Effect of acute limb ischaemia on neuromuscular

ORIGINAL ARTICLE

Effect of Acute Limb Ischaemia on Neuromuscular Function in Rats Konstantinos P. Hatzipantelis1, Konstantinos Natsis2 and Maria Albani3 From the 1 5th Department of Surgery, Division of Vascular Surgery, 2 Department of Anatomy, 3 Department of Physiology, Medical School, Aristotle University of Thessaloniki, Greece

Eur J Surg 2001; 167: 831–838 ABSTRACT Objective: To locate the exact site of the primary lesion in the neuromuscular system in acutely ischaemic extremities. Design: Experimental study. Setting: University hospital, Greece. Animals: 22 adult rats. Interventions: Isometric tensions of extensor digitorum longus muscles were recorded before ischaemia and every 5 minutes after the arterial occlusions by indirect stimulation. When no contractile activity was elicited, the muscle was stimulated directly and recordings made every 5 minutes. The sciatic nerve function was checked by recordings of nerve conduction velocity. Specimens from the muscles were examined under electron microscopy. Main outcome measures: Muscle contractile properties, conduction velocity, and electron microscopic appearance. Results: After a period of about 50 minutes neuromuscular function under indirect stimulation in the ischaemic limbs was lost, whilst under direct stimulation the extensor digitorum longus muscles and the sciatic nerves still functioned. Electron microscopic study showed distinct alterations at the neuromuscular junctions. Conclusions: The response of the neuromuscular system to acute ischaemia indicated that the neuromuscular junction is probably the site most susceptible to acute ischaemia. Key words: neuromuscular junction, peripheral nerve, skeletal muscle, muscle contractile properties, electron microscopy, revascularisation, embolism, thrombosis, trauma.

INTRODUCTION Acute ischaemia of a limb may be caused by embolism, thrombosis, or injury of its large arteries. Acute ischaemia of an extremity is dramatic. The mortality after revascularisation of an acutely ischaemic limb is estimated to be between 15%–48%, with an amputation rate of about 20% (2). This may be related to ischaemia —reperfusion injury rather than the consequence of direct ischaemic damage (21). In addition, adequately revascularised extremities may still develop considerable functional decits. In recent years, new discoveries have led to a better understanding of the mechanisms that contribute to injury during reperfusion, but substantial fewer data have been presented about the primary injury of the peripheral nerve—neuromuscular junction—skeletal muscle during the preceding ischaemic interval. Dyck et al. and Korthals et al. consider that skeletal muscle is more vulnerable to ischaemic injury than nerve (4, 15). Chervu et al. and Hoch et al. suggested that peripheral nerve is more susceptible to ischaemia than skeletal muscle (3, 10). Lundborg noted that the neuromuscular junction is the most susceptible site of the neuromuscular system to ischaemia (19). The fact Ó 2001 Taylor & Francis. ISSN 1102–4151

is that little is known about the inuence of ischaemia on the neuromuscular junction. Our experimental animal model using the electrophysiological method can check simultaneously the neuromuscular function as a whole as well as the function of the nerves and skeletal muscles individually under conditions of acute ischaemia. The purpose of this study was to nd out the exact site of the primary lesion in the neuromuscular system in acutely ischaemic extremities, knowledge that may lead to an improved therapeutic approach to the disease.

MATERIAL AND METHODS Twenty-two male and female Wistar rats, 2–3 months old, were divided into two groups of fteen and seven rats, respectively. The second group served as controls. The rats were anaesthetised with a 4.5% w/v solution of chloral hydrate 1 ml/100 g body weight given intraperitoneally. In the rst (experimental) group the aorta below the renal arteries and both common iliac arteries were mobilised, through a midline abdominal incision, while Eur J Surg 167

832

K. P. Hatzipantelis et al.

Fig. 2. A. Isometric tetanic tensions recorded from a normal extensor digitorum longus muscle under indirect stimulation. B. Isometric tetanic tension recorded from the ischaemic muscle in the same rat under direct stimulation, because the muscle was not responding to indirect stimulation. Calibration bars in both recordings equal to 55 g.

Fig. 1. A–D. Action potentials recorded from the sciatic nerve and isometric twitch tensions recorded from the extensor digitorum longus muscle of a single rat in the experimental group, under indirect stimulation, as the ischaemic time increases. Calibration bars in all recordings equal to 55 g. E. Isometric twitch tension recorded from the same muscle of the same rat under direct stimulation, because the muscle did not respond to indirect stimulation.

both iliolumbar arteries, both inferior epigastric arteries, and the inferior mesenteric artery were ligated. The venous drainage was left intact. A longitudinal incision was made over the distal tendon of the right extensor digitorum longus muscle and the tendon was detached distally. Through a right Eur J Surg 167

posterior thigh incision the sciatic nerve was dissected from its bed. The rat was then placed on another operating board and the right lower limb was immobilised, the distal end of the muscle was attached to a force transducer, with a 3/0 silk suture. Through an oscilloscope (Textronix 5115) the force transducer then recorded the twitch and tetanic tensions of the muscle. Two electrodes, one stimulating and one recording, were also placed along the sciatic nerve 1 cm apart. The recording electrode was connected to the same oscilloscope and placed at the end of the sciatic nerve, while the stimulating electrode was placed 1 cm proximally and compound nerve action potentials were recorded. The sciatic nerve and the muscle were moistened with Krebs solution at 30–35°C, which was anaerobic and contained no glucose (16). Under physiological conditions (time 0) the compound nerve action potentials of the sciatic nerve and the twitch and tetanic tensions of the extensor digitorum longus muscle were recorded after indirect and direct stimulation. The abdominal aorta and both iliac arteries were then ligated, the abdominal incision was closed, and the above recordings (except those recorded by direct stimulation of the muscle) were measured every 5 minutes for approximately 50 minutes. The control group was prepared similarly but without interruption of the blood ow. When no contractile activity was elicited by nerve stimulation, the muscle was stimulated directly and the same experimental procedure was followed (Figs 1 and 2). In the control group recordings after direct stimulation of the muscle started 110 minutes after the onset of the experiment. The contractile properties and conduction velocities measured in this study are as follows: optimal muscle length = the muscle length at which maximal isometric twitch tension occurs; twitch tension = the peak tension developed during a twitch at optimal muscle length; time peak tension = time from the end of the latent period to the peak of the isometric tension; half relaxation time = time of tension decay from peak of

Neuromuscular function after acute ischaemia

Fig. 3. Figures from the radionuclide study in rats. A) ischaemic animal B) control animal.

isometric twitch to one half this value; and tetanic tension = the maximum tension at the optimal stimulating frequency (80 or 100 Hz for extensor digitorum longus). The conduction velocity was measured by the equation V = S/T, where S = 1 cm and T = latency time. Latency time was measured from the stimulus artefact to the point where the negative deection crossed the isopotential line (19), so the ratio of conduction velocity was V = Vt /Vo = T0/Tt. Isometric forces and conduction velocities were expressed as a ratio of the initial values obtained

833

before the induction of ischaemia. Time peak tension and half relaxation time values, which corresponded to 0 g twitch tensions, were not calculated, because this was impossible. At the end of the experiment the muscle was harvested and weighed, and specimens from both groups were xed with 2.5% glutaraldehyde, pH 7.4, for 1 hour, postxed with 1% osmium tetroxide, dehydrated through graded ethanols, cleared in propylene oxide, and embedded in Epon 812 resin. Ultrathin sections were cut, retrieved on to copper grids, stained with uranyl acetate and lead citrate and examined with a Zeiss 9S-2 transmission electron microscope. Representative micrographs were taken of each specimen. Afterwords each animal was killed by intracardiac infusion of a 4.5% w/v solution of chloral hydrate. The completeness of the arterial occlusion was checked by means of radionuclide imaging. 7.4 £ 109 Bq of 99mTc were injected through the descending aorta and then an image was taken (500 000 counts), enough to show lack of radioactivity in the lower limbs and pelvic area. In control animals both lower limbs and the pelvic area were also clearly depicted (Fig. 3). Data from between groups and within groups were compared using Student’s two-tailed t test for unpaired and paired data. All results are presented as mean (SEM). The Kruskal–Wallis one way analysis of variance on ranks and pairwise multiple comparison procedures (Student–Newman–Keuls and Dunn’s)

Table I. Mean (SEM) twitch tension expressed as a ratio of the initial value (time 0) in the two groups, under direct stimulation of the extensor digitorum longus muscle Experimental group (n = 15)

Control group (n = 7)

Time (min)

Group value

Time (min)

Group value

p value Mann-Whitney

0 55.7 (4.3) 60.7 (4.3) 65.7 (4.3)

1 (0) 0.2 (0.04)* 0.2 (0.04)* 0.2 (0.03)*

0 110 115 120

1 (0) 0.7 (0.09) 0.7 (0.1) 0.8 (0.1)

0.001 0.0006 0.0005

* p < 0.05 compared with controls (t test).

Table II. Mean (SEM) twitch tension expressed as a ratio of the initial value (time 0) in the two groups, under indirect stimulation of the extensor digitorum longus muscle Time (min)

Experimental group (n = 15)



0 5 20 35 50 65

1 (0) 0.8 (0.03)* 0.6 (0.06)* 0.4 (0.09)* 0.2 (0.06)* 0*

1 (0) 1 (0.08) 1 (0.07) 0.9 (0.09) 0.8 (0.07) 0.8 (0.08)

0.165 0.0043 0.0019 0.0004 0.0005

* p < 0.05 compared with controls (t test). Eur J Surg 167

834


Table III. Mean (SEM) tetanic tension expressed as a ratio of the initial value (time 0) in the two groups, under indirect stimulation of the extensor digitorum longus muscle Time (min)




0 5 20 35 50 65

1 (0) 0.8 (0.05)* 0.6 (0.07)* 0.3 (0.09)* 0.1 (0.07)* 0*

1 (0) 1 (0.06) 0.9 (0.06) 0.9 (0.09) 0.9 (0.08) 0.9 (0.07)

0.0032 0.009 0.0017 0.0005 0.0002


Table IV. Mean (SEM) tetanic tension expressed as a ratio of the initial value (time 0) in the two groups, under direct stimulation of the extensor digitorum longus muscle Experimental group (n = 15)


Time (min)

Group (n = 15)

Time (min)

Group (n = 7)


0 55.7 (4.3) 60.7 (4.3) 65.7 (4.3)

1 (0) 0.2 (0.11) 0.2 (0.09)* 0.2 (0.04)*

0 110 115 120

1 (0) 0.9 (0.02) 0.8 (0.06) 0.9 (0.06)

0.121 0.014 0.0002


Table V. Mean (SEM) time peak tension expressed as a ratio of the initial value (time 0) in the two groups, under indirect stimulation of the extensor digitorum longus muscle Time (min)




0 5 20 35 50

1 (0) 0.8 (0.04)* 0.8 (0.3)* 0.8 (0.05)* 0.8 (0.08)

1 (0) 1 (0.02) 1.1 (0.09) 1.2 (0.10) 1.1 (0.11)

0.003 0.001 0.001 0.001


were also applied to the data of the ischaemic and control muscles at different times. Comparisons between groups were made also with Mann–Whitney u test. Probabilities of less than 0.05 were accepted as signicant. RESULTS Control extensor digitorum longus: One way repeated measures ANOVA was used to analyse the data at different times (0, 5, 20, 35, 50, and 65 minutes). No value was signicantly different from any other, except the twitch tension that were recorded under direct stimulation (Table I). Nerve conduction velocity: There were no differEur J Surg 167

ences between or within the groups for the 65 minutes of the experiment. Twitch and tetanic tension: Tables I, II, III, and IV show that there were signicant differences between the groups after the 5 minutes of recordings (p < 0.05). The twitch and tetanic tensions of the muscles in the experimental group, which were recorded under indirect stimulation, ceased within 65 minutes of ischaemia at a mean time of 50.7 (4.3) minutes. After that time twitch and tetanic tensions were recorded under direct stimulation. Time peak tension: Tables V and VI show that there were signicant differences between the groups (p < 0.05), except at 50 minutes. Half relaxation time: Tables VII and VIII show that

Neuromuscular function after acute ischaemia

835

Table VI. Mean (SEM) time peak tension expressed as a ratio of the initial value (time 0) in the two groups, under direct stimulation of the extensor digitorum longus muscle Experimental group (n = 15)


Time (min)

Group values

Time (min)

Group values


0 55.7 (4.3) 60.7 (4.3) 65.7 (4.3)

1 (0) 0.8 (0.04)* 0.8 (0.04)* 0.9 (0.06)*

0 110 115 120

1 (0) 1.1 (0.07) 1.1 (0.09) 1.2 (0.14)

0.006 0.013 0.027


Table VII. Mean (SEM) half relaxation time expressed as a ratio of the initial value (time 0) in the two groups, under indirect stimulation of the extensor digitorum longus muscle Time (min)




0 5 20 35 50

1 (0) 0.8 (0.03)* 0.9 (0.07)* 1 (0.09) 1.1 (0.13)

1 (0) 1.1 (0.05) 1.2 (0.06) 1.1 (0.06) 1.1 (0.10)

0.0002 0.018 0.117 0.030


Table VIII. Mean (SEM) half relaxation time expressed as a ratio of the initial value (time 0) in the two groups, under direct stimulation of the extensor digitorum longus muscle Experimental group (n = 15)


Time (min)

Group (n = 15)

Time (min)

Group (n = 7)


0 55.7 (4.3) 60.7 (4.3) 65.7 (4.3)

1 (0) 1.2 (0.13) 1.2 (0.15) 1.3 (0.15)

0 110 115 120

1 (0) 1 (0.06) 1 (0.07) 1.1 (0.06)

0.274 0.500 0.610

there were signicant differences between the groups only at 5 and 20 minutes (p < 0.05). Electron microscopic evaluation: Fifteen minutes after neuromuscular function has ceased because of acute ischaemia, ultrastructural alterations at the neuromuscular junction were seen that were not present in controls (Fig. 4). These were distinct in mitochondrial morphology both in presynaptic and postsynaptic sites, with the mitochondria swollen, structural disorganisation, ruptured cristae, and in some cases the presence of myelin gures and clearing of their matrix. We also noticed that in the presynaptic area the acetylcholine vesicles were of different size and density and were clustered (Fig. 5). In some animals the changes were more prominent with vacuolation of mitochondria in the presynaptic area and disorganisation of the postsynaptic junctional folds (Fig. 6).

DISCUSSION In our study the extensor digitorum longus muscle and the peripheral end of the sciatic nerve were chosen for investigation. The reason was that both of these structures are more vulnerable to ischaemia than other muscles or nerves in the hindlimb of the experimental animal (5, 12, 13, 15, 25, 28). In vivo models have been described of whole hindlimb ischaemia using tourniquets and selective arterial occlusion. We chose selective arterial occlusion because tourniquets could potentially cause direct muscle and neural damage from mechanical compression and venous occlusion (22, 24, 26). Effect of acute ischaemia on peripheral nerve function In the present study we found that acute ischaemia of Eur J Surg 167

836


Fig. 4. Electron micrograph of normal neuromuscular junction. At the left is the nerve terminal (arrow heads). Junctional sarcoplasm (asterisk) contains glycogen granules, ribosomes, and normal mitochondria (large arrows). The small arrow heads mark the synaptic clefts and folds. Note the normal appearance of the sarcomeres (original magnication £19 720).

65 minutes duration did not inuence the function of the sciatic nerve. Zollman et al. found that in vivo ischaemia of the sciatic–tibial nerve of SpragueDawley rats for up to 3 hours was not sufcient to extinguish the nerve action potential. In the same study they found that acute ischaemia for one hour did not inuence the concentrations of ATP and lactate in the sciatic-tibial nerve (33). They stated, therefore, that the maintenance of sciatic–tibial nerve function under condition of acute ischaemia is attributed to the maintenance of the ATP concentrations. This is in accordance with the knowledge that the nerve function (impulse transmission) is axoplasmic ow-dependent and axoplasmic ow is energy-dependent (4, 27). Effect of acute ischaemia on skeletal muscle function We found that acute ischaemia of 65.7 (4.3) minutes duration (maximum 80 minutes) reduced the twitch and tetanic tension (obtained under direct stimulation of the muscle) by over 80%, while time peak tension was reduced and the half relaxation time remained unchanged. This indicates that the ischaemic muscle was weaker but faster than the normal one. The possible mechanism of this is that acute ischaemia produces acidosis, because of increased lactate production (7). The increase in hydrogen ions reduces the time that calcium channels remain open. These calcium channels are located on the membrane of the sarcoplasmic reticulum, so there is a reduction in the release of calcium from the sarcoplasmic reticulum Eur J Surg 167

Fig. 5. Transmission electron micrograph of a neuromuscular junction, 15 minutes after its function ceased because of acute ischaemia. The postsynaptic site contains enlarged, disrupted mitochondria (asterisk) with myelin gures (white arrow). The presynaptic site shows swollen mitochondria, acetylcholine vesicles of different size and density (open star), clustered in zones in some areas (arrow heads) (original magnication £24 000).

Fig. 6. Transmission electron micrograph of a degenerative neuromuscular junction, 15 minutes after its function ceased because of acute ischaemia. In the presynaptic area note a big vacuole (V) and the lack of acetylcholine vesicles. In the postsynaptic area note the clustering of the synaptic vesicles in zones facing the crests of the disorganised junctional folds (arrow heads) (original magnication £24 000).

Neuromuscular function after acute ischaemia to the sarcoplasm; this means that the time peak tension is reduced and the muscle becomes faster. Hydrogen ions compete with calcium ions in the sarcoplasm in their attempts to be bound to troponin C before the formation of the actin-myosin link (11, 14). The energy for muscle contraction comes from ATP hydrolysis. Because the concentrations of ATP in the skeletal muscles after 80 minutes of ischaemia remain stable, it seems that the reduction in twitch and tetanic tension comes from the reduction of the amount of calcium that is bound up to troponin C (7, 32). As the ischaemic time increases, acidosis causes dysfunction of the calcium pump, which is a Ca–ATPase, and reduces the time of release of calcium from troponin C because the amount of troponin C that binds up calcium is low (8, 30). Half relaxation time therefore remains stable, because this depends on the velocity of sequestration of calcium by the sarcoplasmic reticulum (17). Effect of acute ischaemia on neuromuscular function When we stimulate the sciatic nerve and take recordings from the extensor digitorum longus muscle under conditions of acute ischaemia, we simultaneously check the function of the peripheral nerve, the neuromuscular junction, and the skeletal muscle. As the time of ischaemia increases there is a gradual reduction in twitch and tetanic tension. After 50.7 (4.3) min of ischaemia, the muscle stops functioning under indirect stimulation. This is probably the result of ischaemia at the neuromuscular junction and in skeletal muscle, as for this duration of ischaemia the nerve function is intact. For the same period of ischaemia the sciatic nerve and the muscle are still functioning under direct stimulation. The neuromuscular junction is probably the site most susceptible to acute ischaemia (Figs 1 and 2). At the neuromuscular junction, the depolarisation of the presynaptic membrane by an action potential normally opens calcium channels, allowing an inux of calcium to trigger the exocytic release of acetylcholine from synaptic vesicles. The increase of calcium concentration is short-lived, because calcium-binding proteins, calcium-sequestering vesicles, and mitochondria rapidly take up the calcium that has entered the axon terminal, while calcium pumps in the membrane pump it out of the cell (18). The acetylcholine diffuses across the synaptic cleft and binds to receptor proteins in the membrane of the postsynaptic cell. The muscle cell membrane at the synapse behaves as a transducer that converts a chemical signal into an electrical one, because the acetylcholine receptor when it is bound to acetylcholine leads chiey to a large inux of sodium causing depolarisation of the membrane (9). Acetylcholine is then rapidly eliminated from the cleft by diffusion, enzymatic degradation, or by reuptake into

837

the nerve terminal. The depolarisation of the muscle cell membrane causes ion channels in the sarcoplasmic reticulum to open transiently and release calcium into the cytosol. It is the sudden increase in the cytosolic calcium concentration that causes the myobrils in the muscle cell to contract (29). Under ischaemic conditions the neuromuscular function probably stops because acidosis causes a reduction in the time that calcium channels remain open, a reduction in the number of synaptic vesicles and of their acetylcholine contents and reduction in the permeability of muscle membrane to sodium and potassium (1, 6). Electron microscopy study of the neuromuscular junction On the basis of ultrastructural evidence, the neuromuscular junction subjected to 65.7 (4.3) minutes [80 minutes maximum] of ischaemia showed distinct alterations. These changes have also been described by other investigators, but the ischaemic time was from 2 to 4 hours (20, 23, 31). Possibly the alterations of the organelles that we observed earlier in our experiments were caused by the twitch and tetanic tensions that we forced on the ischaemic muscles during the experiment. It is of great value that we noticed these alterations at the neuromuscular junction at the time when the neuromuscular function ceased because of acute ischaemia. In conclusion, it seems that the response of the neuromuscular system to acute ischaemia indicates that the neuromuscular junction is probably the site most susceptible to acute ischaemia.

REFERENCES 1. Badonic T. Electron microscopic changes of synapses after occlusion of the abdominal aorta. Folia Morphol (Praha) 1981; 29: 136–138. 2. Blaisdell FW, Steele M, Allen RE. Management of acute lower extremity arterial ischemia due to embolism and thrombosis. Surgery 1978; 84: 822–830. 3. Chervu A, Moore WS, Homsher E, Quinones-Baldrich WJ. Differential recovery of skeletal muscle and peripheral nerve function after ischemia and reperfusion. J Surg Res 1989; 47: 12–19. 4. Dyck PJ. Hypoxic neuropathy: does hypoxia play a role in diabetic neuropathy? The 1988 Robert Wartenberg Lecture. Neurology 1989; 39: 111–118. 5. Fox JL, Kenmore PI. The effect of ischemia on nerve conduction. Exp Neurol 1967; 17: 403–419. 6. Haralambie G. Importance of humoral changes to physical performance. In: J Keul, ed. Limiting factors of physical performance. Stuttgart: Thieme, 1973: 189. 7. Harris K, Walker PM, Mickle DAG, et al. Metabolic response of skeletal muscle to ischemia. Am J Physiol 1986; 250: H 213–H 220. 8. Hasselbach W, Oetliker H. Energetics and electrogeniEur J Surg 167

838

9. 10.

11.

12.

13. 14. 15. 16. 17.

18. 19.

20. 21. 22.


city of the sarcoplasmic reticulum calcium pump. Annu Rev Physiol 1983; 45: 325–339. Hille B. Ionic channels of excitable membranes. Sunderland, MA: Sinauer, 1984: 117–147. Hoch JR, Stevens RP, Keller MP, Silver D. Recovery of neuromuscular function during reperfusion of the ischemic extremity: effect of mannitol and superoxide dismutase. Surgery 1991; 110: 656–663. Hudlicka O, Pette D, Staudte H. The relation between blood ow and enzymatic activities in slow and fast muscles during development. Pugers Arch 1973; 343: 341–356. Idstrom J-P, Soussi B, Elander A, Bylund-Fellenius A-C. Purine metabolism after in vivo ischemia and reperfusion in rat skeletal muscle. Am J Physiol 1990; 258: H 1668– H 1673. Jennische E. Ischemia—induced injury in glycogen— depleted skeletal muscles: selective vulnerability of the FG-bres. Acta Physiol Scand 1985; 125: 727–734. Klisouras V. Sources of muscle energy. In: Klisouras V, ed. Ergophysiology , Athens: Kegraft E.P.E., 1986: 174. Korthals JK, Maki T, Gieron MA. Nerve and muscle vulnerability to ischemia. J Neurol Sci 1985; 71: 283– 290. Krebs HA, Henseleit K. Untersuchungen uber die harnstoffbildung im tierkorper. Z Physiol Chem 1932; 210: 33–66. Lewis DM, Al-Amood WS, Rosendorff C. Stimulation of denervated muscle: What do isometric and isotonic recordings tell us? In: Nix WA, Virbova G, eds. Electrical stimulation and neuromuscular disorders, Berlin: Springer, 1986: 101–113. Llinas R. Calcium in synaptic transmission. Sci Am 1982; 247: 56–65. Lundborg G. Ischemic nerve injury: experimental studies on intraneural microvascular pathophysiolog y and nerve function in a limb subjected to temporary circulatory arrest. Scand J Plast Reconstr Surg 1970; 6 (Suppl): 1– 113. Makitie J, Teravainen H. Ultrastructure of striated muscle of the rat after temporary ischemia. Acta Neuropathol 1977; 37: 237–245. McCord JM. Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985; 312: 159– 163. Nitz AJ, Dobner JJ, Matulionis DH. Pneumatic tourniquet application and nerve integrity: motor function and electrophysiology. Exp Neurol 1986; 94: 264–279.

Eur J Surg 167

23. Patterson S, Klenerman L. The effect of pneumatic tourniquets on the ultrastructure of skeletal muscle. J Bone Joint Surg 1979; 61 (B): 178–183. 24. Pedowitz RA, Fride`n J, Thornell AE. Skeletal muscle injury induced by a pneumatic tourniquet: an enzyme and immunohistochemical study in rabbits. J Surg Res 1992; 52: 243–250. 25. Petrasek PF, Shervanthi HV, Walker PM. Determinants of ischemic injury to skeletal muscles. J Vasc Surg 1994; 19: 623–631. 26. Petrasek PF, Walker PM. A clinically relevant small— animal model of skeletal muscle ischemia—reperfusion injury. J Invest Surg 1994; 7: 27–38. 27. Rydevic B, McLean WG, Sjoestrand J, Lundborg G. Blockage of axonal transport induced by acute, graded compression of the rabbit vagus nerve. J Neurol Neurosurg Psychiatry 1980; 43: 690–698. 28. Sladky JT, Greenberg JH, Brown MJ. Regional perfusion in normal and ischemic rat sciatic nerves. Ann Neurol 1985; 17: 191–195. 29. Stevens CF. The neuron. Sci Am 1979; 241: 48–59. 30. Takamori M, Hazama R, Tsujihata M. Active state properties of denervated and immobilized muscle: comparison with dystrophic muscle. Neurol 1978; 28: 603– 608. 31. Tountas CP, Bergman RA. Tourniquet ischemia: ultrastructural and histochemical observations of ischemic human muscle and of monkey muscle and nerve. J Hand Surg 1977; 2 (A): 31–37. 32. Walker PM. Ischemia/reperfusion injury in skeletal muscle. Ann Vasc Surg 1991; 5: 399–402. 33. Zollman PJ, Awad O, Schmelzer JD, Low PA. Effect of ischemia and reperfusion in vivo on energy metabolism of rat sciatic-tibial and caudal nerves. Exp Neurol 1991; 114: 315–320.

Submitted September 27, 2000; submitted after revision March 09, 2001; accepted June 29, 2001 Address for correspondence: K. Hatzipantelis, M.D. Kassandrou 8 GR-60100 Katerini Greece E-mail: [email protected]