; Oriental Yeast Co, Tokyo, Japan) and water. During this period, the SB group underwent SB with 0.1 mL of 1% mepivacaine hydrochloride (1% Carbocaine injection, Astra Japan, Osaka, Japan) on the Cut Side once a day, under general anesthesia with sevoflurane, for 30 days starting on the day following nerve division. We treated the Control group in the same manner, but with 0.1 mL of saline. The injection was given from the dorsal side to the left second vertebra. We confirmed the efficacy of SB based on the following criteria. The onset of signs similar to Horner’s syndrome (blepharoptosis, enophthalmos) appeared immediately after the emergence from anesthesia with sevoflurane, in 1 to 2 minutes after the block and an increase in the body surface temperature of the area innervated by the cervical sympathetic nerves on the blocked side.
•
Hanamatsu et al.
269
Physiologic Evaluation by Somatosensory Evoked Potentials We measured somatosensory evoked potentials (SEPs) of the SB and Control groups during an established observation period. We induced anesthesia with sevoflurane and secured the airway using the Teflon outer tube of an intravenous catheter (Quick-Cath; 18 gauge, Baxter, Tokyo, Japan) as an endotracheal tube. An intracoelic injection of 2 mg/kg of pancuronium bromide (Mioblok; Sankyo, Tokyo, Japan) was administered, and ventilation was controlled with pure oxygen using a small animal respirator (SN-480-5; Shinano Manufacturing, Tokyo, Japan). The tidal volume was 10 to 15 mL, and the ventilatory frequency was 55 to 60 breaths per minute. The body temperature was monitored rectally and maintained at 36°C. Systolic and diastolic blood pressure and heart rate were monitored by a noninvasive sphygmomanometer (BP-98; Softron, Tokyo, Japan), according to the tail-cuff method. SEPs were recorded 5 and 10 minutes after the beginning of controlled ventilation on both the Cut and Normal Sides. We measured SEPs using an evoked potential analyzer (Neuropack 2; MEB-7102, Nihon Kohden, Tokyo, Japan) according to the method of Toda et al.5 Stainless steel, stimulatory needle electrodes were placed in bipolar positions in the superior labial region of the SEP elicited side. A reference electrode was placed ipsilaterally, and a recording electrode was placed contralaterally under the periosteum of the temporal calvarium, which corresponds to the somatosensory area of the region governed by the trigeminal nerve. Electric stimuli with an intensity of 1.4 mA and a single square wave pulse of 0.2-ms duration were repeatedly delivered 20 times at 2 Hz.5 We identified the first clearly positive and negative SEP waveforms as P1 and N1 waves, respectively, and measured the peak-to-peak amplitude between P1 and N1. We adopted the mean value of the 2 measurements as an amplitude value for statistics. The amplitude value of the Cut Side was converted into a percentage in relation to that of the Normal Side. The ratio was used as the metric for statistical analysis. Histologic Evaluation After SEP measurement, we removed the ION to compare the histologic changes. We prefixed them with 2.5% glutaraldehyde in 0.05 mol/L cacodylic acid buffer, pH 7.4, postfixed them with 1.0% osmium tetroxide, dehydrated with ethyl alcohol, exposed to propylene oxide, and embedded in epoxy resins. The embedded samples were sliced at a position 2 mm distal to the severed site to form cross-
270
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
sections 1 m thick, then stained with toluidine blue for optical microscopy. Ultrathin sections were prepared from slices taken from the same region, and double stained with uranyl acetate and lead acetate for transmission electron microscopy (JEM2000EX II; JEOL, Tokyo, Japan). Histologic Evaluation of the 1- and 8-Month Subgroups. We examined 24 cross-sections (1 cross section from each animal in each subgroup of the 2 groups) under an optical microscope. Number of Regenerating Myelinated Nerve Fibers per Unit Area. The number of regenerated myelinated nerve fibers was counted, and the area of transection of the fasciculi in the sample was measured under an optical microscope. We compared the number of regenerated myelinated nerve fibers/mm2 between the 2 groups. Mean Diameter and Distribution of the Diameters of Regenerated Myelinated Nerve Fibers. A fasciculus most representative of the average was selected from each sample for evaluation. We measured the diameter of the regenerated myelinated nerve fibers in each selected fasciculus to obtain the average diameter and distribution of diameters of the myelinated nerve fibers and compared the values. Histologic Structure of the ION in 1- and 8-Month Subgroups. After SEP measurement, the ION on the Normal and Cut Sides were removed for histologic evaluation. Data Analysis Differences in mean values were examined using an unpaired t test. The level of significance was set at .05. A probability of P ⬍ .05 was considered significant.
Results Electrophysiologic Recovery of the Cut ION Measured by SEPs The SEP amplitude increased within 1 month after nerve cutting in the SB group, but not in the Control group. After 8 months, recovery of the SEP amplitude in the SB group had reached almost 100%, whereas that of the Control group was about 70% at 8 months and did not reach 100% until 12 months had elapsed (Fig 1). The recovery rate of the SB subgroups was consistently higher, and the standard deviation was consistently smaller than those of the Control subgroups. Differences in the recovery rates after 1, 2, 4, 5, 6, and 8 months between the SB and Control groups were significant.
Fig 1. Changes in SEP amplitude of severed infraorbital nerves in the SB and Control groups. Recovery rate was expressed as a percentage of SEP amplitude recorded from the Normal Side. In the SB group, SEP amplitude was measurable within 1 month after nerve cutting. After 8 months, average recovery rate in the SB group reached almost 100%, whereas that in the Control group reached 70%. Comparison of the recovery rates between the 2 groups showed significant differences after 1, 2, 4, 5, 6, and 8 months.
Histologic Observation by Optical and Electron Microscopy Histologic Evaluation of Infraorbital Nerves 1 and 8 Months After Nerve Cutting. Number of Regenerated Myelinated Nerve Fibers/Unit Area. In the 1-month subgroups, the number of regenerated myelinated nerve fibers was 6,641 ⫾ 2,200.3 in the SB group and 6,367 ⫾ 1,957.3 in the Control group. In the 8-month subgroups, the number was 10,443 ⫾ 3,210.7 in the SB group and 10,715 ⫾ 4,125.7 in the Control group. The number of myelinated nerve fibers on the Normal Side was 10,597 ⫾ 1,117.4. There were significant differences in the number of regenerated fibers between the 1- and 8-month subgroups in both groups. However, there were no significant differences in the number of regenerated fibers between the SB and Control groups either in the 1-month or 8-month subgroups. Mean Diameter and Distribution of the Diameters of Regenerated Myelinated Nerve Fibers. The average diameter was 10 ⫾ 4.53 m on the Normal Side (Fig 2). In the 1-month subgroups, the average was 6 ⫾ 2.75 m in the SB group and 5 ⫾ 2.67 m in the Control group. In the 8-month subgroups, the average was 10 ⫾ 4.83 in the SB group and 9 ⫾ 3.85 m in the Control group. The average diameter of the regenerated myelinated nerve fibers was greater in the SB group in both the 1- and 8-month subgroups though this difference was not significant. The histogram of the myelinated fibers on the Nor-
Sympathetic Block for Nerve Regeneration
Fig 2. Compared recovery rate of infraorbital nerves between the SB and Control groups 1 and 8 months after nerve cutting: average diameter of regenerated myelinated nerve fibers and distribution of the diameters. Distribution of the diameters, average size, and standard deviation of myelinated nerve fibers on the Normal Side and of regenerated myelinated nerve fibers of 1- and 8-month subgroups of the 2 groups. The average diameter of regenerated myelinated nerve fibers in the SB subgroups was greater (not significantly different). Histogram of myelinated fibers on the Normal Side displayed rather double-phased distribution of larger and smaller diameter fibers. Histogram corresponding to 1-month subgroups of the 2 groups showed single-phased distribution with peaks of smaller diameter fibers less than 8 m, and a few had larger diameters. Histogram on the 8-month SB subgroup showed double-phased distribution of smaller and larger diameter fibers.
mal Side displays a double-phased distribution, with the first peak being in the smaller diameter range of less than 8 m and the second being in the larger diameter range of 8 m or greater. Changes in Histologic Structure of the ION 1 and 8 Months After Nerve Cutting 1 Month After Nerve Cutting. Optical microscopy revealed that the myelinated nerve fibers were scattered and scarcer 1 month after nerve
•
Hanamatsu et al.
271
cutting in the SB and Control groups compared with those on the Normal Side (Figs 3 and 4). Immature myelinated nerve fibers that were still undergoing regeneration were stained with toluidine blue. Not all of the regenerated myelinated fibers were inside the fasciculi. Some were located in the connective tissues of the epineurium. Swelling, formation of vacuoles, and shriveled axons in some myelinated nerve fibers were observed even at this stage. The distribution of regenerated myelinated fibers was not equal in all fasciculi. In the regenerating myelinated fibers, some axons, which were surrounded by projections emanating from what seemed to be fibroblasts, formed small fasciculus-like structures (Compartmentation). Some fasciculi contained a few scattered regenerated fibers, while in others, a large number of regenerated fibers was growing uniformly. The vessels of the epineurium, perineurium, and endoneurium seemed to have reassumed their shape and regained their function. Electron microscopy showed that numerous myelinated nerve fibers were in the process of regeneration in both groups. At the same time, a relatively large number of degenerated nerve fibers was evident. We found that some axons formed similar structures in some myelinated fibers that began regeneration outside the fasciculi from which they had originated. In the degenerated nerve fibers, we found the following: gaps between axons and myelin sheaths were formed owing to axon shrinking; axons had totally disappeared and as a result, the surrounding myelin sheaths turned into wavy myelin balls; cells that appeared to be macrophages ingested and phagocytosed the degenerated nerve fibers. Eight Months After Nerve Cutting Judging from the number and size of myelinated nerve fibers and the thickness of myelin sheaths, regeneration had progressed considerably in the 8-month subgroups compared with the findings from the 1-month subgroups (Figs 5 and 6). Under optical microscopy, the myelinated nerve fibers that had almost completed regeneration occupied most of the fasciculi in both groups. In some fasciculi, the myelinated nerve fibers were not mature; myelin sheaths were irregularly shaped and seemed to be in the premature stage of regeneration. Several axons that had regenerated outside the original fasciculi were occasionally surrounded by a structure like the perineuria to form small fasciculi. Electron microscopy showed that a large number of myelinated nerve fibers had almost completed regeneration in the SB group (Fig 7). This finding coincided with the findings of optical microscopy. However, a
272
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
Fig 3. Optical micrograph of infraorbital nerves 1 month after severance in the SB group. Smaller diameter myelinated nerve fibers, seemingly undergoing regeneration, have densely stained myelin sheath (toluidine blue; ⫻60).
large number of insufficiently myelinated nerve fibers were still in the process of regeneration in the Control group (Fig 8). No nerve fibers had degenerated in either group at this stage.
Discussion The function of the sensory afferent pathway of the trigeminal nerve can be objectively measured by stimulating the trigeminal nerve to assess the
Fig 4. Optical micrograph of infraorbital nerves 1 month after severance in the Control group. Regeneration rate of each fasciculus in the same nerve trunk differed (toluidine blue; ⫻60).
functional recovery of the severed nerve. We used SEPs to objectively and neurophysiologically observe the functional recovery of severed nerve. SEPs are elicited from a region on the skull surface corresponding to the somatosensory field by stimulating the sensory nerve at the periphery and the sum of conduction or the compound potentials of a group of nerve fibers, thus reflecting the function of stimulus conduction.6 The P1-N1 amplitude of SEPs
Sympathetic Block for Nerve Regeneration
•
Hanamatsu et al.
273
Fig 5. Optical micrograph of infraorbital nerve from the SB group 8 months after nerve cutting. Many myelinated fasciculi with regeneration almost completed were seen (toluidine blue; ⫻60).
is dependent on the intensity of the stimulus and is often adopted as an index of the sum of sensation to stimulus in humans or animals.5 In the SB group, the SEP amplitude recovered significantly earlier in comparison to the Control group (Fig 1). The variability of the SEP amplitude tended to be smaller in the SB group than the Control group. We compared the degree of regen-
eration 1 and 8 months after nerve cutting by measuring myelinated nerve fibers/unit area, but differences were not significant in any subgroup in the 2 groups. We found that the number of regenerated myelinated fibers of the 1-month subgroups determined by electron microscopy was slightly different from that determined by optical microscopy. We believe that optical microscopy failed to distinguish
Fig 6. Optical micrograph of infraorbital nerve from the Control group 8 months after nerve cutting. Some fasciculi contain numerous myelinated nerve fibers still regenerating. Myelin sheaths are irregularly shaped and seem to be in the premature stage of regeneration process (toluidine blue; ⫻80).
274
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
Fig 7. Electron micrograph of infraorbital nerve from the SB group 8 months after nerve cutting. Myelinated nerve fibers show the regeneration process is almost complete.
between regenerated fibers and myelin sheaths that had escaped phagocytosis by macrophages, and thus included these sheaths in counting of the regenerated fibers and thereby overestimated the number of myelinated fibers. In the 8-month subgroups, we assumed that optical microscopic measurements were not overestimated, because electron microscopy did not reveal anything resembling remnants of myelin sheaths. The number of myelinated nerve fibers/unit area between the Cut and Normal Sides did not significantly differ in the
8-month subgroups of both groups. We concluded that regeneration in terms of number of fibers was almost complete 8 months after nerve cutting. We measured the number of myelinated nerve fibers, not of axons, as a tool with which to assess the degree of nerve regeneration to avoid the influence of excessive sprouting resulting from the disturbance of axonal growth. The results showed that the distribution of myelinated nerve fibers on the Normal Side was 2-phased; smaller and larger diameter fibers of up to 8 m and over 8 m, respec-
Sympathetic Block for Nerve Regeneration
•
Hanamatsu et al.
275
Fig 8. Electron micrograph of infraorbital nerve from the Control group 8 months after nerve cutting. Many myelinated nerve fibers are still immature.
tively (Fig 2). In both 1-month subgroups, the distribution was single-phased, with the myelinated fibers with smaller diameter being more prevalent than in the Normal Side and only a few fibers of the larger diameter. We attributed these findings to immature myelinization and assumed that the myelin sheath of the smaller fibers would thicken over time and that their diameters would accordingly increase. We also presumed that nerve fibers that had not been myelinated sufficiently (resulting from delayed regeneration) to be recognized under
optical microscopy would gradually join the larger diameter fibers. We considered that the SB 8-month subgroup had attained restoration and regeneration that was comparable to those of the Normal Side, although this subgroup contained fewer nerve fibers between 11- and 13-m diameter compared with the Normal Side (Fig 5). Therefore, we concluded that the myelinated nerve fibers reached maturity at 8 months in the SB group. The Control 8-month subgroup seemd to lack a group of large nerve fibers, especially those larger than
276
Regional Anesthesia and Pain Medicine Vol. 27 No. 3 May–June 2002
15-m diameter in comparison to the corresponding SB subgroup of the Normal Side. Consequently, it was assumed that the SB group was still undergoing regeneration even after 8 months (Fig 6). Compared with the electrophysiologic degree of recovery measured by SEPs, fibers with the highest diameter among all the regenerated myelinated fibers played a crucial role in the high recovery rate. Wallerian degeneration occurs at the distal portion of the severed region,7 which we observed even 1 month after the nerve transection in the present study. Blood flow in the endoneurium does not necessarily flow in a regular direction owing to widespread anastomosis and changes direction as soon as pressure or a disturbance of blood flow arises. The microcirculation of a fragment of the rabbit tibial nerve separated by cutting at the central and peripheral sides remains unaffected if the blood vessels from outside are retained.8 That report suggests that the vascular network has considerable reserve capacity. The intraneural vascular system contains adrenergic nerve endings regulated by sympathetic nerves, and postsynaptic conduction causes vasoconstriction via alpha receptors.9 The peripheral vessels constantly receive impulses from vasoconstrictors (resting tone). Therefore, interruption of sympathetic activity inevitably causes vasodilation. A cervical SB temporarily interrupts outflow from the cervical sympathetic nerves and, consequently, produces vasodilation of peripheral vessels governed by the cervical sympathetic nerve. We believe that the effects of the SB are further extended from nutrient vessels around the injured area to the vascular network of the nerve fragment both central and peripheral to the severed region. Our results using SEPs and the distribution of fiber
diameters suggest that the SB promotes myelinization following axonal growth and the maturation of myelin sheaths. In conclusion, we demonstrated that SB promoted the recovery and regeneration of severed ION. The clinical relevance of these findings in humans remains indeterminate.
Acknowledgment The authors thank Prof. Shigeo Aiyama for reviewing the manuscript.
References 1. Hegtvedt AK. Evaluation and management of orofacial sensory nerve injuries. Ohio Dent J 1992;66:16-17. 2. Shankland WE. Anatomical considerations for mandibular implant surgery: Two common post-operative problems. Dent Today 1991;10:48-49. 3. Moore DC. Stellate ganglion block. Springfield, IL: Thomas; 1954. 4. Eriksson E. Atlas der Lokalanasthesia. New York, NY: Thieme; 1970. 5. Toda K, Ichioka M, Suda H, Iriki A. Effects of electroacupuncture on the somatosensory evoked response in rat. Exp Neurol 1979;63:652-658. 6. Liveson JA, Spielholz NI. Interpretation of electrodiagnostic data. In: Liveson JA, Spielholz NI, eds. Peripheral Neurology: Case Studies in Electrodiagnosis. Philadelphia, PA: Davis; 1979:3-8. 7. Gilliatt RW, Hjorth RJ. Nerve conduction during Wallerian degeneration in the baboon. J Neurol Neurosurg Psychiatry 1972;35:335-341. 8. Lundborg G. Structure and function of the intraneural microvessels as related to trauma, edema formation and nerve function. J Bone Joint Surg 1975;57A:938948. 9. Lundborg G. Nerve Injury and Repair (Japanese edition). Ikuta Y, Ochi M, Japanese eds. Tokyo, Japan: Hirokawa; 1991:4-57.