Journal of Clillical Neuro-ophthalmology 12(2): 121-127, 1992. '91992 Raven Press, Ltd., New York

Effects of Repeated Botulinum Toxin Injections on Orbicularis Oculi Muscle

Gary E. Borodic, M. D. and Robert Ferrante, M.S.

Histologic evaluation was conducted on 12 orbicularis Botulinum A toxin has been used to treat a num­ oculi specimens from 11 patients with essential blepha­ ber of movement disorders, including benign es­ rospasm and Meige's disease who had received an av­ sential blepharospasm, facial dyskinesia associated erage of 11.3 injections of botulinum A toxin over 3.5 years. Denervation was demonstrated by the spread of with Meige's disease, spasmodic torticollis, adult­ acetylcholinesterase staining on muscle fibers when onset spasmodic torticollis, occupational hand specimens were evaluated within 11 weeks of the last dystonias, and spasmodic dysphonia (1-9). Many injection. When specimens were taken after 12 weeks, of these conditions are chronic movement disor­ spread of acetylcholinesterase was confined to the neu­ ders that require repeated injections of the toxin to romuscular junctions, with little fiber size variability re­ sembling normal muscle. Fibrosis seen in three speci­ maintain a beneficial effect. The toxin effect gener­ mens could be correlated to prior surgery. Repeated in­ ally lasts 2--4 months (1-2). jections of botulinum toxin into human muscle do not The U.s. Food and Drug Administration has re­ appear to cause irreversible muscle atrophy or other de­ cently approved botulinum A toxin for use in generative changes. Denervation changes (fiber size blepharospasm after 8 years of clinical trials. Short­ variability, acetylcholinesterase spread) appear to corre­ late to the time interval since the last injection. term side effects of botulinum A toxin for many of Key Words: Blepharospasm-Meige's syndrome, Botu­ these conditions result from local toxin spread to linum A toxin-Dystonia-Acetylcholinesterase-Hemi­ contiguous muscles (7,8). Dysphagia results from facial spasm toxin spread to the peripharyngeal muscles after higher dose injections over the sternocleiodomas­ toid muscle used for the treatment of spasmodic torticollis (8). Diplopia may be a consequence of lower eyelid injections when toxin used to treat involuntary blepharospasm spreads to the inferior oblique muscle (7). Long-term or latent clinical complications of bot­ ulinum toxic injections have not been reported. However, because therapy for many of these movement disorders requires repeated injections, there is a potential for a long-term effect of toxin administration on human muscle, In this study, we report the chronic histopathologic effects in pa­ tients who have received repeated botulinum A toxin injections into the orbicularis oculi muscle for benign essential blepharospasm or Meige's disease for at least several years.

MATERIALS AND METHODS From the Massachusetts Eye and Ear Infirmary (G.E.B.) and Neuropathology Unit (R.F.), Massachusetts General Hospital, Patients involved in this study required surgical Boston, Massachusetts, U.S.A. Address correspondence and reprint requests to Dr. Gary E. intervention in two clinical situations. The first Borodic, 100 Charles River Plaza, Boston, MA 02114, U.S.A. was involutional ptosis caused by disinsertion of

121 122 G. E. BORODIC AND R. FERRANTE the attachment of the levator aponeurosis to the late the histochemical findings using various tarsal plate. To correct this condition, which often stains. involves reinsertion of the levator aponeurosis to Morphometric measurements were made with the tarsal plate, a small strip of orbicularis muscle the bioquant II system. Muscle fiber size was eval­ is removed to debulk the lid. The second clinical uated by measuring cross-sectional diameters in situation was unsatisfactory or incomplete symp­ 200 fibers in each specimen. Statistical compari­ tomatic response to botulinum toxin, which re­ sons were made using an F test, and the results sulted in the selection of an orbicularis myectomy were expressed as the mean +/- standard devia­ procedure as an alternative method of therapy. tion with variance. Each patient who became a surgical candidate The age of patients and indications for surgical for the correction of involutional disinsertion pto­ therapy are shown in Table 1. The number of bot­ sis or a myectomy procedure was advised that a ulinum toxin injections and the duration of ther­ histologic analysis of resected muscle for botuli­ apy are shown in Table 2, which also presents the num toxin activity would be conducted. Comori intervals between the last injection and the surgi­ trichrome stain and hematoxylin and eosin were cal procedure. The control values were taken from used to evaluate fiber size variability, and acetyl­ specimens removed during routine ptosis proce­ cholinesterase enzyme histochemistry was done to dures from patients who never received botulinum assess evidence of denervation. The acetylcholin­ toxin injections. Control data were derived from esterase stain has been previously reported as be­ four separate muscle specimens from three differ­ ing a valuable method of assessing denervation in ent patients. Each control patient demonstrated animal (9,10) and human striated muscle after the ptosis associated with elevated lid crease, and re­ injection of botulinum toxin (11,12). tracted preaponeurotic fat pads with good levator Specimens taken from orbicularis oculi muscle function indicating the diagnosis of involutional were immediately placed in cold (4°C) formol­ ptosis. None of the control patients had blepharo­ calcium (Baker's solution) and fixed for 6-12 hours spasm. at 4°C. Muscle specimens were cryoprotected in gum sucrose solution for 3 hours or until the tissue sank in solution. The muscle was oriented in a RESULTS longitudinal plane and frozen in a cryostat chuck using OTC compound (Tissue Tek). When the tis­ Human Specimens Treated with Repeated sue sample size permitted, the muscle was subdi­ Botulinum Toxin Injections vided and cut in longitudinal and cross-section. Cut tissue sections (10 f.lm) were adhered to gel­ The histologic response appeared to be depen­ coated slides, allowed to air dry for 2 minutes, and dent on the time interval between the last injection subsequently stained for acetylcholinesterase ac­ and the surgical procedure. The degree of acetyl­ tivity, using the method of Ceneser-Jensen and cholinesterase activity on muscle fibers was most Blackstad (13). Sections were incubated in a solu­ prominent within a period of 7-15 weeks following tion containing 13 ml maleic buffer (1.96 g maleic a therapeutic injection of botulinum toxin. Nor­ acid, 0.8 g NaOH, 10.8 ml IN NaOH, 200 ml dis­ mally, the cholinesterase staining pattern is con- tilled water), 10 mg acetylthiocholine iodide, 2 ml 0.03 M cupric sulfate, 1 ml 0.1 sodium citrate, and TABLE 1. Case history summary 0.5 mM potassium ferricyanide for 1 hour at 37°C. Contiguous cryostat sections were stained either Patient Indication for with hematoxylin and eosin or with Comori tri­ number Age Prior surgery intervention chrome stain to assess tissue morphology. 1 71 None Botox failure" To correlate the acetylcholinesterase staining 1a Myectomy Botox failure 2 59 None p~ttern Ptosis with other histochemical enzyme stains, 3. 65 None Botox failure histochemistry for myofiber ATPase activity was 4. 91 Blepharoplasty Botox failure conducted according to the method of Brooke and 5. 71 Myectomy Partial botox failure 6. 76 None Ptosis Kaiser (14), and NADH activity was assessed ac­ 7. 61 Blepharoplasty Botox failure cording to the method of Scarpelli, Hess, and 8. 78 None Ptosis Pearse (15). Acetylcholinesterase, ATPase, and 9. 71 None Ptosis 10 56 Blepharoplasty Botox failure NADH staining were conducted on five albino rab­ 11 62 None Ptosis bit longissimus dorsi muscle specimens at 5 weeks " Botulinum toxin failures underwent orbicularis myec­ after 10 lU botulinum toxin was injected to corre- tomy procedures of the upper lid and brow.

! Clin NeurfHJPhthalmol, Vol. 12, No.2, 1992 STRIATED MUSCLE AFTER BOTULINUM TOXIN INJECTIONS 123

TABLE 2. Effects of repeated botulinum toxic injections

Histologic Findings Number of Cholinesterase Patient Time since injections/duration Fiber size voriability fibrosis staining number last injection of therapy (median diameter in microns) (median size in microns) Control 28.43 s = 5.29, v = 28 Focal, 12.5, s = 5.4 1:H01 1 wk 7 injections, 1.25 yr 31.8 s = 8.2, v = 66 Focal, 21, s = 8 2:H02 8wk 16 injections, 3.3 yr 30 s = 14, v = 213 Diffuse, 35.2, s = 18 + 3:AK 7wk 14 injections, 3.4 yr 27.0 s = 15.5, v = 237 Diffuse, ND + 4:RG 22wk 19 injections, 5.5 yr 25.7 s = 6.8, v = 46 Focal, 20, s = 8 5:MS 52wk 11 injections, 5.0 yr 34.4 s = 7, v = 48 Focal,25 + 6:RN 52wk 4 injections, 2.7 yr 28.1 s = 8, v = 64 Focal, 22.5, s = 6 7:EP 18 wk 12 injections, 3.0 yr 32.0 s = 9, v = 81 Focal, 19.3, s = 6.7 8:TC 7wk 11 injections, 4.0 yr 28.5 s = 17.0, v = 289 Diffuse, 161.3, s =124 9:GC 16 wk 16 injections, 4.5 yr 26.0 s = 13.5, v = 172 Diffuse, 34, s = 13 10:TP 22wk 6 injections, 2.0 yr 28.1 s = 8, v = 63 ND 11 :CP 7wk 2 injections, 1.0 yr 23.5 s = 11, v = 121 Diffuse, 147, s = 70

S, standard deviation; v, variance; ND, not determined. fined to a focal area on the muscle fiber within the The degree of acetylcholinesterase spread on mus­ neuromuscular junction (see Fig. 2). Within 15 cle fibers was inversely related to the time interval weeks after botulinum toxin injection however, from injection to biopsy (Fig. 3). there appeared to be extensive spread of enzyme Muscle fiber variability was directly related to activity throughout large surface areas of muscle the time from the last injection. Muscle fiber diam­ fibers, indicating denervation (Fig. 1). This obser­ eter variability in a pattern consistent with neuro­ vation was especially evident in patient 7 (TC) and genic atrophy was noted in specimens within 3 patient 11 (CP). When the time interval between months of botulinum toxin injection (Fig. 4). Sta­ these procedures and the last botulinum toxin in­ tistical quantification demonstrated a significant jection was longer than 3 months, there was no difference in fiber diameter variability relative to substantial spread of acetylcholinesterase activity controls (p < .05, F test) for patients injected from the neuromuscular junction (Fig. 2) as com­ within 15 weeks of biopsy (see patients 1 (H02), 3, pared with normal control patterns (see Table 2). 8, 11). The degree of muscle fiber size variability

FIG. 1. Spread of acetylcholinesterase staining activity on human orbicularis oculi muscle tissue 7 weeks after therapeutic botulinum toxin injections.

] Clin Neuro-ophthalmol. Vol. 12, No.2, 1992 124 G. E. BORODIC AND R. FERRANTE

FIG. 2. Acetylcholinesterase activity is confined to the neuromuscular junction (ar­ row) after 52 weeks since the last botulinum toxin injection. relative to intervals since injections is shown in of type l:type 2 fibers increased significantly as Fig. 5. In patients not receiving botulinum toxin for compared to control specimens, with type 1 fibers 6 months, the fiber size variability was not signif­ representing 22% of the total population. icantly different from controls (F test, specimens The NADH activity also demonstrated alter­ in cases 4, 5, 6, 10) after an average of 10.0 injec­ ations in fiber size and fiber typing. In addition, tions over an average of 3.8 years. Acetylcholines­ this method identified changes in the intermyo­ terase staining pattern was also not significantly fibrillary network that were consistent with dener­ different from control specimens in this group. vation. These observations indicate no evidence of perma­ Acetylcholinesterase staining of these animal nent denervation after multiple botulinum injec­ specimens after 5 weeks demonstrated diffuse tions over several years if botulinum toxin had not staining characteristic on fiber surfaces that was been injected within 6 months before the evalua­ associated with increased fiber diameter size vari­ tion. ation (Fig. 6B). Fibrosis was apparent, however, in specimens 1, ATPase staining and NADH reactivity showed 3, 4, each of which were from patients who under­ findings typically associated with denervation in went previous surgical procedures (see Table 1). animal studies, which correlated and substantiated

ATPase and NADH Staining on Animal Tissues Showing Intense Spread of Acetylcholinesterase Activity After Botulinum Toxin Injections '50 In control specimens, myofibrillary ATPase ac­ tivity at pH = 9.4 demonstrated both type 1 and 100 type 2 fibers. The number or percent ratio of type 1-2 fibers was 3.5% of the total. Type 1 fibers were 50 distributed throughout the muscle specimens in saline injected controls. ~--+---.----+

At the injection site there was marked variation o 8 8 16 18 16 23 24 52 52 of muscle fiber size of both fiber types. The pattern Time Since last Injection of fiber typing was altered in that more grouped --+-- Choll nest Spread fiber typing and the presence of grouped atrophy FIG. 3. Denervation reflected by cholinesterase strongly suggested denevation (Fig. 6A). The ratio spread.

I elin NeuTo-vplll/w!nwL VoL 12, No.2, 1992 STRIATED MUSCLE AFTER BOTULINUM TOXIN INJECTIONS 125

the result found with acetylcholinesterase staining temporally with the clinical response seen after ad­ pattern after 5 weeks. ministering botulinum A toxin for the treatment of movement disorders in humans (16). Generally, the duration of effect of botulinum toxin for the DISCUSSION treatment of benign essential blepharospasm and After injection of nonlethal doses of botulinum facial dyskinesia associated with Meige's disease is toxin, Duchen (9,10) had demonstrated denerva­ 2-4 months. The temporal effect of botulinum tion followed by reinnervation with collateral ax­ toxin in the treatment of this disease correlates onal sprouting. The initial effect of botulinum very well with the histologic phenomena of func­ toxin is to block the release of acetylcholine from tional denervation as reflected by the histologic the presynaptic membrane at the neuromuscular specimens demonstrated in this study over a sev­ junction (9). Shortly after this block, within several eral month period. weeks, collateral axonal sprouting has been dem­ Two muscle specimens from patient 1 were an­ onstrated in both animal (10) and human orbicu­ alyzed: one taken 8 weeks after injection and one laris oculi tissue (11,12). This collateral sprouting is taken 1 week after injection. After 8 weeks, spread followed by the regeneration of neuromuscular of acetylcholinesterase was noted with increased junctions on muscle fibers. With the denervation, spread of acetylcholinesterase enzyme activity on Microns muscle fibers is demonstrated after 3--4 weeks and 350r------appears in normal levels after 12-14 weeks in ro­ 300 dent studies with botulinum A injections (9,10). 250 Earlier reports have indicated that sprouting from the preterminal axon can be noted 6 weeks to 200 3 years since the last botulinum toxin injection 150 (12). Although reversible muscle fiber atrophy and 100 acetylcholinesterase activity spread were demon­ 50 /' strated in this study, permanent dysmorphic alter­ I o ~---'-~~---'-----'------'----'------'-_ _'______'_ __'__L_ ations in the preterminal axon or ultrastructural Contro 1 wI- 16 18 22 22 52 52 changes of muscle fiber are still possible. Time from Injection The histologic cycle, consisting of denervation­ reinnervation and increased acetylcholinesterase FIG. 5. Orbicularis muscle fiber size, variations after staining activity on muscle fibers, correlates well botulinum toxin.

JClin Neuro-ophthalmol, Vol. 12, No.2, 1992 126 G. E. BORODIC AND R. FERRANTE

A

B

FIG.6. A: ATPase staining showing an increased proportion of type I fibers compared to controls (longissimus dorsi in albino rabbit). Also. notice type I fiber grouping and fiber atrophy suggesting denervation. B: Acetylcholinesterase staining after 5 weeks showing diffuse staining characteristic on fiber surfaces away from the neuromuscu­ lar junctions. The diffuse staining pattern was associated with increased fiber size and fiber variation.

fiber size variability, which indicated active dener­ longissimus dorsi muscles in albino rabbits (8) and vation. After 1 week, no appreciable spread in ace­ the soleus muscle in white mice (9,10). tylcholinesterase was seen, and there was minimal Orbicularis oculi muscle fiber size variability has fiber size variability. It appears that a latency pe­ been found to be greater than in other muscles riod of longer than 2 weeks is needed for acetyl­ (16,17). Hence, fiber size variability alone must be cholinesterase spread to occur after botulinum viewed with caution unless the variation is so great toxin injections. A similar latency was noted in the as to represent grouped muscular atrophy occur-

] Clm N£u,o'Dphtlralmo! Vol. 12, No.2, 1992 STRIATED MUSCLE AFTER BOTULINUM TOXIN INJECTIONS 127 ring with denervation. Histochemical analysis REFERENCES with the acetylcholinesterase stain is therefore par­ ticularly useful in confirming the denervation ef­ 1. Scott AB, Kennedy RA, Stubbs AA. Botulinum A toxin for the treatment of blepharospasm. Arch OphthalmoI1984;102: fect created by botulinum toxin. 1464--68. The potential long-term effects of repeated injec­ 2. Borodic GE, Cozzolino D. Blepharospasm and its treatment tions of botulinum toxin on the structure and in­ with emphasis on botulinum toxin. Plast Reconstr Surg 1989; 83:546-54. tegrity of human muscle are of particular concern. 3. Tsui JK, Eisen A, Stoessl L CaIne S, CaIne DB. Double-blind When muscle is subjected to repeated botulinum study of botulinum toxin in spasmodic torticollis. Lancet toxin injections, long-term chronic irreversible 1986;2:245-77. 4. Jankovic LOrman J. Botulinum toxin for cranial-cervical changes may occur; for example, permanent neu­ dystonias: a double-blind controlled study. Neurology 1987; rogenic muscle atrophy with resulting fibrosis and 37:616-23. permanent muscular scarring. These phenomena, 5. Borodic GE, Mills L, Joseph M. Therapy for spasmodic tor­ however, were not demonstrated in the group of ticollis with botulinum A toxin. Plast Reconstr Surg (in press). patients who had received multiple botulinum 6. Ludlow CL, Naunton, RF, Sedory SE, et al. Effects of Bot­ toxin injections over a number of years. One ex­ ulinum A toxin on speech: adduction spasmodic dyspho­ planation could be that during reinnervation there nia. Neurology 1988;38:1220-25. 7. Frueh BR, Nelson CC, Kapustiak JF, Musch DC. The effect is enough neurotrophic influence on muscle fibers of omitting botulinum toxin from the lower eyelid in bleph­ to prevent long-term contractures from denerva­ arospasm treatment. Am JOphthalmoI1988;106:45-7. tion. This may explain why botulinum toxin rarely 8. Borodic GE, Cozzollino D, Fay L, Joseph M, Ferrante R. causes complete paralysis of the muscles injected Botulinum A toxin for the treatment of spasmodic torticol­ lis: Dysphagia and regional toxin spread. Head Neck 1990; for therapeutic purposes. Antibody formation may 12:392-8. be instrumental in the reduced effect of repeated 9. Duchen LW. Changes in motor innveration and cholines­ botulinum toxin in some patients (16). In such sit­ terase localization of mouse: difference between fast and slow muscles. J Neurol Neurosurg Psychiat 1970;33:40-54. uations, repeated injections of botulinum toxin 10. Duchen LW. Histologic differences between soleus and may be less effective and may result in lesser de­ gastrocnemius muscles in mouse after local injection of bot­ grees of denervation over a period of time. ulinum toxin. J Physiol (Lond) 1969;204:17-8. 11. Borodic G, Ferrante R. Histopathology of human orbicu­ Some evidence of fibrosis was noted in several of laris oculi muscle after multiple injections of botulinum the muscle specimens. However, it could not be toxin. Presented at the American Academy of Ophthalmol­ clearly attributed to the botulinum toxin injections. ogy, 1988. The presence of fibrosis in the muscles studied ap­ 12. Holds JB, Alderson K, Fogg SG, Anderson RL. Motor nerve sprouting in human orbicularis muscle after botulinum A peared to correlate most closely to the prior lid injection. Invest Ophthalmol Vis Sci 1990;31:964-7. surgery, and it could have represented a portion of 13. Geneser-Jensen FA, Blackstad TW. Distribution of acetyl­ the surgical scar. cholinesterase in the hippocampal regions of the guinea pig: I. entorhinal area, parasabiculum, and presubiculum. In summary, repeated injections of botulinum A Z Zellforsch Mikrosk Anat1971;114:460-81 toxin over several years do not seem to be associ­ 14. Dubowitz V, Brooke MH. Histological and histochemical stains ated with irreversible denervation or changes in and reactions in muscle biopsy; A modern approach. London: the morphologic character of human orbicularis oc­ WB Saunders 1977:20-122. 15. Scarpelli DG, Hess R, Pearse AGE. The cytochemical local­ uli muscle fiber size and acetylcholinesterase stain­ ization of oxidative enzymes. I. diphophorpyridine nucle­ ing pattern. The diffuse atrophy seen in muscle otide diaphorase and triphophopyridine nucleotide diaph­ fibers in the short term appears to be a reversible orase. J Biophys Biochem CytoI1958;4:747 16. Brin MF, Fahn S, Moskowitz CB, et al. Localized injections phenomenon. Reduced acetylcholinesterase activ­ of botulinum toxin for the treatment of focal dystonias and ity increases to normal levels as the effects of the hemifacial spasm. Mov Disord 1987;2:237-54. toxin diminish. These histologic studies generally 17. Nelson C, Blaivas M. Histologic and histochemical charac­ teristics of eyebrow and eyelid muscle. J Neuropathol Exp parallel the clinical duration of action of botulinum NeuroI1990;49:298. toxin when used to treat blepharospasm, as well as 18. Blaivas M, Nelson C. Aging changes in human orbicularis other forms of segmental dystonias. oculi muscle. JNeuropathol Exp Neurol 1990;49:298.

I Clin Neuro-ophthalmol, Vol. 12, No.2, 1992