Experimental Corticosteroid Myopathy By MICHAEL G. SHEAHAN AND PAUL J. VIGNOS, JR., M.D. Triamcinolone treatment in rabbits using 1 mg./Kg. doses produces a differential metabolic effect, with predominant involvement of glycolytic-type fibers of muscle manifested by a significant atrophy of type I1 fibers. A consequent atrophy of muscle containing a large T of large doses of corticosteroids in the treatment of various conditions, such as connective tissue diseases, may cause muscle weakness in and Although it has been suggested that halogenated steroids produce the greatest changes: studies comparing the effects of various steroids have shown that many steroid compounds will produce decreases in total body and individual muscle weight in both dogs5 and mice.? Histological studies on human biopsy material have shown minimal and variable nonspecific changes in the muscle fibers. Histochemical staining methods have not been applied in a systematic study of steroid myopathy in human subjects because many patients with steroid-induced myopathy also have a primary disease capable of causing histochemical alterations in muscle. Steroid myopathy in experimental animals is manifested by histological muscle fiber changes ranging from no pathologic change to fiber necrosis, regeneration, and severe muscle atrophy. The type of fiber atrophy has varied widely in past reports. Generalized muscle fiber atrophy was reHE USE Supported in part by a grant from the Muscular Disease Society of Northeastern Ohio. MICHAEL G. SHEAHAN:Fourth year medical student, Case Western Reserve University School of Medicine, Cleveland, Ohio. PAUL J. VIGNOS, JR., M.D.: Associate Professor of Medicine, and ARmmms AND proportion of type II fibers also occurs. These changes take place before there is any signscant alteration in size of type I fibers, and before there is any apparent loss of either glycolytic or oxidative enzyme activity. ported by Faludi and co-workers in a histological study of mice treated with a variety of steroids, including 2 mg. of triamcinolone ( approximately 71 mg./ Kg. ) six times a week for 4 weeks.? DAgostino and Chiga, using both light and electron microscopy, reported primary involvement of granular (type I, red) muscle fibers in rabbits receiving 10 mg./Kg. of cortisone for 12-14 days8 Smith has observed primarily agranular (type 11, white) muscle fiber changes in histochemical studies of rabbits receiving 6-10 mg./Kg. triamcinolone for 3-15 days6 The present study tries to document the earliest changes in animal muscles associated with corticosteroid administration. Triamcinolone was administered to rabbits using dosage levels more comparable to those used clinically. It was hoped that administration of amounts of corticosteroid smaller than previously used might clarify the primary type of muscle fiber involvement. MATERIALS AND METHODS White, male New Zealand rabbits weighing approximately 2 Kg. each were divided into test Director, Arthritis Unit, University Hospitals o/ Cleuelund, Chelund, Ohio. Reprint requests should be addressed'io Or. Vignos at the University Hospitals of Cleveland, 2065 Adelbert Road, Clevelund, Ohio 44106. RHEUMATISM, VOL. 12, No. 5 (OCTOBER 1969) 491 492 SHEAHAN AND VIGNOS Table 1.-Effect Animal no. 1 2 3 4 5 6 7 8 9 Mean SD P of Corticosteroid on Muscle and Total Body Weight Initial body wt., Kg. Final body wt., Kg. Control Test Control 2.27 1.84 2.09 2.05 1.88 2.18 2.18 2.39 2.58 2.16 0.23 2.47 1.90 2.02 2.03 1.92 2.07 2.36 2.28 2.67 2.19 0.26 >0.8 Test 2.66 2.62 2.27 1.87 2.58 2.38 2.25 1.93 2.43 1.80 2.58 1.68 2.63 2.08 3.16 2.20 3.27 2.80 2.65 2.15 0.35 0.38 <0.05 and pair-fed control groups. The test rabbits received an injection of 1 mg./Kg. triamcinolone daily for 3 weeks. Injection was into the hamstring muscle group, alternating the side injected each day. The test rabbits and the pair-fed controls were sacrificed after 3 weeks by an intravenous injection of 180 mg. pentobarbital sodium. Muscles to be examined were immediately excised, weighed, and frozen in isopentane immersed in liquid nitrogen. The entire left and right soleus muscles were excised, and a defined section of each gastrocnemius muscle was utilized. The portion of the gastrocnemius excised was that part of the lateral head bounded by the peroneus nerve, the vena saphena parva, and the plantaris muscle. The weight of both soleus muscles was expressed as per cent of total body weight at time of sacrifice. The weight of the excised portion of both gastrocnemius muscles was similarly expressed. Frozen sections from the midportion of the gastrocnemius and soleus muscles were cut at -25OC in a cryostat and stained for fiber morphology using a modified trichrome reaction.9 Sections of each muscle were stained according to reported histochemical methods for phosphorylase,lOJl succinic dehydrogenase,lZ NADH diaphorase,13J* and myofibrillar A”F’ase.15 Sections of the appropriate control muscle were handled in an identical manner. Muscle fiber size was estimated by measuring the diameters of more than 30 fibers of each type ( I and 11) in at least five different high power fields. Only fibers which appeared to be cut in transverse section were measured. Maximum vertical and horizontal diameters of each fiber were used to obtain an average maximum diameter. Fibers on the edge of the section and those contracted from the e n d o m y s i h were not Gastrocnemius, % body wt. Control Test 0.33 0.19 0.27 0.18 0.30 0.17 0.29 0.15 0.26 0.20 0.22 0.21 0.33 0.15 0.31 0.19 0.31 0.18 0.29 0.18 0.04 0.02 <0.001 Soleus - % body &. Control Test 0.09 0.11 0.10 0.10 0.10 0.11 0.10 0.11 0.10 0.11 0.10 0.11 0.11 0.12 0.09 0.10 0.09 0.09 0.10 0.11 0.01 0.01 <0.05 measured. Although there are various ways to measure muscle fiber thickness, the determination of maximum fiber diameter was used in this study to minimize any changes caused by fiber atrophy. Since the atrophic fibers were found to assume what has been termed an “angular” rather than polygonal shape, the comparison of maximum diameters minimized the observed changes in fiber transverse sectional area. All measurements were made with an eyepiece micrometer calibrated with a stage micrometer. Distinction of fiber type was based on the ATPase (6 p sections incubated for 10 min.) or phosphorylase ( 16 p sections incubated for 10 min.) reactions. ATPase and phosphorylase staining reactions give good agreement in fiber diameter measurements. Only obvious type I and I1 fibers were used. RESULTS There was a s i g d c a n t difference ( p < 0.05) in the terminal body weight of the steroid-treated animals as compared to the controls. There was no difference in initial weights between the two groups (Table 1)There was a significant decrease in average weight of the gastrocnemius muscle in the steroid-treated animals when compared to the controls (Table 1). Conversely, the average weight of the soleus muscle from the steroid-injected animals showed a slight but s i g d c a n t increase in weight when compared to its control (Table 1). There was no difference in histologic ap- 493 EXPERIMENTAL CORTICOSTEROID MYOPATHY Table 2.-Muscle Animal no. 1 2 3 4 5 6 7 8 9 Mean SD P _.__ Fiber Type Sizes in Control and Corticosteroid-Treated Rabbits Gastrocnemius, p Type I TypeU Control Test Control Test 42.0 39.0 45.9 41.5 41.2 49.9 49.8 45.0 48.1 44.7 4.0 > __ 46.7 42.3 49.0 41.1 39.2 43.5 40.1 37.8 44.0 42.6 3.6 0.2 57.7 46.3 53.3 40.0 65.8 40.8 54.4 34.1 58.9 41.8 58.8 39.1 68.4 35.7 64.0 39.6 66.1 44.8 60.8 40.2 5.4 3.9 <0.001 pearance between the test and control muscles in the tissues stained with the modified trichrome reaction except for the appearance in the steroid-treated animals of randomly distributed slightly angular fibers, especially in the gastrocnemius muscle. The angular-shaped fibers were found to be histochemical type I1 fibers. In the gastrocnemius, the average size of the type I1 fiber in the steroid-treated animals is significantly smaller than the average control type I1 fiber. Similar changes are found in the soleus muscle, where the average size of the type I1 fibers in the steroid-treated animals is again significantly smaller than the average control type I1 fiber. There was no statistical difference between the average size of the injected and control type I fibers in gastrocnemius or soleus (Table 2). A sample photomicrograph of a control gastrocnemius muscle showed that the type I1 fibers in the control animal appeared to be larger than the type I fibers (Figure 1A). In the steroid-treated animals, the type I1 fibers were .either the same size as the type I fibers or smaller (Figure 1B). I n the control soleus muscle, the two fiber types have a variable size relationship, although in general the two fiber types appear to be of equal size (Figure - Soleus, p Type I Control Test 63.8 52.7 62.2 59.0 63.6 68.4 67.9 58.0 56.0 61.3 5.3 57.3 53.7 66.3 57.1 42.9 59.7 54.9 50.4 63.9 56.2 7.0 >O.l Type I1 C&&ol 64.0 49.7 67.7 64.2 68.1 55.3 72.6 66.3 65.0 63.7 7.0 <O.Ol Test 49.2 46.0 50.7 46.2 45.1 47.3 51.7 58.0 71.2 51.7 8.3 2A). The soleus type I1 fibers of the steroid-injected animals generally appeared somewhat smaller than the type I fibers (Figure 2B). The difference in size between the two fiber types was not as striking as in the gastrocnemius, although actual fiber measurement did show a significant change. The muscle from the steroid-treated animals did not show any decrease in enzyme activity as judged by the four histochemical stains for oxidative and glycolytic enzyme activity. The quality, intensity, and localization of all the stains used appeared similar in the injected and control rabbit muscles. The loss of phosphorylase activity reported to occur with higher doses of triamcinolonee was not observed at the corticosteroid dosage levels used in this study. DISCUSSION Normal human and animal muscles are composed of muscle fibers which can be differentiated according to color, physiologic response to stimulus, and enzyme content as judged by biochemical and histochemical technics. A number of muscle fiber classifications have been proposed. In this study, the histochemical classification of Dubowitz and Pearse16 and Engel" has been used. Type I fibers are red, slow in response, and have predominantly oxidative 494 SHEAHAN AND VIGNOS Fig. 1.-Control and corticosteroid-treated gastrocnemius muscle, ATPase reaction. (A) Control muscle. (B) Steroid-treated muscle. Dark fibers are type I1 and light fibers are type I. 126 X. EXPERIMENTAL CORTICOSTEROID MYOPATHY 495 Fig. 2.-Control and corticosteroid-treated soleus muscle, ATPase reaction. (A) Control muscle. (B) Steroid-treated muscle. Dark fibers are type I1 and light fibers are type I. 126 X. 496 metabolism. Type I1 fibers are white, fast, and are predominantly glycolytic. The results of this study indicate that the primary change in experimental steroid myopathy is a decrease in type I1 fiber size. The change was seen in both the gastrocnemius and soleus muscles but was more marked in the former. Smiths and DAgostino and ChigaR reported an apparent reversal in the normal size relationship between type I and I1 fibers in steroid-treated rabbit muscle. The larger fibers in the control animals were type 11, while the larger fibers in steroidtreated animals were type I. Fiber size measurements were not given by these investigators. The reversal of the normal fiber size relationship was interpreted by Smith as an atrophy of type I1 fibers. This interpretation is supported by the fiber size measurements reported in our study. DAgostino and Chiga suggested that the change primarily involved an increase in type I fiber size. No increase in type I fiber size was found in our experiments. Triamcinolone was used in both our study and that of Smith, while D’Agostino and Chiga used cortisone. Although it is possible that the different results may be explained by difference in steroids used, other studies in experimental steroid myopathy make this unlikely. Faludi and coworkers report that seven different steroids cause similar minimal changes in mouse muscle.r It has been suggested that steroid myopathy may be associated with an inhibition of muscle phosphorylase, resulting in an increased dependence on the mitochondrial enzyme systems of the fibers for energy production.6 Conversely, the mitochondrial oxidative phosphorylation system has been implicated in steroid myopathy.s Results of our study indicate that the decrease in size of the type I1 fibers occurs before any decrease in phosphorylase, oxidative, or other enzyme activity can be de- SHEAHAN AND VIGNOS tected by the histochemical reactions used in this study. The ratio of type I to type I1 fibers differs in the gastrocnemius and soleus muscles as it does in other animal musThe gastrocnemius is composed mainly of type I1 fibers, and the soleus is composed almost entirely of type I fibers. If the type I1 fibers are the principal fiber type undergoing atrophy, one would expect the gastrocnemius muscle to be more atrophic than the soleus. A comparison of the relative weight losses of the two muscles suggests that this is the case. Since muscle weight was calculated as per cent of total body weight, the weight of the gastrocnemius muscle decreased out of proportion to total body weight in the steroidtreated animals to produce a relative weight decrease in this muscle. The increase in weight of the soleus muscle in the steroid-treated animals is interpreted as a greater relative decrease in total body weight than in soleus weight, due to the high concentration of type I fibers in this muscle. The gastrocnemius and soleus muscles were chosen for study since each represents a muscle composed predominantly of one fiber type. Therefore, relative weight changes should be more apparent if one fiber type were preferentially affected by steroid treatment. Furthermore, since Smith reports that proximal muscles are affected by triamcinolone to a greater extent than distal muscles,6 it seemed preferable to study distal muscles in the hope that these muscles would show less severe alterations. This might make it easier to distinguish early and possibly primary alterations in muscle fiber types. “Type I1 fiber atrophy” has been reported in a number of human muscle diseases. These include cachetic atrophy, polymyositis, corticosteroid atrophy, subacute inclusion body encephalitis, and myasthenia gravk20 Patients with dissimilar 497 EXPERIMENTAL CORTICOSTEROID MYOPATHY chronic disease also have type I1 fiber atrophy.*l However, the majority of these last patients were also taking corticosteroids. It is possible that type I1 is the more vulnerable fiber type, and thus may be af- fected preferentially by many different disease processes acting through a common pathway or through multiple metabolic derangements. SUMMARIO IN INTERLINGUA Tractamento con triamcinolona in conilios in un dosage de 1 mg per kg produce un effect0 metabolic differential con affection predominante de fibras muscular del typo glycolytic, manifeste in un atrophia significative de fibras de typo 11. Occurre etiam un consequente atrophia de musculo continente un alte proportion de fibras de typo 11. Iste alterationes occurre ante que ulle significative alteration &mensional es marliteste in fibras de typo I e ante omne apparente perdita de activitate enzymatic de typo glycolytic o oxidative. REFERENCES 1. Dubois, E. L.: Triamcinolone in the treatment of systemic lupus erythematosus. J. A. M. A. 167:1590, 1958. 2. Freyberg, R. H., Bemtsen, C. A., Jr., and Hellman, L.: Further experiences with A 1,9 alpha fluoro, 16 alpha hydroxyhydrocortisone (triamcinolone) in treatment of patients with rheumatoid arthritis. Arthritis Rheum. 1:215, 1958. 3. MacLean, K., and Schw, P. H.: Reversible amyotrophy complicating treatment with fluorocortisone. Lancet 1:701, 1959. 4. Ellis, J. T.: Degeneration and regeneration in the muscles of cortisone-treated rabbits. Amer. J. Phys. Med. 34940, 1955. 5. Faludi, G., Mills, L. C., and Chayes, Z. W.: Effect of steroids on muscle. Acta Endocr. 45: 68, 1964. 6. Smith, B.: Histological and histochemical changes in the muscles of rabbits given the corticosteroid triamcinolone. Neurology 14:857, 1964. 7. Faludi, G., Gotlieb, J., and Meyers, J.: Factors influencing the development of steroidinduced myopathies. Ann. N. Y. Acad. Sci. 138: 61, 1966. 8. D’Agostino, A. N., and Chiga, M.: Cortisone myopathy in rabbits: a light and electron microscopic study. Neurology 16:257, 1966. 9. Engel, W. K., and Cunningham, G. G.: Rapid examination of muscle tissue: an improved trichrome method for fresh-frozen biopsy sections. Neurology 13:919, 1963. 10. Takeuchi, T., and Kuriaka, H.: Histochemical detection of phosphorylase in animal tissues. J. Histochem. Cytochem. 3: 153, 1955. 11. Takeuchi, T.: Histochemical demonstration of branching enzyme (amylo-l,lk 1,6-transglucosidase) in animal tissues. J. Histochem. Cytochem. 6:208, 1958. 12. Nachlas, M. M., Tsou, K. C., DeSouza, E., Cheng, C. S., and Seligman, A. M.: Cytochemical demonstration of succinic dehydrogenase by the use of a new p-nitrophenyl substituted ditetrazole. J. Histochem. Cytochem. 5:420, 1957. 13. Friede, R. L., Fleming, L. M., and Knoller, M.: A quantitative appraisal of enzyme histochemical methods in brain tissue. J. Histochem. Cytochem. 11:232, 1963. 14. Scarpelli, D. G., Hess, R., and Pearse, A. G. E.: Cytochemical localization of oxidative enzymes. I. Diphosphopyridine nucleotide diaphorase and triphosphopyridine nucleotide diaphorase. J. Biophys. Biochem. Cytol. 4:747, 1958. 15. Padykuda, H. A., and Herman, E.: The specificity of the histochemical method of adenosine triphosphatase. J. Histochem. Cytochem. 3: 170, 1955. 16. Dubowitz, V., and Pearse, A. G. E.: A comparative histochemical study of oxidative enzyme and phosphorylase activity in skeletal muscle. Histochemie 2: 105, 1960. 17. Engel, W. K.: The essentiality of histoand cytochemical studies of skeletal muscle in the investigation of neuromuscular disease. Neurology 12:778, 1962. 18. Bullard, H. H.: Histological as related to physiological and chemical differences in certain muscles of the cat. Hopkins Hosp. Rep. 18:323, 1919. 19. Ogata, T.: A histochemical study of the red and white muscle fibers: I. Activity of the succinoxydase system in muscle fibers. Acta Med. Okayama 12:216, 1958. 20. Engel, W. K., Brooke, M. H., and Nelson, P. G.: Histochemical studies of denervated or tenotomized cat muscle: illustrating difficulties in relating experimental animal conditions to human neuromuscular diseases. AM. N. Y. Acad. Sci. 138: 160, 1966. 21. Engel, W. K.: Muscle biopsy. C h . Orthop. 39:80, 1965.