- Email: [email protected]
Nerve conduction and microanatomy in the rabbit sciatic nerve after gradual limb lengthening–distraction neurogenesis
ELSEVIER
Journal of Orthopaedic Research Journal of Orthopaedic Research 21 (2003) 3 6 4 3
www.elsevier.com/locate/orthres
Nerve conduction and microanatomy in the rabbit sciatic nerve after gradual limb lengthening-distraction neurogenesis Atsushi Yokota *, Munekazu Doi, Hisashi Ohtsuka, Muneaki Abe * D q m tnzent of 01tlzopedic Surgery, Oraha Medicul College, 2-7, Daigahu-machi, Takatrukr 569-8686, Japan
Received 7 November 2001, accepted 12 June 2002
Abstract
T o clarify how the peripheral nerve adapts to elongation during gradual limb lengthening, electrophysiological and histomorphometric examinations were performed on the sciatic nerves in 18 rabbits. External fixators were used to lengthen the right femora by 30 mm (30'!h), at a daily rate of 0.5 mm (Group 1) or 2.0 mm (Group 2). Examinations were performed immediately after the limb lengthening procedure. Electrophysiologically, mild conduction slowing was observed in Group 1; a conduction block was evident in Group 2. Histologically, the mean diameter of myelinated fibers was unchanged in Group 1, but a significantly decreased diameter was observed in Group 2. Electron microscopy revealed that mild degenerative change of unmyelinated axons occurred sporadically in two cases in Group 2, but neither group showed evidence of thinning of myelin sheath of myelinated fibers. The mean internodal length (between nodes of Ranvier) of teased fibers was 1216 & 295 pm in the control contralateral side, 1484 & 347 pm in Group 1, and 1467 f 322 pm in Group 2. Thus the internodes were lengthened by 22.1'%1(Group 1) and 20.7% (Group 2) in comparison with those of the controls. Straightening of the geometry of paranodal myelin sheath was significantly correlated with the rate of distraction. These results indicate that myelinated nerve fibers adapt to gradual elongation by lengthening each Schwann cell body, not by proliferation of Schwann cells. 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved.
Introduction
Owing to improvements in surgical techniques and external fixation devices, limb lengthening has become a common procedure to correct many kinds of congenital anomalies as well as acquired limb length discrepancies [3,13,15]. Unfortunately, experimental data about the effect of gradual stretching on peripheral nerves are few [1,7,9,11,12,14-16,1S,20-23] despite the fact that this technique has the potential for nerve injury [6,8]. In 1989, Ilizarov [12] reported that lengthening of the Schwann cell and myelinic membrane occurs during limb lengthening. He concluded that the mechanism involved was identical to that seen in embryonic peripheral nerve formation. In his study, however, he
* Corresponding authors. Tel.: +81-726-83-1221; fax: +81-726-836265. E-mail addresses: [email protected] (A. Yokota), [email protected] ( M . Abe).
observed only ultrastructural changes in transverse sections of nerve specimens, and the longitudinal structural changes in the lengthened fiber were not observed. The longitudinal structure at the microscopic level is important for understanding the distraction neurogenesis of the lengthened nerve fibers. Observations of teased nerve fibers, which involve the isolation of single myelinated fibers, permit evaluation of consecutive internodes or segments of myelinated nerve fibers. This procedure can also be used to establish the frequency of certain neuropathologic abnormalities in a manner that is generally more sensitive than that found when fixed sections are used [5]. Thus, the procedure can be used to provide detailed information about changes in the histomorphometric parameters of lengthened myelinated fibers. The purpose of this study was to characterize the structural and functional response of in vivo distraction neurogenesis and to investigate the influence of the rate of distraction on changes in the conduction property and histomorphometric parameters of the peripheral nerve during limb lengthening.
0736-0266/03/$ - see front matter 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 2 ) 0 0 1 0 0 - 6
A. Yokota et al. I Journal of Orthopaedic Reseurch 21 (2003) 36-43
Materials and methods
37
lyzer system was then used to evaluate these digitized histologic images.
Study design
Tran.sverse sections Small cubes (approximately 1 mm in length) of the specimens were prefixed by immersing in the same fixative as described above for 2 h a t 4 "C. The specimens were washed in several changes of buffer and postfixed by immersion in 1% Os04 in 0.1 M phosphate buffer, p H 7.4. for 2 h at 4 "C. After dehydration in serial ethanol solutions and propylene oxide, specimens were embedded in Epon 812. Transverse sections of thickness 1 pm were stained with 1% toluidine blue. The diameters of all myelinated fibers were determined. Ultrathin sections of the nerve were cut with a diamond knife (Diatome, Bienne, Switzerland) on an ultramicrotome (OM u2; Reichart, Leica Scientific Instruments, Nussloch, Germany). Then these sections were double stained with uranyl acetate and lead citrate and were examined by electron microscopy (H-7100: Hitachi, Tokyo, Japan).
This study was performed o n 18 skeletally mature male Japanese white rabbits (JAPAN SLC Inc., HamdmdtSu, Japan) with a body weight of 3.0-3.5 kg. The rabbits were randomly separated into two test groups. with nine animals in each group. The right femora were lengthened by set amounts once per day over the course of the study to give a total lengthening of 30 mm (approximately 30'!4) of the total length of the femur). A distraction rate of 0.5 mm per day (Group 1) or 2.0 mm per day (Group 2) from the first day was used because pilot studies had indicated that early union of the osteotomy site could occur in Group 1 when lengthening began after a waiting period of several days. Full weight-bearing was allowed immediately after surgery. All analyses were performed immediately after completion of the limb lengthening procedure. We planned to harvest 10 rabbits for electrophysiological analysis and the remaining eight rabbits for histological and histomorphometric analysis. The protocol used was approved by the Osaka Medical College animal care and use committee.
Nerae teasing We measured the distance between the nodes of Ranvier (internodal length; Fig. 1) of myelinated fibers to clarify whether proliferation of Schwann cells occurs o r each Schwann cell is lengthened. After a 1.5-h post-fixation period in 1%) Os04 in 0.1 M phosphate buffer, bundles were passed through 450/11, 66%, and 100% glycerin for 24 h at each concentration and at a temperature of 37 "C. Under a surgical microscope, single myelinated fibers were teased apart with fine forceps. Care was taken to obtain the fibers in a random fashion. Four hundred fibers were taken from the two test groups and the control group (a total of 1200 fibers), and both their diameter and internodal length were measured. The diameter was classified as the mean value of three counts made at different parts of each internode excluding the paranode (Fig. 1). The internodal length was classified according to the length of the whole internode including the paranode. We observed the geometry of the paranode- node-paranode (PNP) region [24] (Fig. I). This region was classified into one of three types: (A) roundthe shape of the paranodal myelin sheath keeps its physiological curvature, and the nodes of Ranvier d o not widen; (B) straight-the shape of the paranodal myelin sheath straightens, but the nodes of Ranvier d o not widen; (C) wide-the shape of the paranodal myelin sheath straightens, and the nodes of Ranvier widen (Fig. 2A-C). Morphologic changes in the internode were also evaluated.
Surgical procedure With the rabbit under intravenous anesthesia with sodium pentobarbital (20 mg/kg body weight), the right femur was explored through a lateral approach and a muscle-splitting technique. A unilateral external fixatordistractor (M101; Orthofix, Verona, Italy) was applied to the lateral aspect of the femur under aseptic technique. With use of a reciprocating saw under irrigation with saline solution, the femur was osteotomized at the mid-section, through the gap between the middle two transfixing pins. The bone was completely covered with local muscle and fascia and a tension-free skin closure was performed. Electrol,liysiological analysis We assessed nerve function by measuring the compound nerve action potential (CNAP) of the sciatic nerve with use of an electromyography device (DANTEC Neuromatic-2000; Medtronic Functional Diagnotics AIS, Skovlunde, Denmark). After the animals were anesthetized with sodium pentobarbital, the sciatic nerve and proximal part of the tibial nerve were explored through a lateral approach. Two stainless steel bipolar electrodes (UM2-5050; Unique Medical Inc., Osaka, Japan) were used: one for stimulating the nerve supramaximally at the point where the tibial nerve entered the gastrocnemius and the other for recording the CNAP at the sciatic notch. The electrical stimulation used was a 10-Hz, 0.1-ms rectangular wave. Room temperature was maintained at 24 "C. We recorded and averaged 20 responses and measured peak amplitude and conduction velocity by measuring the distance between the stimulating and recording electrodes. The same procedure was performed on the contralateral untreated side. Conduction velocity and peak amplitude in these recordings were expressed as a percentage of the values obtained for the contralateral side of each animal. Histological and Izistonzorpliometric analysis After the animals were anesthetized with sodium pentobarbital, they were perfused through the heart with 3 I of fixative solution containing 2.5% glutaraldehyde and 2%) paraformaldehyde in 0. I M phosphate buffer, p H 7.4. This technique let us maintain the in situ length of the elongated nerve. A 10-mm long sample of the sciatic nerve was dissected out at the level of the mid-thigh (osteotomy). The tibial bundle was isolated from the fibular bundle and divided into two segments, one for transverse sectioning and the other for teased fiber analysis. Four specimens from the contralateral side were randomly chosen as a control group. Image capture analysis of the finished sections was performed with use of light microscopy (VANOX-T; OLYMPUS, Tokyo, Japan) in conjunction with a C C D video camera (DXC-950; Sony, Tokyo, Japan). All quantitative histomorphometric parameters were obtained through use of a semiautomatic image analyzer system (MCID version 3.0; Imaging Research Inc., Ont., Canada). Microscopy images of the histologic specimens were captured by camera and then digitally transferred to a computer. The image ana-
Statistical analysis Results are presented as mean values and the standard deviation. An unpaired t-test (two-tailed) was used to compare the percentage value of conduction velocity and peak amplitude of both test groups. A paired t-test was used to compare the experimental and control sides with regard to conduction velocity and peak amplitude. Analysis of variance (ANOVA) with Scheffe's post-hoc t-test was used to compare quantitative, histomorphometric parameters of each group. Chisquare test with Yates continuity correction was used to compare the
.
.
i
T
I1
.
. t
t
NOR
NOR
~
PN . _PN
~
Fig. 1. A schematic drawing of the teased myelinated nerve fiber to show the terminology used. IL; internodal length, NOR; node of Ranvier, PNP; pafanode-node-paranode region, PN; paranode, MS; myelin sheath, Ax; axon.
38
A. Yokota el
(11.
I Journal oj Orthopaedic Research 21 (2003) 36-43
Electrophysiological analysis The results of Group 1 showed mild changes in response, averaging an 18% delay in conduction velocity and a 20% reduction in amplitude in comparison with results of the contralateral side ( P < 0.001 and P < 0.009, respectively). The results of Group 2 showed an average 23% delay in conduction velocity and 71% reduction in amplitude in comparison with those of the contralateral side ( P < 0.008 and P < 0.004, respectively). The percentage of conduction velocity for the elongated sciatic nerve was similar for the two test groups (P = 0.29; Table 1). However, the percentage of amplitude in Group 2 was significantly lower than that found for Group 1 (P < 0.0001; Table 2). Histological and histornorphometric analysis Transverse sections The mean diameter of myelinated fibers is shown in Table 3. Myelinated fibers in Group 2 significantly narrowed in comparison to those of the control group and Group 1 (P < 0.OOOl). However, the mean myelinated fiber diameter did not differ between the control group and Group 1 ( P = 0.42). Electron microscopy revealed no evidence of degeneration, fibrosis, or endoneurial edema in the control group or Group 1 (Fig. 3A-D). Thinning of the myelin sheath of myelinated fibers was not observed in either test group (Fig. 3C and E), but a degenerative change of unmyelinated axons occurred sporadically in two cases in Group 2 (Fig. 3F).
Fig. 2. Classification of the shape of the PNP region of teased myelinated fibers: (A) round (B) straight (C) wide. Arrowheads indicate straightened myelin sheath. Arrow indicates the widening of the node of Ranvier. Original magnification 160x .
categorical data of geometry of the PNP region of each group. Values of P < 0.05 were considered to be significant.
Results
One animal from each test group had to be substituted, one because of femoral fracture and one because of respiratory infection (drop-out rate of 1 IYn). The remaining 16 rabbits showed good tolerance to the lengthening procedure. Therefore, eight rabbits were used for electrophysiological analysis and another eight were used for histomorphometric analysis. Pin loosening, infection, or soft-tissue related problems did not develop in any animal. The external fixation devices caused no additional injuries.
Teased jibers The mean diameters and internodal lengths of teased fibers for each group are shown in Table 3 . There was no difference between the mean diameters of each group; however, the internodal length of both test groups was significantly longer than that of the control group ( P < 0.0001). The percentage of elongation of internodes compared to the control group was 22.1% in Group 1 and 20.7% in Group 2. There was no evidence of segmental demyelination, remyelination, or dilatation of the Schmidt-Lantermann incisures (Fig. 4A-C). A straight type of PNP region was observed in 3.3% of PNP regions in the control group, 12.1%)in Group 1, and 21.6% in Group 2 thus a tendency toward straightening in the geometry of this region was significantly correlated with the rate of distraction (P < 0.0001, Table 3).
Discussion Little is known about how the peripheral nerve adapts to gradual elongation since only a few reports are available on histomorphometric analysis in peripheral nerves during gradual limb lengthening [1,7,16]. In the
A. Yokota et al. / Journal of Orthopaedic Reseurcl~21 (2003) 36-43
39
Table 1 Conduction velocity of the sciatic nerve Group 2
Group 1 Rabbit 1
2 3 4
Treatment (mls)
Contralateral (mls)
Treatmenticontralateral
59.8 64.2 69.2 66.3
77.4 77.4 83.3 79.4
77.3 82.9 83.1 83.5
64.9 i4.0
79.4 i2.8
81.7
Rabbit
WU)
1 2 3 4
* 2.9
Treatment (m/s)
Contralateral (m/s)
(%)
Treatmentkontralateral
51.4 70.0 70.5 53.5
71.4 86.6 82.2 78.5
72.0 80.8 85.8 68.2
61.4 zt 10.3
79.7 f 6.4
76.7 f 8.0
Treatment (PV)
Contralateral (FV)
Treatmentkontralaterdl
70 124 52 100
264 400 216 292
26.5 31.0 24.1 34.2
94 f 31.9
293 i 77.9
29.0 f 4.5"
The date in the bottom row are mean f SD
Table 2 Peak amplitude of the sciatic nerve
Group 1 Rabbit 1
2 3
4
Group 2 Treatment
Contralateral (UV)
Treatmentlcontralateral
(WV)
240 208 204 336
292 280 244 420
82.2 74.3 83.6 80.0
247 f 61.5
309 f 76.8
80.0 zt 4.1
Rabbit
(%)
1 2 3 4
(%)
The data in the bottom row are mean f SD. " P < 0.0001 compared with Group 1
Table 3 Quantitative, histomorphometric parameters of each specimen Group I ( n = 4)
Group 2 ( n = 4)
7.528
6.877"
7136 i421
7753 f 344
11.8 f 3.9
12.4 i 3.9
12.2 i4.0
1216 f 295
1484 i 347h
1467 =?z 322'
Control ( n = 4) Traiwwrsr .section Mean fiber diam- 7.606 eter (pm) Total myelinated 7504 & 306 fibers (no.)
Tiused ,fiber Mean fiber diameter (pm) Mean internodal length (pm)
The geometry of the PNP region 95.6 86.0 Round ('%) 76.9 ( n = 61 5)' ( n = 765) ( n = 688)d Straight (%) 3.3 12.1 21.6 ( n = 173)' (TI = 97)d ( P I = 26) Wide (YO) 1.1 1.9 1.5 ( n = 12) ( n = 15) in = 9) " P < 0.0001 compared with Group 1 and the control group. b f < 0.0001 compared with the control group, and P = 0.75 compared with Group 2. ' P < 0.0001 compared with the control group. f < 0.0001 compared with Group 2 and the control group. ' P < 0.0001 compared with Group 1 and the control group.
current study, the internodal length of myelinated fibers of both test groups was significantly longer than that of
the control group. In addition, the diameter of myelinated fibers of Group 1 (distraction rate 0.5 mm per day) was similar to that of the untreated contralateral side. These results differ from those of acute elongation studies in which elongated nerve fibers become narrower [lo], thereby suggesting that some nerve fiber grows. However, if proliferation of Schwann cells occurs in stretched nerve fibers, then the internodal length should be shorter and the myelin sheath should become thinner [2]. Ultrastructual examination of the current study disclosed no evidence of thinning of the myelin sheath in either test group. Thus, the results of the current study primarily indicate that in vivo distraction neurogenesis is achieved not by means of proliferation of Schwann cells, but by the elongation of each Schwann cell body itself. In the study of Battiston et al. [I], the increase in internodal length was not as much as the total amount of lengthening of the nerve. They therefore postulated that Schwann cell proliferation occurs in addition to a lengthening of the myelinated fibers. Elongated nerve fibers, however, usually lose their wavy structure and become straight [20]. In addition, bowstringing of the nerve across the joints could occur during limb lengthening [21]. These alterations could compensate for some amount of lengthening that the nerve undergoes. In the current study, we did not measure the total amount of lengthening of the nerve by means of epineural marking
40
A . Yoliolu ('I (11. I Joirrnrrl of' Orthopuerlic~Resrrrrcli 21 (2003) 3 6 ~43
(E)
(F)
Fig. 3 . Electron micrographs of transverse section of sciatic nerves in low magnitude (A, C. E, 1OOOx. the scale bar indicates 5 pm) and high magnitude (B. D. F:, 12.500 x , the scale bar indicates 0.5 pm): (A, B) control; (C, D) Group I ; (E, F) Group 2. No degenerative changes were seen in Group 1. In Group 2, one axon in a cluster of unmyelinated fiber has a watery appearance (F, *). indicating axonal degeneration. But many other unmyelinated fibers appear normal.
suture because the exploration at the first procedure would cause adhesion around the nerve. The quotient of
increase of the internode was approximately 70-75% of that of bone lengthening.
A . Yokotci et ul. I Journul of' Orthopuedic Re~c.ar.ch21 12003) 36 43
41
Fig. 4. Consecutive lengths of teased myelinated fibers: (A) control. ( R ) Group I . (C) Group 2. The scale bar indicates 200 pin. Arrows indicate the nodes of Ranvier. Note the internodal lengths of Group 1 and Group 2 fibers are longer than that of the control group. Original magnification 16x
Histological study of transverse sections of the current study indicated only modest morphologic changes in the peripheral nerves after gradual limb lengthening. N o evidence of degenerative changes was observed in Group 1. In two cases in Group 2 (distraction rate 2.0 mm per day), mild degeneration of unmyelinated axons [4] was shown, a finding similar to the observations by Fink et al. [7] in canine tibia1 and peroneal nerves. Previous histological studies yield contradictory results. Some studies concluded no degenerative changes [1,9, 1 I], whereas other studies observed clear-cut degeneration after gradual limb lengthening [14,20,22]. The difference between these results was probably caused by differences in the amount or rate of distraction. lppolito et al. [14] and Skoulis et al. [22] have shown a clear relationship between the amount or rate of distraction and the degree of degeneration. In addition, the differences between species and the difference in vulnerability of each peripheral nerve to the distraction force [l8] could affect these results. In an electrophysiological analysis of the current study, we recorded the CNAP orthodromically. The clinical significance of this method is identical to that of a sensory nerve conduction study [19], which has been reported to be a more sensitive index to detect subtle conduction slowing than motor nerve conduction [17]. The results showed that the conduction property of the nerves was affected by the rate of lengthening. The results of Group 1 showed mild conduction slowing; on the other hand, the results of Group 2 showed
an apparent conduction block, with significant diminution of the amplitude in the presence of mild slowing of conduction velocity. This result was similar to results of other studies [I 1,20,22,23]. Shibukawa et al. [20] concluded that reduction of amplitude is a more sensitive predictor of nerve dysfunction than delay of velocity during limb lengthening. Because the PNP region is responsible for impulse conduction [24], structural alterations of this region would result in the change of the conduction property of the nerve. In the current study, straightening of the geometry of the paranodal myelin sheath was significantly correlated with the rate of distraction. We assumed that geometry of this area would reflect the capability of each Schwann cell to lengthen its body against a given rate of distraction. In addition, the mean diameter of myelinated fibers of Group 2 was significantly decreased compared to that of Group I and the control group. The hypothesis about the development of nerve impairment supported by these findings is as follows: (1) If the rate of distraction is slow, then the Schwann cells can adapt to elongation, which maintains the geometry of the PNP regions. As a result, nerve dysfunction does not manifest (Fig. 5B). (2) If the rate of distraction is too rapid, then the Schwann cells can no longer adapt to elongation, and straightening of the paranodal myelin sheath results. In addition, the nerve fibers become thinner under the tension, like a stretched elastic band (Fig. 5C). As a result, nerve function is impaired.
A. Yokota et 01. I Journal of Orthopciedic Research 21 (2003) 3 6 4 3
42
a
0
A t NOR
t NOR
a‘
aa’
b
0
B b’
bb’
C
0
C
cc’ C’
Fig. 5. A schematic drawing of the longitudinal (left) and transverse (right) sections of the myelinated nerve fiber: (A) normal, (B) after slow elongation, (C) after rapid elongation. LoS: length of the Schwann cell; NOR: node of Ranvier. Broken lines indicate the transection line. Note the geometry of the paranode becomes straight and the diameter of fiber decreases after rapid elongation.
The current study was limited to assessing the gross geometry of the PNP region by means of teased fibers. The study could not quantify the gap length at the nodes of Ranvier although its procedure enabled us to observe numerous samples in each group. Ikeda et al. [1 11 attained an elongation of 300/0 in rabbit sciatic nerves with the daily rate of 4.0 mm. Kern et al. [I61 obtained a 20Y0elongation of rabbit tibias with the rate of 0.7 mm per day. Both of them demonstrated widening of the nodes of Ranvier with use of longitudinal thin sections from the fixed specimens. The straight type of PNP region in our classification for teased nerve fibers might have some increase in the nodal gap. In conclusion, the current study suggests that if lengthening is performed slowly enough as not to manifest nerve dysfunction, then peripheral nerves can accommodate to the elongated new length with appropriate elongation of the internodal length and preservation of the diameter of the nerve fibers. This indicates that distraction neurogenesis is achieved by the lengthening of each Schwann cell body, and not by the proliferation of Schwann cells.
Acknowledgements The authors would like to thank Toshio Hasegawa, M.D., Ph.D., for advice and encouragement throughout the current study. The authors also thank Tomoyuki Tsujimura, M.D., Taisuke Horiguchi, M.D., and Mr. Tsuguya Naito for technical assistance. The valuable comments made by Nozomu Inoue M.D., Ph.D., were greatly appreciated. The authors have received nothing
of value from a private or commercial entity for the data described in this study. References Battiston B, Buffori P, Vigasio A, et al. The effects of lengthening of nerves. Ital J Orthop Traumat 1992;18:79-86. Cragg BG, Thomas PK. The conduction velocity of regenerated peripheral nerve fibers. J Physiol 1964;171:164-75. De Bastiani G, Aldegheri R , Renji-Brivio L, Trivella G. Limb lengthening by callus distraction (callotasis). J Pediatr Orthop 1987;7: 129-34. Dyck PJ, Hopkins AP. Electron microscopic observations on degeneration and regeneration of unmyelinated fibers. Brain 1972;95:223-34. Dyck PJ, Giannini C , Lais A. Pathologic alterations of nerves. In: Dyck PJ, Thomas PK, Griffin JW, et al., editors. Peripheral neuropathy. 3rd ed. Philadelphia: WB Saunders; 1993. Faber FWM, Keessen W, van Roermund PM. Complications of leg lengthening: 46 procedures in 28 patients. Acta Orthop Scand 1991;62:327-32. Fink B, Neuen-Jacob E, Lehmann J. et al. Changes in canine peripheral nerves during experimental callus distraction. Clin Orthop 2000;376:252--67. Galardi G, Conii G, Lozza L, et al. Peripheral nerve damage during limb lengthening. J Bone Joint Surg [Br] 1990;72:1214. Gil-Albarova J, Melgosa M , Gil-Albarova 0. Canadell J. Soft tissue behavior during limb lengthening: an experimental study in lambs. J Pediatr Orthop B 1997;6:266-73. Haftek J, Poland W. Stretch injury of peripheral nerve. J Bone Joint Surg [Br] 1970;52:354--65. lkeda K, Tomita K , Tanaka S. Experimental study of peripheral nerve injury during gradual limb elongation. Hand Surg 2000; 5:41-7. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues. Part I. The influence of stability of fixation and soft-tissue preservation. Clin Orthop 1989;238:249- 8 I . Ilizarov GA. Clinical application of the tension-stress effect for limb lengthening. Clin Orthop 1990;2S0:8 -26. Tppolito E, Peretti G , Bellocci M, et al. Histology and ultrastructure of arteries, veins, and peripheral nerves during limb lengthening. Clin Orthop 1994;308:5&62. Kawamura B, Hosono S. Takahashi T, et al. Limb lengthening by means of subcutaneous osteotomy: experimental and clinical studies. J Bone Joint Surg [Am] 1968;50:851-78. Kerns J, Urabe T, Bleasdale T, et al. Effects of gradual bone lengthening on the rabbit tibia1 nerve. Scand J Plast Reconstr Hand Surg 2001;35:361-8. Ludin HP, Lutschg J, Valsangiacomo F. Vergleichende untersuchung orthodromer und antidromer sensibler nervenleitgeschwindigkeiten: 2. befunde bei polyneuropathien und bei status nach polyradikulitis. Z EEG E N G 1977;8:180-6. Makarov M R , Birch JG, Delgado M R , et al. Effects of external fixation and limb lengthening on peripheral nerve function. Clin Orthop 1996;329:310-6. Oh SJ. General concepts of electrodiagnostic studies in neuromuscular disease. In: Oh SJ, editor. Clinical electromyography: nerve conduction studies. 2nd ed. Baltimore: Williams & Wilkins; 1993. Shibukawa M, Shirai Y. Experimental study on slow-speed elongation injury of the peripheral nerve: electrophysiological and histological changes. J Orthop Sci 2001;6:262-8. Simpson AHRW, Kenwright J. The response of nerve to different rates of distraction. J Bone Joint Surg [Br] 1992:74(Suppl 3):326. Skoulis TG, Vekris MD, Terzis JK. Effect of distraction osteogenesis on the peripheral nerve: experimental study in the rat. J Reconstr Microsurg 1998;14:565-74.
A Yokota et rrl. I Journul of Orthopedic Reseurclz 21 (2003) 36-43
[23] Strong M, Hruska J, Czyrny J, et al. Nerve palsy during femoral lengthening: MRI, electrical, and histologic findings in the central and peripheral nervous systems-a canine model. J Pediatr Orthop 1994;14:347-51.
43
[24] Thomas PK, Berthold C-H, Ochoca J. Microscopic anatomy of the peripheral nerve system. In: Dyck PJ, Thomas PK, Griffin JW, et al., editors. Peripheral neuropathy. 3rd ed. Philadelphia: WB Saunders; 1993.
Journal of Orthopaedic Research Journal of Orthopaedic Research 21 (2003) 3 6 4 3
www.elsevier.com/locate/orthres
Nerve conduction and microanatomy in the rabbit sciatic nerve after gradual limb lengthening-distraction neurogenesis Atsushi Yokota *, Munekazu Doi, Hisashi Ohtsuka, Muneaki Abe * D q m tnzent of 01tlzopedic Surgery, Oraha Medicul College, 2-7, Daigahu-machi, Takatrukr 569-8686, Japan
Received 7 November 2001, accepted 12 June 2002
Abstract
T o clarify how the peripheral nerve adapts to elongation during gradual limb lengthening, electrophysiological and histomorphometric examinations were performed on the sciatic nerves in 18 rabbits. External fixators were used to lengthen the right femora by 30 mm (30'!h), at a daily rate of 0.5 mm (Group 1) or 2.0 mm (Group 2). Examinations were performed immediately after the limb lengthening procedure. Electrophysiologically, mild conduction slowing was observed in Group 1; a conduction block was evident in Group 2. Histologically, the mean diameter of myelinated fibers was unchanged in Group 1, but a significantly decreased diameter was observed in Group 2. Electron microscopy revealed that mild degenerative change of unmyelinated axons occurred sporadically in two cases in Group 2, but neither group showed evidence of thinning of myelin sheath of myelinated fibers. The mean internodal length (between nodes of Ranvier) of teased fibers was 1216 & 295 pm in the control contralateral side, 1484 & 347 pm in Group 1, and 1467 f 322 pm in Group 2. Thus the internodes were lengthened by 22.1'%1(Group 1) and 20.7% (Group 2) in comparison with those of the controls. Straightening of the geometry of paranodal myelin sheath was significantly correlated with the rate of distraction. These results indicate that myelinated nerve fibers adapt to gradual elongation by lengthening each Schwann cell body, not by proliferation of Schwann cells. 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved.
Introduction
Owing to improvements in surgical techniques and external fixation devices, limb lengthening has become a common procedure to correct many kinds of congenital anomalies as well as acquired limb length discrepancies [3,13,15]. Unfortunately, experimental data about the effect of gradual stretching on peripheral nerves are few [1,7,9,11,12,14-16,1S,20-23] despite the fact that this technique has the potential for nerve injury [6,8]. In 1989, Ilizarov [12] reported that lengthening of the Schwann cell and myelinic membrane occurs during limb lengthening. He concluded that the mechanism involved was identical to that seen in embryonic peripheral nerve formation. In his study, however, he
* Corresponding authors. Tel.: +81-726-83-1221; fax: +81-726-836265. E-mail addresses: [email protected] (A. Yokota), [email protected] ( M . Abe).
observed only ultrastructural changes in transverse sections of nerve specimens, and the longitudinal structural changes in the lengthened fiber were not observed. The longitudinal structure at the microscopic level is important for understanding the distraction neurogenesis of the lengthened nerve fibers. Observations of teased nerve fibers, which involve the isolation of single myelinated fibers, permit evaluation of consecutive internodes or segments of myelinated nerve fibers. This procedure can also be used to establish the frequency of certain neuropathologic abnormalities in a manner that is generally more sensitive than that found when fixed sections are used [5]. Thus, the procedure can be used to provide detailed information about changes in the histomorphometric parameters of lengthened myelinated fibers. The purpose of this study was to characterize the structural and functional response of in vivo distraction neurogenesis and to investigate the influence of the rate of distraction on changes in the conduction property and histomorphometric parameters of the peripheral nerve during limb lengthening.
0736-0266/03/$ - see front matter 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 2 ) 0 0 1 0 0 - 6
A. Yokota et al. I Journal of Orthopaedic Reseurch 21 (2003) 36-43
Materials and methods
37
lyzer system was then used to evaluate these digitized histologic images.
Study design
Tran.sverse sections Small cubes (approximately 1 mm in length) of the specimens were prefixed by immersing in the same fixative as described above for 2 h a t 4 "C. The specimens were washed in several changes of buffer and postfixed by immersion in 1% Os04 in 0.1 M phosphate buffer, p H 7.4. for 2 h at 4 "C. After dehydration in serial ethanol solutions and propylene oxide, specimens were embedded in Epon 812. Transverse sections of thickness 1 pm were stained with 1% toluidine blue. The diameters of all myelinated fibers were determined. Ultrathin sections of the nerve were cut with a diamond knife (Diatome, Bienne, Switzerland) on an ultramicrotome (OM u2; Reichart, Leica Scientific Instruments, Nussloch, Germany). Then these sections were double stained with uranyl acetate and lead citrate and were examined by electron microscopy (H-7100: Hitachi, Tokyo, Japan).
This study was performed o n 18 skeletally mature male Japanese white rabbits (JAPAN SLC Inc., HamdmdtSu, Japan) with a body weight of 3.0-3.5 kg. The rabbits were randomly separated into two test groups. with nine animals in each group. The right femora were lengthened by set amounts once per day over the course of the study to give a total lengthening of 30 mm (approximately 30'!4) of the total length of the femur). A distraction rate of 0.5 mm per day (Group 1) or 2.0 mm per day (Group 2) from the first day was used because pilot studies had indicated that early union of the osteotomy site could occur in Group 1 when lengthening began after a waiting period of several days. Full weight-bearing was allowed immediately after surgery. All analyses were performed immediately after completion of the limb lengthening procedure. We planned to harvest 10 rabbits for electrophysiological analysis and the remaining eight rabbits for histological and histomorphometric analysis. The protocol used was approved by the Osaka Medical College animal care and use committee.
Nerae teasing We measured the distance between the nodes of Ranvier (internodal length; Fig. 1) of myelinated fibers to clarify whether proliferation of Schwann cells occurs o r each Schwann cell is lengthened. After a 1.5-h post-fixation period in 1%) Os04 in 0.1 M phosphate buffer, bundles were passed through 450/11, 66%, and 100% glycerin for 24 h at each concentration and at a temperature of 37 "C. Under a surgical microscope, single myelinated fibers were teased apart with fine forceps. Care was taken to obtain the fibers in a random fashion. Four hundred fibers were taken from the two test groups and the control group (a total of 1200 fibers), and both their diameter and internodal length were measured. The diameter was classified as the mean value of three counts made at different parts of each internode excluding the paranode (Fig. 1). The internodal length was classified according to the length of the whole internode including the paranode. We observed the geometry of the paranode- node-paranode (PNP) region [24] (Fig. I). This region was classified into one of three types: (A) roundthe shape of the paranodal myelin sheath keeps its physiological curvature, and the nodes of Ranvier d o not widen; (B) straight-the shape of the paranodal myelin sheath straightens, but the nodes of Ranvier d o not widen; (C) wide-the shape of the paranodal myelin sheath straightens, and the nodes of Ranvier widen (Fig. 2A-C). Morphologic changes in the internode were also evaluated.
Surgical procedure With the rabbit under intravenous anesthesia with sodium pentobarbital (20 mg/kg body weight), the right femur was explored through a lateral approach and a muscle-splitting technique. A unilateral external fixatordistractor (M101; Orthofix, Verona, Italy) was applied to the lateral aspect of the femur under aseptic technique. With use of a reciprocating saw under irrigation with saline solution, the femur was osteotomized at the mid-section, through the gap between the middle two transfixing pins. The bone was completely covered with local muscle and fascia and a tension-free skin closure was performed. Electrol,liysiological analysis We assessed nerve function by measuring the compound nerve action potential (CNAP) of the sciatic nerve with use of an electromyography device (DANTEC Neuromatic-2000; Medtronic Functional Diagnotics AIS, Skovlunde, Denmark). After the animals were anesthetized with sodium pentobarbital, the sciatic nerve and proximal part of the tibial nerve were explored through a lateral approach. Two stainless steel bipolar electrodes (UM2-5050; Unique Medical Inc., Osaka, Japan) were used: one for stimulating the nerve supramaximally at the point where the tibial nerve entered the gastrocnemius and the other for recording the CNAP at the sciatic notch. The electrical stimulation used was a 10-Hz, 0.1-ms rectangular wave. Room temperature was maintained at 24 "C. We recorded and averaged 20 responses and measured peak amplitude and conduction velocity by measuring the distance between the stimulating and recording electrodes. The same procedure was performed on the contralateral untreated side. Conduction velocity and peak amplitude in these recordings were expressed as a percentage of the values obtained for the contralateral side of each animal. Histological and Izistonzorpliometric analysis After the animals were anesthetized with sodium pentobarbital, they were perfused through the heart with 3 I of fixative solution containing 2.5% glutaraldehyde and 2%) paraformaldehyde in 0. I M phosphate buffer, p H 7.4. This technique let us maintain the in situ length of the elongated nerve. A 10-mm long sample of the sciatic nerve was dissected out at the level of the mid-thigh (osteotomy). The tibial bundle was isolated from the fibular bundle and divided into two segments, one for transverse sectioning and the other for teased fiber analysis. Four specimens from the contralateral side were randomly chosen as a control group. Image capture analysis of the finished sections was performed with use of light microscopy (VANOX-T; OLYMPUS, Tokyo, Japan) in conjunction with a C C D video camera (DXC-950; Sony, Tokyo, Japan). All quantitative histomorphometric parameters were obtained through use of a semiautomatic image analyzer system (MCID version 3.0; Imaging Research Inc., Ont., Canada). Microscopy images of the histologic specimens were captured by camera and then digitally transferred to a computer. The image ana-
Statistical analysis Results are presented as mean values and the standard deviation. An unpaired t-test (two-tailed) was used to compare the percentage value of conduction velocity and peak amplitude of both test groups. A paired t-test was used to compare the experimental and control sides with regard to conduction velocity and peak amplitude. Analysis of variance (ANOVA) with Scheffe's post-hoc t-test was used to compare quantitative, histomorphometric parameters of each group. Chisquare test with Yates continuity correction was used to compare the
.
.
i
T
I1
.
. t
t
NOR
NOR
~
PN . _PN
~
Fig. 1. A schematic drawing of the teased myelinated nerve fiber to show the terminology used. IL; internodal length, NOR; node of Ranvier, PNP; pafanode-node-paranode region, PN; paranode, MS; myelin sheath, Ax; axon.
38
A. Yokota el
(11.
I Journal oj Orthopaedic Research 21 (2003) 36-43
Electrophysiological analysis The results of Group 1 showed mild changes in response, averaging an 18% delay in conduction velocity and a 20% reduction in amplitude in comparison with results of the contralateral side ( P < 0.001 and P < 0.009, respectively). The results of Group 2 showed an average 23% delay in conduction velocity and 71% reduction in amplitude in comparison with those of the contralateral side ( P < 0.008 and P < 0.004, respectively). The percentage of conduction velocity for the elongated sciatic nerve was similar for the two test groups (P = 0.29; Table 1). However, the percentage of amplitude in Group 2 was significantly lower than that found for Group 1 (P < 0.0001; Table 2). Histological and histornorphometric analysis Transverse sections The mean diameter of myelinated fibers is shown in Table 3. Myelinated fibers in Group 2 significantly narrowed in comparison to those of the control group and Group 1 (P < 0.OOOl). However, the mean myelinated fiber diameter did not differ between the control group and Group 1 ( P = 0.42). Electron microscopy revealed no evidence of degeneration, fibrosis, or endoneurial edema in the control group or Group 1 (Fig. 3A-D). Thinning of the myelin sheath of myelinated fibers was not observed in either test group (Fig. 3C and E), but a degenerative change of unmyelinated axons occurred sporadically in two cases in Group 2 (Fig. 3F).
Fig. 2. Classification of the shape of the PNP region of teased myelinated fibers: (A) round (B) straight (C) wide. Arrowheads indicate straightened myelin sheath. Arrow indicates the widening of the node of Ranvier. Original magnification 160x .
categorical data of geometry of the PNP region of each group. Values of P < 0.05 were considered to be significant.
Results
One animal from each test group had to be substituted, one because of femoral fracture and one because of respiratory infection (drop-out rate of 1 IYn). The remaining 16 rabbits showed good tolerance to the lengthening procedure. Therefore, eight rabbits were used for electrophysiological analysis and another eight were used for histomorphometric analysis. Pin loosening, infection, or soft-tissue related problems did not develop in any animal. The external fixation devices caused no additional injuries.
Teased jibers The mean diameters and internodal lengths of teased fibers for each group are shown in Table 3 . There was no difference between the mean diameters of each group; however, the internodal length of both test groups was significantly longer than that of the control group ( P < 0.0001). The percentage of elongation of internodes compared to the control group was 22.1% in Group 1 and 20.7% in Group 2. There was no evidence of segmental demyelination, remyelination, or dilatation of the Schmidt-Lantermann incisures (Fig. 4A-C). A straight type of PNP region was observed in 3.3% of PNP regions in the control group, 12.1%)in Group 1, and 21.6% in Group 2 thus a tendency toward straightening in the geometry of this region was significantly correlated with the rate of distraction (P < 0.0001, Table 3).
Discussion Little is known about how the peripheral nerve adapts to gradual elongation since only a few reports are available on histomorphometric analysis in peripheral nerves during gradual limb lengthening [1,7,16]. In the
A. Yokota et al. / Journal of Orthopaedic Reseurcl~21 (2003) 36-43
39
Table 1 Conduction velocity of the sciatic nerve Group 2
Group 1 Rabbit 1
2 3 4
Treatment (mls)
Contralateral (mls)
Treatmenticontralateral
59.8 64.2 69.2 66.3
77.4 77.4 83.3 79.4
77.3 82.9 83.1 83.5
64.9 i4.0
79.4 i2.8
81.7
Rabbit
WU)
1 2 3 4
* 2.9
Treatment (m/s)
Contralateral (m/s)
(%)
Treatmentkontralateral
51.4 70.0 70.5 53.5
71.4 86.6 82.2 78.5
72.0 80.8 85.8 68.2
61.4 zt 10.3
79.7 f 6.4
76.7 f 8.0
Treatment (PV)
Contralateral (FV)
Treatmentkontralaterdl
70 124 52 100
264 400 216 292
26.5 31.0 24.1 34.2
94 f 31.9
293 i 77.9
29.0 f 4.5"
The date in the bottom row are mean f SD
Table 2 Peak amplitude of the sciatic nerve
Group 1 Rabbit 1
2 3
4
Group 2 Treatment
Contralateral (UV)
Treatmentlcontralateral
(WV)
240 208 204 336
292 280 244 420
82.2 74.3 83.6 80.0
247 f 61.5
309 f 76.8
80.0 zt 4.1
Rabbit
(%)
1 2 3 4
(%)
The data in the bottom row are mean f SD. " P < 0.0001 compared with Group 1
Table 3 Quantitative, histomorphometric parameters of each specimen Group I ( n = 4)
Group 2 ( n = 4)
7.528
6.877"
7136 i421
7753 f 344
11.8 f 3.9
12.4 i 3.9
12.2 i4.0
1216 f 295
1484 i 347h
1467 =?z 322'
Control ( n = 4) Traiwwrsr .section Mean fiber diam- 7.606 eter (pm) Total myelinated 7504 & 306 fibers (no.)
Tiused ,fiber Mean fiber diameter (pm) Mean internodal length (pm)
The geometry of the PNP region 95.6 86.0 Round ('%) 76.9 ( n = 61 5)' ( n = 765) ( n = 688)d Straight (%) 3.3 12.1 21.6 ( n = 173)' (TI = 97)d ( P I = 26) Wide (YO) 1.1 1.9 1.5 ( n = 12) ( n = 15) in = 9) " P < 0.0001 compared with Group 1 and the control group. b f < 0.0001 compared with the control group, and P = 0.75 compared with Group 2. ' P < 0.0001 compared with the control group. f < 0.0001 compared with Group 2 and the control group. ' P < 0.0001 compared with Group 1 and the control group.
current study, the internodal length of myelinated fibers of both test groups was significantly longer than that of
the control group. In addition, the diameter of myelinated fibers of Group 1 (distraction rate 0.5 mm per day) was similar to that of the untreated contralateral side. These results differ from those of acute elongation studies in which elongated nerve fibers become narrower [lo], thereby suggesting that some nerve fiber grows. However, if proliferation of Schwann cells occurs in stretched nerve fibers, then the internodal length should be shorter and the myelin sheath should become thinner [2]. Ultrastructual examination of the current study disclosed no evidence of thinning of the myelin sheath in either test group. Thus, the results of the current study primarily indicate that in vivo distraction neurogenesis is achieved not by means of proliferation of Schwann cells, but by the elongation of each Schwann cell body itself. In the study of Battiston et al. [I], the increase in internodal length was not as much as the total amount of lengthening of the nerve. They therefore postulated that Schwann cell proliferation occurs in addition to a lengthening of the myelinated fibers. Elongated nerve fibers, however, usually lose their wavy structure and become straight [20]. In addition, bowstringing of the nerve across the joints could occur during limb lengthening [21]. These alterations could compensate for some amount of lengthening that the nerve undergoes. In the current study, we did not measure the total amount of lengthening of the nerve by means of epineural marking
40
A . Yoliolu ('I (11. I Joirrnrrl of' Orthopuerlic~Resrrrrcli 21 (2003) 3 6 ~43
(E)
(F)
Fig. 3 . Electron micrographs of transverse section of sciatic nerves in low magnitude (A, C. E, 1OOOx. the scale bar indicates 5 pm) and high magnitude (B. D. F:, 12.500 x , the scale bar indicates 0.5 pm): (A, B) control; (C, D) Group I ; (E, F) Group 2. No degenerative changes were seen in Group 1. In Group 2, one axon in a cluster of unmyelinated fiber has a watery appearance (F, *). indicating axonal degeneration. But many other unmyelinated fibers appear normal.
suture because the exploration at the first procedure would cause adhesion around the nerve. The quotient of
increase of the internode was approximately 70-75% of that of bone lengthening.
A . Yokotci et ul. I Journul of' Orthopuedic Re~c.ar.ch21 12003) 36 43
41
Fig. 4. Consecutive lengths of teased myelinated fibers: (A) control. ( R ) Group I . (C) Group 2. The scale bar indicates 200 pin. Arrows indicate the nodes of Ranvier. Note the internodal lengths of Group 1 and Group 2 fibers are longer than that of the control group. Original magnification 16x
Histological study of transverse sections of the current study indicated only modest morphologic changes in the peripheral nerves after gradual limb lengthening. N o evidence of degenerative changes was observed in Group 1. In two cases in Group 2 (distraction rate 2.0 mm per day), mild degeneration of unmyelinated axons [4] was shown, a finding similar to the observations by Fink et al. [7] in canine tibia1 and peroneal nerves. Previous histological studies yield contradictory results. Some studies concluded no degenerative changes [1,9, 1 I], whereas other studies observed clear-cut degeneration after gradual limb lengthening [14,20,22]. The difference between these results was probably caused by differences in the amount or rate of distraction. lppolito et al. [14] and Skoulis et al. [22] have shown a clear relationship between the amount or rate of distraction and the degree of degeneration. In addition, the differences between species and the difference in vulnerability of each peripheral nerve to the distraction force [l8] could affect these results. In an electrophysiological analysis of the current study, we recorded the CNAP orthodromically. The clinical significance of this method is identical to that of a sensory nerve conduction study [19], which has been reported to be a more sensitive index to detect subtle conduction slowing than motor nerve conduction [17]. The results showed that the conduction property of the nerves was affected by the rate of lengthening. The results of Group 1 showed mild conduction slowing; on the other hand, the results of Group 2 showed
an apparent conduction block, with significant diminution of the amplitude in the presence of mild slowing of conduction velocity. This result was similar to results of other studies [I 1,20,22,23]. Shibukawa et al. [20] concluded that reduction of amplitude is a more sensitive predictor of nerve dysfunction than delay of velocity during limb lengthening. Because the PNP region is responsible for impulse conduction [24], structural alterations of this region would result in the change of the conduction property of the nerve. In the current study, straightening of the geometry of the paranodal myelin sheath was significantly correlated with the rate of distraction. We assumed that geometry of this area would reflect the capability of each Schwann cell to lengthen its body against a given rate of distraction. In addition, the mean diameter of myelinated fibers of Group 2 was significantly decreased compared to that of Group I and the control group. The hypothesis about the development of nerve impairment supported by these findings is as follows: (1) If the rate of distraction is slow, then the Schwann cells can adapt to elongation, which maintains the geometry of the PNP regions. As a result, nerve dysfunction does not manifest (Fig. 5B). (2) If the rate of distraction is too rapid, then the Schwann cells can no longer adapt to elongation, and straightening of the paranodal myelin sheath results. In addition, the nerve fibers become thinner under the tension, like a stretched elastic band (Fig. 5C). As a result, nerve function is impaired.
A. Yokota et 01. I Journal of Orthopciedic Research 21 (2003) 3 6 4 3
42
a
0
A t NOR
t NOR
a‘
aa’
b
0
B b’
bb’
C
0
C
cc’ C’
Fig. 5. A schematic drawing of the longitudinal (left) and transverse (right) sections of the myelinated nerve fiber: (A) normal, (B) after slow elongation, (C) after rapid elongation. LoS: length of the Schwann cell; NOR: node of Ranvier. Broken lines indicate the transection line. Note the geometry of the paranode becomes straight and the diameter of fiber decreases after rapid elongation.
The current study was limited to assessing the gross geometry of the PNP region by means of teased fibers. The study could not quantify the gap length at the nodes of Ranvier although its procedure enabled us to observe numerous samples in each group. Ikeda et al. [1 11 attained an elongation of 300/0 in rabbit sciatic nerves with the daily rate of 4.0 mm. Kern et al. [I61 obtained a 20Y0elongation of rabbit tibias with the rate of 0.7 mm per day. Both of them demonstrated widening of the nodes of Ranvier with use of longitudinal thin sections from the fixed specimens. The straight type of PNP region in our classification for teased nerve fibers might have some increase in the nodal gap. In conclusion, the current study suggests that if lengthening is performed slowly enough as not to manifest nerve dysfunction, then peripheral nerves can accommodate to the elongated new length with appropriate elongation of the internodal length and preservation of the diameter of the nerve fibers. This indicates that distraction neurogenesis is achieved by the lengthening of each Schwann cell body, and not by the proliferation of Schwann cells.
Acknowledgements The authors would like to thank Toshio Hasegawa, M.D., Ph.D., for advice and encouragement throughout the current study. The authors also thank Tomoyuki Tsujimura, M.D., Taisuke Horiguchi, M.D., and Mr. Tsuguya Naito for technical assistance. The valuable comments made by Nozomu Inoue M.D., Ph.D., were greatly appreciated. The authors have received nothing
of value from a private or commercial entity for the data described in this study. References Battiston B, Buffori P, Vigasio A, et al. The effects of lengthening of nerves. Ital J Orthop Traumat 1992;18:79-86. Cragg BG, Thomas PK. The conduction velocity of regenerated peripheral nerve fibers. J Physiol 1964;171:164-75. De Bastiani G, Aldegheri R , Renji-Brivio L, Trivella G. Limb lengthening by callus distraction (callotasis). J Pediatr Orthop 1987;7: 129-34. Dyck PJ, Hopkins AP. Electron microscopic observations on degeneration and regeneration of unmyelinated fibers. Brain 1972;95:223-34. Dyck PJ, Giannini C , Lais A. Pathologic alterations of nerves. In: Dyck PJ, Thomas PK, Griffin JW, et al., editors. Peripheral neuropathy. 3rd ed. Philadelphia: WB Saunders; 1993. Faber FWM, Keessen W, van Roermund PM. Complications of leg lengthening: 46 procedures in 28 patients. Acta Orthop Scand 1991;62:327-32. Fink B, Neuen-Jacob E, Lehmann J. et al. Changes in canine peripheral nerves during experimental callus distraction. Clin Orthop 2000;376:252--67. Galardi G, Conii G, Lozza L, et al. Peripheral nerve damage during limb lengthening. J Bone Joint Surg [Br] 1990;72:1214. Gil-Albarova J, Melgosa M , Gil-Albarova 0. Canadell J. Soft tissue behavior during limb lengthening: an experimental study in lambs. J Pediatr Orthop B 1997;6:266-73. Haftek J, Poland W. Stretch injury of peripheral nerve. J Bone Joint Surg [Br] 1970;52:354--65. lkeda K, Tomita K , Tanaka S. Experimental study of peripheral nerve injury during gradual limb elongation. Hand Surg 2000; 5:41-7. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues. Part I. The influence of stability of fixation and soft-tissue preservation. Clin Orthop 1989;238:249- 8 I . Ilizarov GA. Clinical application of the tension-stress effect for limb lengthening. Clin Orthop 1990;2S0:8 -26. Tppolito E, Peretti G , Bellocci M, et al. Histology and ultrastructure of arteries, veins, and peripheral nerves during limb lengthening. Clin Orthop 1994;308:5&62. Kawamura B, Hosono S. Takahashi T, et al. Limb lengthening by means of subcutaneous osteotomy: experimental and clinical studies. J Bone Joint Surg [Am] 1968;50:851-78. Kerns J, Urabe T, Bleasdale T, et al. Effects of gradual bone lengthening on the rabbit tibia1 nerve. Scand J Plast Reconstr Hand Surg 2001;35:361-8. Ludin HP, Lutschg J, Valsangiacomo F. Vergleichende untersuchung orthodromer und antidromer sensibler nervenleitgeschwindigkeiten: 2. befunde bei polyneuropathien und bei status nach polyradikulitis. Z EEG E N G 1977;8:180-6. Makarov M R , Birch JG, Delgado M R , et al. Effects of external fixation and limb lengthening on peripheral nerve function. Clin Orthop 1996;329:310-6. Oh SJ. General concepts of electrodiagnostic studies in neuromuscular disease. In: Oh SJ, editor. Clinical electromyography: nerve conduction studies. 2nd ed. Baltimore: Williams & Wilkins; 1993. Shibukawa M, Shirai Y. Experimental study on slow-speed elongation injury of the peripheral nerve: electrophysiological and histological changes. J Orthop Sci 2001;6:262-8. Simpson AHRW, Kenwright J. The response of nerve to different rates of distraction. J Bone Joint Surg [Br] 1992:74(Suppl 3):326. Skoulis TG, Vekris MD, Terzis JK. Effect of distraction osteogenesis on the peripheral nerve: experimental study in the rat. J Reconstr Microsurg 1998;14:565-74.
A Yokota et rrl. I Journul of Orthopedic Reseurclz 21 (2003) 36-43
[23] Strong M, Hruska J, Czyrny J, et al. Nerve palsy during femoral lengthening: MRI, electrical, and histologic findings in the central and peripheral nervous systems-a canine model. J Pediatr Orthop 1994;14:347-51.
43
[24] Thomas PK, Berthold C-H, Ochoca J. Microscopic anatomy of the peripheral nerve system. In: Dyck PJ, Thomas PK, Griffin JW, et al., editors. Peripheral neuropathy. 3rd ed. Philadelphia: WB Saunders; 1993.