Amiodarone was injected endoneurially at increasing doses into the exposed tibia1 nerve of rats to study its electrophysiologic and pathologic effects on peripheral nerve fibers. Forty-five male Wistar rats were used, and each of the following concentrations was injected into 15 nerves: 25 pg/mL, 50 pg/mL, and 100 pg/mL. Microinjection of a 25 pg/mL concentration of amiodarone resulted in a subacute, incomplete conduction block evident at day 3 postinjection. This conduction block remained stable until day 10 and recovery was complete at day 35. Microinjection of a 50 pg/mL concentration of amiodarone produced a faster evolving conduction block, and significant axon degeneration (approximately 40% of fibers). Injection of a 100 pg/mL concentration resulted in severe acute motor axon degeneration followed by complete but delayed regeneration. Results of morphological studies closely correlated with electrophysiological findings. Amiodarone thus seems to have a direct toxic effect on axons at high concentrations in the peripheral nerve, and we suggest that different pathological changes described in human amiodarone neuropathy could be related to different concentrations of the drug in the nerve, perhaps due to variability of bloodnerve barrier efficacy. Key words: amiodarone acute neuropathy demyelination axon degeneration schwannopathy MUSCLE & NERVE 151788-795 1992
AMIODARONE-INDUCED EXPERIMENTAL ACUTE NEUROPATHY IN RATS LUClO SANTORO, FABRlZlO BARBIERI, RENATO NUCCIOTTI, FRANCESCO BATTAGLIA, FRANCESCO CRISPI, M. RAGNO, P. GRECO, and G. CARUSO
Amiodarone, a diiodinated benzofuran derivative, is an a- and P-antagonist, which is effective in treatment of cardiac arrhythmia^.^^" A number of cases of peripheral neuropathy associated with amiodarone treatment have been reported since 19'74.' Based on neuropathological studies, this neuropathy has ben defined as axonal, 13,14 almost purely demyelinating,' and both axonal and demyelinating in type.' Jacobs and Costa-Jussa' have suggested that in man the Schwann-cell changes precede myelin breakdown and eventual axon degeneration. However, they failed to produce a neuropathy in
From the Department of Clinical Neurophysiology and Neurologic Clinic, Second School of Medicine, Federico It University of Naples, Naples, Italy (Drs. Santoro, Barbieri, Nucciotti, Battaglia, Crispi, and Caruso); and Department of Pharmaceutical Chemistry. Faculty of Pharmacy, Federico II University of Naples, Naples, Italy Address reprint requests lo Lucio Santoro, MD, Cattedra di Neurofisiopatologia II Facolta di Medicina e Chirurgia, UniversitA degli Studi di Napoli Federico 11, via Sergio Pansini 5, 80131 Napoli, M i a . Accepted for publication November 5. 1991 CCC 0148-639X1921070788-08 $04 00 0 1992 John Wiley & Sons, Inc
788
Amiodarone, Experimental Neuropathy
amiodarone-fed rats because of a functional blood-nerve barrier.' In this study we demonstrate that acute endoneurial injection of' different doses of amiodarone in rats produce various electrophysiologic and pathologic changes ranging from pure demyelination to complete axon degeneration. MATERIALS AND METHODS
Forty-five male Wistar rats weighing 200 to 300 g were used in these experiments. All procedures were carried out under anesthesia by intraperitoneal injection of a mixture of ketamine 110 mg/ kg, xylazine 13.4 mg/kg, and atropine 0.05 mg/kg. T h e sciatic nerve was exposed by an aseptic surgical incision from sciatic notch to below popliteal fossa. Using a micrometer syringe, 30 FL of amiodarone solution was injected into the subperineurial region of the single fascicle of the posterior tibial nerve under a surgical microscope (OPMI 99, Zeiss, Milan). The commercial product, Cordarone (Sigma-Tau, Pomezia, Rome), composed of amiodarone 150 mg, benzylic alcohol 60.6 mg, and polysorbate 300 mg was used in 3 mL of injectable solution; it was injected at each of the fol-
MUSCLE 8, NERVE
July 1992
lowing concentrations into 15 nerves: (1) 25 pg/mL (0.75 pg), (2) 50 pg/mL (1.5 pg), (3) I00 pg/mL (3 pg). The contralateral nerves were injected with 30 pL of physiological saline solution in 15 rats and with 30 pL of amiodarone solvents (i.e., benzylic alcohol 1212 pg and polysorbate 6 pg) in 20 rats. Five contralateral nerves were injected with 30pL of a benzylic alcohol solution (1212 Fg) and the remaining 5 with polysorbate solution (6 pg). T h e accuracy of injection was judged according to swelling of the nerve either proximal or distal to the injection point, and
to the disappearance of perineurial transverse striae.'" ElectrophysiologicalTechiques. Electrophysiological studies were performed on all animals immediately pre- and postinjection, at daily intervals until 10 days, and then every 3 to 5 days until observations were complete. Platinum needle electrodes (type 13L70, Dantec Elektronik, Skovlunde, Denmark) were used for stimulating and recording. Compound muscle action potentials (CMAPs) were recorded from small foot muscles. Proximal
I
$
.
DAY 0 (a)
.
!
'
1
. !:.
I?
ii
1 ,:i
DAY 7
DAY35
!
:
&
-
=
j +
my17
mV
1 4 msec
FIGURE 1. Electrophysiological follow-up of a rat injected with amiodarone 25 pglmL. (a) Day 0: preinjection. (b) Day 0: 5 to 60 minutes postinjection.
Amiodarone, Experimental Neuropathy
MUSCLE & NERVE
July 1992
789
stimulating electrodes were placed adjacent to the sciatic nerve at the sciatic notch, and the distal pair was inserted at the ankle (i.e., above and below the injection site). Using a Dantec Electromyographer Type 1500 system, square-wave pulses of 0.2 ms and supramaximal strength were delivered at 1-second intervals. Presence of conduction block was determined by comparing peak-to-peak amplitude of proximal and distal CMAP." Limb temperature was maintained between 37" and 38" C by a thermoster probe inserted subcutaneously and a thermostatically controlled infrared heat lamp (Dantec, 31B40). The axon regeneration rate was determined by the first appearance of an evoked distal CMAP, i.e., the onset of recordable reinnervation of foot muscle. Normal regeneration rate was previously obtained, according to the sanie criteria, in 14 control rats, whose sciatic nerves had been tied for 2 days at the popliteal fossa at the same point where injection was performcd.18 Morphological Techniques. For each drug concentration, at least two sciatic nerves and the contralateral nerve were studied at 5, 10, 15, 20, and 40 days after injection. Animals were killed by pentobarbitone overdose. T h e whole exposed sciatic nerve was fixcd in situ for 5 minutes with 0.1 niol/L phosphale-buffered 2.5% glutaraldehyde (GTA) solution. 'The nerve was then excised and cut into small blocks, fixed in 2.5% GTA solution for 1 hour and postfixed with 2% osmium tetroxide, dehydrated in graded ethanol and propylene oxide, and embedded in epoxy plastic. For light and electron microscopy, and teased fibers nerve preparation was performed according to the technique previously described.
> E
7
40
days
FIGURE 2. Amiodarone, 25 pg/mL injection. Amplitude changes in distal (m) and proximal ( + ) recorded CMAPs are plotted against days postinjection.
clearly differed according to amiodarone concentration. At a 25 kg/mL concentration, amiodarone resulted in a subacute, incomplete conduction block evident at day 3 postinjection. Proximal response amplitude decreased to 25% of initial value, while distal response amplitude was only slightly and temporarily decreased, probably because of partial drug diffusion below the injection site (Fig. 1). This conduction block remained stable for a further 7 days, and then reversed slowly; CMAP amplitude for proximal stimulation was completely normal at day 35 postinjection (Fig. 2). There was no electrophysiological evidence of axon degeneration at this concentration. The 50 kg/mL concentration of amiodarone produced a faster evolving conduction block and evidence of parallel axon degeneration: at day 3 postinjection,
'
18
RESULTS
N o significant change in conduction velocity was seen in any of the 15 nerves injected with saline solution. Injection of benzylic alcohol and polysorbate solution determined an acute conduction block immediately after injection with disappearance of proximal CMAPs for 40 to 55 minutes in all 20 nerves, but amplitude was normal in all nerves at electrophysiological tests by the next day. T h e same sequence was observed after injection ol' benzylic alcohol alone, but not after polysorbate solution. Acute but temporary conduction block was seen in all nerves injected with amiodarone solution at any concentration (Fig. l), but subsequent events Electrophysiological Observations.
790
Amiodarone, Experimental Neuropathy
4
14
-
12
J
10
-
T
>
E
8 6 -
42 -
0 ) 0
I
5
,
,
1
I
1 0 1 5 2 0 25
I
I
30 3 5
I
I
4 0 45
I
I
I
I
50 5 5 60 6 5
days
FIGURE 3. Amiodarone, 50 pg/mL injection. Amplitude changes in distal (W) and proximal (A)recorded CMAPs are plotted against days postinjection.
MUSCLE & NERVE
July 1992
proximal and distal CMAP amplitudes decreased to 20% and 35%, respectively, of initial values (Fig. 3), and a slow recovery began at day 10 but was complete only at day 60 (Fig. 3 ) . At a 100 pg/mL concentration, there was no response at day 3 postinjection to proximal stimulation in any nerve. By day 5 no CMAP could be obtained by proximal or distal stimulation in any nerve, indicating a severe acute motor axon degeneration. Electrophysiological follow-up revealed axon regeneration in all nerves beginning 36.9 '. 5.5 days after injection. The amplitude of proximal responses returned to previous levels about 50 days after the beginning of muscle reinnervation, which was significantly delayed ( P < 0.001) in comparison with that observed in 14 normal controls (28.7 4.2 days).I6
*
Neuropathological Studies.
Light and Electron Mi-
croscopy. Morphological and electrophysiological findings were closely correlated. At day 5 post-
injection, when cross-sections of nerves injected with 25 pg/mL of amiodarone showed only minimal aspecific changes. By day 10, extensive demyelination, myelin debris surrounding denuded axons, and many macrophages containing vacuoles and myelin debris were seen in more than 75% of the fibers (Figs. 4 and 5). Sixty-one percent of teased fibers showed segmental demyelination or widening of nodal gaps, although there were some denuded axons surrounded by activated Schwann cells with prominent profiles of rough endoplasmic reticulum (RER), which could precede remyelination (Fig. 6). Clear signs of remyelination were first noted on day 15 and, at day 40, cross-section was completely normal. Posterior tibia1 nerve segments 2 cm distal to injection site examined at days 5, 10, 15, and 20 postinjection revealed no significant axon degeneration. In the nerves injected with a 50 pg/mL concentration, there were signs at day 5 of demyelination and axon degeneration, which became
FIGURE 4. Electron micrograph showing a myelinated fiber entered by two vacuolated macrophage cells and some cell processes. A macrophage process ( t ) has penetrated the myelin sheath, which is undergoing demyelinating changes. The whole is surrounded by the basal lamina ( t ) of the nerve fiber. Uranyl acetate and lead citrate; day 10 postinjection of 25 pg/mL; x6900.
Amiodarone, Experimental Neuropathy
MUSCLE & NERVE
July 1992
791
FIGURE 5. Electron micrograph showing more advanced demyelinating changes: two denuded axons (*) are surrounded by macrophage cells containing vacuoles and debris of myelin sheaths. Uranyl acetate and lead citrate; day 10 postinjection of 25 pg/mL; ~6900.
marked by day 10. Almost complete axon degeneration was observed in the nerve injected with 100 pg/mL concentration of amiodarone (Fig. 7), but axon regeneration was present by day 20 postinjection, and was coniplete by day 40. Because of the acute and unique intoxication of the animals we did not observe any lysosoinal inclusions in Schwann, endoneurial, or endothelial cells. DlSCUSSION
Amiodarone has an inhibitory effect on lysosomal phospholipasis A1 and A2,4 which leads to the formation of lysosomal phospholipid-containing inclusions in many cell types. These drug-induced cell inclusions are present only in regions of the nervous system without a blood-brain or bloodnerve barrier.2 Failure to produce a neuropathy in animals fed with the drug2 confirmed an efficient blood-nerve barrier to amiodarone in rats. Human amiodarone-induced neuropathy could be
792
Amiodarone, Experimental Neuropathy
due to alterations in nerve vascular permeability, induced by the drug, or normally increased with age.7 Jacobs and Costa-Jussa6 hypothesize that, in humans, the peripheral nerve changes consist of initial Schwann cell abnormality and subsequent myelin breakdown, and that the “axonal degeneration could result from compressive effects of the increased volume of Schwann cell cytoplasm.” Our study clearly demonstrates that amiodarone, when injected into rats intraneurally at different concentrations, can produce different electrophysiologic and histologic changes. The injection of 25 pg/mL concentration resulted in electrophysiological evidence of a profound subacute demyelination starting on day 3 postinjection. The conduction block, an electrophysiological hallmark of a focal demyelinative lesion,12 persists for a further 7 days after which a slow but complete return to normal conduction ensues. In the peripheral nerves examined at day 10 postinjection we observed prominent Schwann-cell cytoplasmic degeneration. The most frequent abnormalities at
MUSCLE & NERVE
July 1992
FIGURE 6. A demyelinated axon completely surrounded by cytoplasm rich in RER profiles of a Schwann cell (top), a stage propedeutic to remyelination. Another Schwann cell is at bottom. The Schwann cells lie within their own basal lamina and both are encircled by a corrugated ancient basement membrane ( t ). Uranyl acetate and lead citrate; day 15 postinjection of 25 pg/mL; ~9000.
the teased fibers were segmental demyelination and the widening of nodal gaps, which were consistent with the observed decline of proximal CMAP amplitude. The fact that remyelination was complete at day 35 suggests that the nerve segments were never devoid of Schwann cells able to divide. A similar, although slightly slower, sequence of electrophysiologic abnormalities has been described after intraneurial injection of tunicamycin"; the drug was thought to inhibit some metabolic pathways in the Schwann cells selective for myelin maintenance, without affecting the cell's ability to proliferate and remyelinate." At this lowest amiodarone concentration, we saw no significant axon degeneration and, thus, a relatively selective effect of amiodarone producing myelin breakdown- but sparing Schwann cells' ability to proliferate and remyelinate. Nor did we find any clear proof of Schwann-cell cytoplasmic changes preceding myelin breakdown at least at day 5 after injection. Moreover, there was clear
Amiodarone, Experimental Neuropathy
histological and electrophysiological evidence of dernyelination in all 15 rats injected with a 25 pg/mI, concentration of amiodarone. Therefore, we cannot describe this neuropathy as a schwannopathy, as suggested by Jacobs and Costa-Jussa.' These apparently discordant results could be due to the different nerve concentration of the drug. As suggested in the Said and Duckett,I5 and Lampert and Garrett" studies using tellurium, the occurrence of a schwannopathy or a myelinopathy could depend on the amount of amiodarone that reach Schwann cells. At higher doses, in our study, myelin changes were associated with partial (50 pg/mL) or complete (100 p,g/mL) axon degeneration to an extent that has never been described as a consequence of primary demyelinating neuropathy. For the same reason, the swollen Schwann cells are unlikely to exert a compressive effect. Amiodarone at high concentrations in the peripheral nerve could thus have a direct toxic effect on axons. Different pathological changes de-
MUSCLE & NERVE
July 1992
793
FIGURE 7. Electron micrograph showing three myelinated fibers undergoing Wallerian-like degeneration. Uranyl acetate and lead citrate; day 10 postinjection of 100 WgimL; x5400.
scribed in human amiodarone neuropathy could be related to different drug concentrations in the nerve due to variable blood-nerve barrier efficacy. A temporal sequence from Schwann-cell cytoplasmic alterations to axon degeneration, independent of drug concentration is unlikely. The normal although delayed axon regeneration after 100 pg/mL amiodarone injection, and the electrophysiological evidence of normal CMAP amplitude recovery suggest that the delay is limited to the beginning of the regeneration process, when the number of Schwann cells may be insufficient to guide the regenerating axons. By the time the surviving Schwann cells duplicate, and others probably migrate from adjacent nerve segments, the regeneration process has normal characteristics. However, we cannot exclude the possibility that amiodarone has a direct effect upon axons themselves and delays regeneration on this basis. We have no experimental evidence to explain the acute temporary conduction block observed constantly with injection of benzylic alcohol. The very short duration of electrophysiologic abnormalities
794
Arniodarone, Experimental Neuropathy
and absence of pathological findings in the nerves studied suggest a possible action on the sodium voltage-gated channels.
REFERENCES 1. Barbicri F, Santoro L, Crisci C, Massini R, Kusso E, Cam-
panella G: Is the sensory neuropathy in ataxia-telangiectasia distinguishable from that in Friedreich’s ataxia? Actu Neuropathol ( B e d ) 1986;69:2 13-2 19. 2. Costa-Jussa FR, Jacobs JM: T h e pathology of aniiodarone neurotoxicity. I. Experimental studies with reference to changes in other tissues. Brain 1985;108:735-752. 3. Dudognon P, Hauw JJ, De Baecque C , Derrida JP, Escourolle R, Nick J: Neuropathie an chlorhydrate cl’amiodarone. Rev Nrurol 1979;135:527-540. 4. Heath MF, Costa-Jussa FR, Jacobs JM, Jacobson W: T h e induction of pulmonary phospholipidosis and the inhihition of lysosonial phospholipases by arniodarone. B17:t.j E x p Palhol 1985;66:391-398. 5. Heger JJ, Prystowsky EN, Jackman WM: Amiodarone clinical efficacy and electrophysiology during long-term therapy for recurrent ventricular tachycardia or ventricular fibrillation. N EnglJ Med 1981;305:539-545. 6. Jacobs JM, Costa-Jussa FR: T h e pathology of amiodarone neurotoxicity. 11. Peripheral neuropathy in man. Bruin 1985; 1081753- 769. 7. Jacobs JM, Love S: Qualitative and quantitative morphol-
MUSCLE & NERVE
July 1992
ogy of human sural nerve at different ages. Bruin 1985;108:897-924. 8. Kaeser HE: Amiodaron-Neuropathie. Schweizerzsche Mcdizinische Wochenschrzft. 1974; 104:606- 608. 9. Kaeser HE, Ulrich J , Wuthrich R: Aniiodarone-Neuropathie. Pruxis 3976;65:1121-1122. 10. Lampert PW, Garrett RS: Mechanism of demyelination in tellurium neuropathy: Electron microscopic observations. Lab Invest 1971;25:380-388. I 1. Marcus F1, Fontaine GH, Frank R, Grosgogeat Y: Clinical pharmacology and therapeutic applications of the antiarrhythmic agent, amiodarone. Ant Hearl J 1981;101:480493. 12. McDonald WI: Physiological consequences of dernyelinatioti, in Sumner AJ (cd): Th.r P h y . d o a of Perifiheral NPTW Disease. Philadelphia, Saunders, 1980, p p 265-286. 13. Mcier C, Kauer B, Miller U, Ludin HP: Neuromyopathy during chronic amiodarone treatment: A case report. J Nrurol 1979;220:231-239.
Amiodarone, Experimental Neuropathy
14. Pellissier JF, Fonget J, Cros D, De Victor B, Serratrice G, Toga M: Peripheral neuropathy induced by amiodarone chlorhydrate: A clinicopathological study. J Neurol Sci 1984;63:25 1-266. 15. Said J, Duckett S: Tellurium-induced myelinopathy in adult rats. Muscle Nerve 1981;4:319-325. 16. Santoro L, England J , Rehee E, Caruso G, Sumner AJ: Effects on axon regeneration of doxorubicin microin.jected into the sciatic nerve of rats. ElectroencephaloLgr Clin Neurophysiology 1989;72 :8 1P. 17. Sumner AJ, Saida K, Saida T, Silberberg DH, Asbury AK: Acute conduction block associated with experimental antiserum-mediated demyelination of peripheral nerve. Ann Neurol 1982;11:469-477. 18. Sumner AJ, Jensen JJ, Polard JD: Tunicamycin-induced inhibition of in vivo Schwann cell metabolic maintenance of myelin. Neurology 1983;33(suppl 2):99.
MUSCLE & NERVE
July 1992
795
AMIODARONE-INDUCED EXPERIMENTAL ACUTE NEUROPATHY IN RATS LUClO SANTORO, FABRlZlO BARBIERI, RENATO NUCCIOTTI, FRANCESCO BATTAGLIA, FRANCESCO CRISPI, M. RAGNO, P. GRECO, and G. CARUSO
Amiodarone, a diiodinated benzofuran derivative, is an a- and P-antagonist, which is effective in treatment of cardiac arrhythmia^.^^" A number of cases of peripheral neuropathy associated with amiodarone treatment have been reported since 19'74.' Based on neuropathological studies, this neuropathy has ben defined as axonal, 13,14 almost purely demyelinating,' and both axonal and demyelinating in type.' Jacobs and Costa-Jussa' have suggested that in man the Schwann-cell changes precede myelin breakdown and eventual axon degeneration. However, they failed to produce a neuropathy in
From the Department of Clinical Neurophysiology and Neurologic Clinic, Second School of Medicine, Federico It University of Naples, Naples, Italy (Drs. Santoro, Barbieri, Nucciotti, Battaglia, Crispi, and Caruso); and Department of Pharmaceutical Chemistry. Faculty of Pharmacy, Federico II University of Naples, Naples, Italy Address reprint requests lo Lucio Santoro, MD, Cattedra di Neurofisiopatologia II Facolta di Medicina e Chirurgia, UniversitA degli Studi di Napoli Federico 11, via Sergio Pansini 5, 80131 Napoli, M i a . Accepted for publication November 5. 1991 CCC 0148-639X1921070788-08 $04 00 0 1992 John Wiley & Sons, Inc
788
Amiodarone, Experimental Neuropathy
amiodarone-fed rats because of a functional blood-nerve barrier.' In this study we demonstrate that acute endoneurial injection of' different doses of amiodarone in rats produce various electrophysiologic and pathologic changes ranging from pure demyelination to complete axon degeneration. MATERIALS AND METHODS
Forty-five male Wistar rats weighing 200 to 300 g were used in these experiments. All procedures were carried out under anesthesia by intraperitoneal injection of a mixture of ketamine 110 mg/ kg, xylazine 13.4 mg/kg, and atropine 0.05 mg/kg. T h e sciatic nerve was exposed by an aseptic surgical incision from sciatic notch to below popliteal fossa. Using a micrometer syringe, 30 FL of amiodarone solution was injected into the subperineurial region of the single fascicle of the posterior tibial nerve under a surgical microscope (OPMI 99, Zeiss, Milan). The commercial product, Cordarone (Sigma-Tau, Pomezia, Rome), composed of amiodarone 150 mg, benzylic alcohol 60.6 mg, and polysorbate 300 mg was used in 3 mL of injectable solution; it was injected at each of the fol-
MUSCLE 8, NERVE
July 1992
lowing concentrations into 15 nerves: (1) 25 pg/mL (0.75 pg), (2) 50 pg/mL (1.5 pg), (3) I00 pg/mL (3 pg). The contralateral nerves were injected with 30 pL of physiological saline solution in 15 rats and with 30 pL of amiodarone solvents (i.e., benzylic alcohol 1212 pg and polysorbate 6 pg) in 20 rats. Five contralateral nerves were injected with 30pL of a benzylic alcohol solution (1212 Fg) and the remaining 5 with polysorbate solution (6 pg). T h e accuracy of injection was judged according to swelling of the nerve either proximal or distal to the injection point, and
to the disappearance of perineurial transverse striae.'" ElectrophysiologicalTechiques. Electrophysiological studies were performed on all animals immediately pre- and postinjection, at daily intervals until 10 days, and then every 3 to 5 days until observations were complete. Platinum needle electrodes (type 13L70, Dantec Elektronik, Skovlunde, Denmark) were used for stimulating and recording. Compound muscle action potentials (CMAPs) were recorded from small foot muscles. Proximal
I
$
.
DAY 0 (a)
.
!
'
1
. !:.
I?
ii
1 ,:i
DAY 7
DAY35
!
:
&
-
=
j +
my17
mV
1 4 msec
FIGURE 1. Electrophysiological follow-up of a rat injected with amiodarone 25 pglmL. (a) Day 0: preinjection. (b) Day 0: 5 to 60 minutes postinjection.
Amiodarone, Experimental Neuropathy
MUSCLE & NERVE
July 1992
789
stimulating electrodes were placed adjacent to the sciatic nerve at the sciatic notch, and the distal pair was inserted at the ankle (i.e., above and below the injection site). Using a Dantec Electromyographer Type 1500 system, square-wave pulses of 0.2 ms and supramaximal strength were delivered at 1-second intervals. Presence of conduction block was determined by comparing peak-to-peak amplitude of proximal and distal CMAP." Limb temperature was maintained between 37" and 38" C by a thermoster probe inserted subcutaneously and a thermostatically controlled infrared heat lamp (Dantec, 31B40). The axon regeneration rate was determined by the first appearance of an evoked distal CMAP, i.e., the onset of recordable reinnervation of foot muscle. Normal regeneration rate was previously obtained, according to the sanie criteria, in 14 control rats, whose sciatic nerves had been tied for 2 days at the popliteal fossa at the same point where injection was performcd.18 Morphological Techniques. For each drug concentration, at least two sciatic nerves and the contralateral nerve were studied at 5, 10, 15, 20, and 40 days after injection. Animals were killed by pentobarbitone overdose. T h e whole exposed sciatic nerve was fixcd in situ for 5 minutes with 0.1 niol/L phosphale-buffered 2.5% glutaraldehyde (GTA) solution. 'The nerve was then excised and cut into small blocks, fixed in 2.5% GTA solution for 1 hour and postfixed with 2% osmium tetroxide, dehydrated in graded ethanol and propylene oxide, and embedded in epoxy plastic. For light and electron microscopy, and teased fibers nerve preparation was performed according to the technique previously described.
> E
7
40
days
FIGURE 2. Amiodarone, 25 pg/mL injection. Amplitude changes in distal (m) and proximal ( + ) recorded CMAPs are plotted against days postinjection.
clearly differed according to amiodarone concentration. At a 25 kg/mL concentration, amiodarone resulted in a subacute, incomplete conduction block evident at day 3 postinjection. Proximal response amplitude decreased to 25% of initial value, while distal response amplitude was only slightly and temporarily decreased, probably because of partial drug diffusion below the injection site (Fig. 1). This conduction block remained stable for a further 7 days, and then reversed slowly; CMAP amplitude for proximal stimulation was completely normal at day 35 postinjection (Fig. 2). There was no electrophysiological evidence of axon degeneration at this concentration. The 50 kg/mL concentration of amiodarone produced a faster evolving conduction block and evidence of parallel axon degeneration: at day 3 postinjection,
'
18
RESULTS
N o significant change in conduction velocity was seen in any of the 15 nerves injected with saline solution. Injection of benzylic alcohol and polysorbate solution determined an acute conduction block immediately after injection with disappearance of proximal CMAPs for 40 to 55 minutes in all 20 nerves, but amplitude was normal in all nerves at electrophysiological tests by the next day. T h e same sequence was observed after injection ol' benzylic alcohol alone, but not after polysorbate solution. Acute but temporary conduction block was seen in all nerves injected with amiodarone solution at any concentration (Fig. l), but subsequent events Electrophysiological Observations.
790
Amiodarone, Experimental Neuropathy
4
14
-
12
J
10
-
T
>
E
8 6 -
42 -
0 ) 0
I
5
,
,
1
I
1 0 1 5 2 0 25
I
I
30 3 5
I
I
4 0 45
I
I
I
I
50 5 5 60 6 5
days
FIGURE 3. Amiodarone, 50 pg/mL injection. Amplitude changes in distal (W) and proximal (A)recorded CMAPs are plotted against days postinjection.
MUSCLE & NERVE
July 1992
proximal and distal CMAP amplitudes decreased to 20% and 35%, respectively, of initial values (Fig. 3), and a slow recovery began at day 10 but was complete only at day 60 (Fig. 3 ) . At a 100 pg/mL concentration, there was no response at day 3 postinjection to proximal stimulation in any nerve. By day 5 no CMAP could be obtained by proximal or distal stimulation in any nerve, indicating a severe acute motor axon degeneration. Electrophysiological follow-up revealed axon regeneration in all nerves beginning 36.9 '. 5.5 days after injection. The amplitude of proximal responses returned to previous levels about 50 days after the beginning of muscle reinnervation, which was significantly delayed ( P < 0.001) in comparison with that observed in 14 normal controls (28.7 4.2 days).I6
*
Neuropathological Studies.
Light and Electron Mi-
croscopy. Morphological and electrophysiological findings were closely correlated. At day 5 post-
injection, when cross-sections of nerves injected with 25 pg/mL of amiodarone showed only minimal aspecific changes. By day 10, extensive demyelination, myelin debris surrounding denuded axons, and many macrophages containing vacuoles and myelin debris were seen in more than 75% of the fibers (Figs. 4 and 5). Sixty-one percent of teased fibers showed segmental demyelination or widening of nodal gaps, although there were some denuded axons surrounded by activated Schwann cells with prominent profiles of rough endoplasmic reticulum (RER), which could precede remyelination (Fig. 6). Clear signs of remyelination were first noted on day 15 and, at day 40, cross-section was completely normal. Posterior tibia1 nerve segments 2 cm distal to injection site examined at days 5, 10, 15, and 20 postinjection revealed no significant axon degeneration. In the nerves injected with a 50 pg/mL concentration, there were signs at day 5 of demyelination and axon degeneration, which became
FIGURE 4. Electron micrograph showing a myelinated fiber entered by two vacuolated macrophage cells and some cell processes. A macrophage process ( t ) has penetrated the myelin sheath, which is undergoing demyelinating changes. The whole is surrounded by the basal lamina ( t ) of the nerve fiber. Uranyl acetate and lead citrate; day 10 postinjection of 25 pg/mL; x6900.
Amiodarone, Experimental Neuropathy
MUSCLE & NERVE
July 1992
791
FIGURE 5. Electron micrograph showing more advanced demyelinating changes: two denuded axons (*) are surrounded by macrophage cells containing vacuoles and debris of myelin sheaths. Uranyl acetate and lead citrate; day 10 postinjection of 25 pg/mL; ~6900.
marked by day 10. Almost complete axon degeneration was observed in the nerve injected with 100 pg/mL concentration of amiodarone (Fig. 7), but axon regeneration was present by day 20 postinjection, and was coniplete by day 40. Because of the acute and unique intoxication of the animals we did not observe any lysosoinal inclusions in Schwann, endoneurial, or endothelial cells. DlSCUSSION
Amiodarone has an inhibitory effect on lysosomal phospholipasis A1 and A2,4 which leads to the formation of lysosomal phospholipid-containing inclusions in many cell types. These drug-induced cell inclusions are present only in regions of the nervous system without a blood-brain or bloodnerve barrier.2 Failure to produce a neuropathy in animals fed with the drug2 confirmed an efficient blood-nerve barrier to amiodarone in rats. Human amiodarone-induced neuropathy could be
792
Amiodarone, Experimental Neuropathy
due to alterations in nerve vascular permeability, induced by the drug, or normally increased with age.7 Jacobs and Costa-Jussa6 hypothesize that, in humans, the peripheral nerve changes consist of initial Schwann cell abnormality and subsequent myelin breakdown, and that the “axonal degeneration could result from compressive effects of the increased volume of Schwann cell cytoplasm.” Our study clearly demonstrates that amiodarone, when injected into rats intraneurally at different concentrations, can produce different electrophysiologic and histologic changes. The injection of 25 pg/mL concentration resulted in electrophysiological evidence of a profound subacute demyelination starting on day 3 postinjection. The conduction block, an electrophysiological hallmark of a focal demyelinative lesion,12 persists for a further 7 days after which a slow but complete return to normal conduction ensues. In the peripheral nerves examined at day 10 postinjection we observed prominent Schwann-cell cytoplasmic degeneration. The most frequent abnormalities at
MUSCLE & NERVE
July 1992
FIGURE 6. A demyelinated axon completely surrounded by cytoplasm rich in RER profiles of a Schwann cell (top), a stage propedeutic to remyelination. Another Schwann cell is at bottom. The Schwann cells lie within their own basal lamina and both are encircled by a corrugated ancient basement membrane ( t ). Uranyl acetate and lead citrate; day 15 postinjection of 25 pg/mL; ~9000.
the teased fibers were segmental demyelination and the widening of nodal gaps, which were consistent with the observed decline of proximal CMAP amplitude. The fact that remyelination was complete at day 35 suggests that the nerve segments were never devoid of Schwann cells able to divide. A similar, although slightly slower, sequence of electrophysiologic abnormalities has been described after intraneurial injection of tunicamycin"; the drug was thought to inhibit some metabolic pathways in the Schwann cells selective for myelin maintenance, without affecting the cell's ability to proliferate and remyelinate." At this lowest amiodarone concentration, we saw no significant axon degeneration and, thus, a relatively selective effect of amiodarone producing myelin breakdown- but sparing Schwann cells' ability to proliferate and remyelinate. Nor did we find any clear proof of Schwann-cell cytoplasmic changes preceding myelin breakdown at least at day 5 after injection. Moreover, there was clear
Amiodarone, Experimental Neuropathy
histological and electrophysiological evidence of dernyelination in all 15 rats injected with a 25 pg/mI, concentration of amiodarone. Therefore, we cannot describe this neuropathy as a schwannopathy, as suggested by Jacobs and Costa-Jussa.' These apparently discordant results could be due to the different nerve concentration of the drug. As suggested in the Said and Duckett,I5 and Lampert and Garrett" studies using tellurium, the occurrence of a schwannopathy or a myelinopathy could depend on the amount of amiodarone that reach Schwann cells. At higher doses, in our study, myelin changes were associated with partial (50 pg/mL) or complete (100 p,g/mL) axon degeneration to an extent that has never been described as a consequence of primary demyelinating neuropathy. For the same reason, the swollen Schwann cells are unlikely to exert a compressive effect. Amiodarone at high concentrations in the peripheral nerve could thus have a direct toxic effect on axons. Different pathological changes de-
MUSCLE & NERVE
July 1992
793
FIGURE 7. Electron micrograph showing three myelinated fibers undergoing Wallerian-like degeneration. Uranyl acetate and lead citrate; day 10 postinjection of 100 WgimL; x5400.
scribed in human amiodarone neuropathy could be related to different drug concentrations in the nerve due to variable blood-nerve barrier efficacy. A temporal sequence from Schwann-cell cytoplasmic alterations to axon degeneration, independent of drug concentration is unlikely. The normal although delayed axon regeneration after 100 pg/mL amiodarone injection, and the electrophysiological evidence of normal CMAP amplitude recovery suggest that the delay is limited to the beginning of the regeneration process, when the number of Schwann cells may be insufficient to guide the regenerating axons. By the time the surviving Schwann cells duplicate, and others probably migrate from adjacent nerve segments, the regeneration process has normal characteristics. However, we cannot exclude the possibility that amiodarone has a direct effect upon axons themselves and delays regeneration on this basis. We have no experimental evidence to explain the acute temporary conduction block observed constantly with injection of benzylic alcohol. The very short duration of electrophysiologic abnormalities
794
Arniodarone, Experimental Neuropathy
and absence of pathological findings in the nerves studied suggest a possible action on the sodium voltage-gated channels.
REFERENCES 1. Barbicri F, Santoro L, Crisci C, Massini R, Kusso E, Cam-
panella G: Is the sensory neuropathy in ataxia-telangiectasia distinguishable from that in Friedreich’s ataxia? Actu Neuropathol ( B e d ) 1986;69:2 13-2 19. 2. Costa-Jussa FR, Jacobs JM: T h e pathology of aniiodarone neurotoxicity. I. Experimental studies with reference to changes in other tissues. Brain 1985;108:735-752. 3. Dudognon P, Hauw JJ, De Baecque C , Derrida JP, Escourolle R, Nick J: Neuropathie an chlorhydrate cl’amiodarone. Rev Nrurol 1979;135:527-540. 4. Heath MF, Costa-Jussa FR, Jacobs JM, Jacobson W: T h e induction of pulmonary phospholipidosis and the inhihition of lysosonial phospholipases by arniodarone. B17:t.j E x p Palhol 1985;66:391-398. 5. Heger JJ, Prystowsky EN, Jackman WM: Amiodarone clinical efficacy and electrophysiology during long-term therapy for recurrent ventricular tachycardia or ventricular fibrillation. N EnglJ Med 1981;305:539-545. 6. Jacobs JM, Costa-Jussa FR: T h e pathology of amiodarone neurotoxicity. 11. Peripheral neuropathy in man. Bruin 1985; 1081753- 769. 7. Jacobs JM, Love S: Qualitative and quantitative morphol-
MUSCLE & NERVE
July 1992
ogy of human sural nerve at different ages. Bruin 1985;108:897-924. 8. Kaeser HE: Amiodaron-Neuropathie. Schweizerzsche Mcdizinische Wochenschrzft. 1974; 104:606- 608. 9. Kaeser HE, Ulrich J , Wuthrich R: Aniiodarone-Neuropathie. Pruxis 3976;65:1121-1122. 10. Lampert PW, Garrett RS: Mechanism of demyelination in tellurium neuropathy: Electron microscopic observations. Lab Invest 1971;25:380-388. I 1. Marcus F1, Fontaine GH, Frank R, Grosgogeat Y: Clinical pharmacology and therapeutic applications of the antiarrhythmic agent, amiodarone. Ant Hearl J 1981;101:480493. 12. McDonald WI: Physiological consequences of dernyelinatioti, in Sumner AJ (cd): Th.r P h y . d o a of Perifiheral NPTW Disease. Philadelphia, Saunders, 1980, p p 265-286. 13. Mcier C, Kauer B, Miller U, Ludin HP: Neuromyopathy during chronic amiodarone treatment: A case report. J Nrurol 1979;220:231-239.
Amiodarone, Experimental Neuropathy
14. Pellissier JF, Fonget J, Cros D, De Victor B, Serratrice G, Toga M: Peripheral neuropathy induced by amiodarone chlorhydrate: A clinicopathological study. J Neurol Sci 1984;63:25 1-266. 15. Said J, Duckett S: Tellurium-induced myelinopathy in adult rats. Muscle Nerve 1981;4:319-325. 16. Santoro L, England J , Rehee E, Caruso G, Sumner AJ: Effects on axon regeneration of doxorubicin microin.jected into the sciatic nerve of rats. ElectroencephaloLgr Clin Neurophysiology 1989;72 :8 1P. 17. Sumner AJ, Saida K, Saida T, Silberberg DH, Asbury AK: Acute conduction block associated with experimental antiserum-mediated demyelination of peripheral nerve. Ann Neurol 1982;11:469-477. 18. Sumner AJ, Jensen JJ, Polard JD: Tunicamycin-induced inhibition of in vivo Schwann cell metabolic maintenance of myelin. Neurology 1983;33(suppl 2):99.
MUSCLE & NERVE
July 1992
795
