- Email: [email protected]
Subscapularis muscle enervation: The effect of arm position
Subscapularis muscle enervation: The effect of arm position Sergio L. Checchia, MD, Pedro Doneux S., MD, Marcelo G. Martins, MD, and Flamarion S. Meirdes, MD, S~o Paulo, Brazil Extensive subscapularis muscle release is common in major shoulder reconstructive surgery, and manipulation of the muscle's anterior surface may damage its enervation. The subscapularis nerves and their points of entry into the muscle have rarely been described. The objective of this study was to define these points of entry into the subscapularis muscle and the possible modifications that may occur in such points according to the common positions in which the arm is placed during shoulder surgery, that is, internal, neutral, and external rotation. Twenty-five cadavers' shoulders were studied by an extensive anterior approach. The nerves were dissected and marked at their points of entry into the muscle. Measurements were taken of the distances of these nerves in the three previously mentioned arm positions. The points of entry of the subscapularis nerves into the muscle may be very close to the surgical field (as close as 1 cm from the glenoid border). Therefore extensive release of the anterior surface of the subscapularis muscle may cause damage to the subscapularis nerves. The risk can be potentially increased by the position of the arm, especially for the upper nerve when the arm is in external rotation. (J SHOULDERELBOWSURG 1996;5:214-8.) The shoulder has a large range of motion. The anatomic characteristics of this joint make it extremely dependent on its muscles for stability and movement. Therefore it is essential to obtain correct tension and balance of the muscles of the rotator cuff during surgery. Almost all the diseases that cause restriction to the shoulder movements have an internal rotation contracture. Examples of such diseases are degenerative arthritis, sequela of trauma, and the frozen shoulder. In these cases the subscapularis muscle goes through a progressive process of shortening and retraction. For this reason it is important to perform a proper lengthening of the subscapularis muscle during surgery to achieve good postoperative function of the shoulder. Studies have described methods of subscapularis lengthening during surgery. Z-plasty has been From Santa Casa Hospitals, School of Medicine, Department of Orthopedics and Traumatdogy, Sgo Paulo. Presented during the Eleventh Annual Closed Meeting of the American Shoulder and Elbow Surgeons, Manchester, Vermont, September 26-28, 1994. Reprint requests: Sergio L. Checchia, MD, Ah Brilhante 406 Alphaville 09, Santana do Parnaiba, Sao Paulo, Brazil CEP 06482-090. Copyright 9 1996 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/96/$5.00 + 0 3 2 / ! / 6 8 7 0 9
214
performed by fractional lengthening of the tendon 24 or by medial transfer of the tendon's insertion7 s These procedures have been associated with extensive and careful release of the anterior, posterior, superior, and inferior surfaces of the muscle. This release becomes very important in reoperations, where adherence of the subscapularis muscle to the base of the coracoid process and the conjoined tendon and to the neck of the scapula can be found. These attachments must be released to avoid postoperative restriction of motion. Recent literature 4' 24 also describes transfer of part of the subscapularis tendon, especially superiorly and laterally to close massive rotator cuff lesions. All these aggressive procedures manipulating the subscapularis muscle may cause injury to its enervation, provided by branches of the posterior fascicule of the brachial plexus, that is, roots C5 and C6, called the subscapularis nerves. In 1918 Kerr 19 studied 157 brachial plexuses. In all specimens ~ o subscapularis nerves were found. They were called the upper and lower subscapularis nerves. The upper nerve was found to be single in 53.5%, double in 40.76%, and triple in 5.73%. The lower subscapularis nerve varied greatly in its origin; it was found to be single in approximately 50% and double in the re-
J. Shoulder Elbow Surg. Volume 5, Number 3
maining specimens. Kerr's study showed that the lower nerve enervates only the axillary part of the muscle. Thus the upper subscapularis nerve regardless of the number of its branches is responsible for motor enervation of the subscapularis muscle. That study does not mention, however, the distance of these branches from any anatomic reference point. An extensive descriptive and topographic anatomic literature review was carried out, but most textbooks do not mention the subscapularis nerves.* Some authors describe the presence of ~ o subscapularis nerves.t Gardner et al. 11 were the only ones to mention the variation in the number of branches. None of the anatomic literature reviewed mentioned reference points for entry of these branches into the muscle. The recent interest in dynamic shoulder electromyography and subscapular nerve block has increased the importance of the anatomic position of the nerve branches in the subscapularis muscle.3, s. 15 In 1989 Kato 18 studied the nerve branches of the subscapularis muscle and had findings similar to those of Kerr. 19 In 1992 Kadaba et al. 17 described a technique for electromyographic study of this muscle by introducing two electrodes into the muscle, thus separating the upper and lower nerve branches. In 1994 McCann et al. 23 continued the previous study by dissecting 50 shoulders of 36 cadavers. They found three nerve branches in approximately 82% of shoulders, four nerve branches in 16%, and two nerve branches in 2% of the specimens studied. They also measured the mean distance between the branches and three bony points: the base of the coracoid process, the medial border of scapula (at the base of the scapula), and the anterior border of the glenoid. Nonetheless they did not specify the degree of rotation of the joint at the time measurements were taken. The objective of this article was to answer the following two questions. What is the distance between the point of entry of these nerves into the muscle mass and the border of the glenoid? Does this distance change according to the position of the shoulder, especially external rotation? METHODS
Twenty-five shoulders of 25 fresh cadavers without morbid rigidity were studied. The group's age *References 1, 6, 8-10, 12, 20, 26, 28. TReferences2, 7, 13, 14, 16, 21, 27, 29-31.
Checchia et al.
215
at death ranged from 35 to 88 years with a mean of 61.3 years; 64% were male, and 36% were female. Nineteen were white, and six were not white. The mean weight of the cadavers was 69.5 kg (range 54 to 100 kg). Their mean height was 171 cm (range 160to 180cm t. The dissection was performed alternating right and left shoulders regardless of the reported predominance with the cadavers in the supine position. An extended deltopectoral approach was used. After the deltopectoral sulcus was opened, part of the deltoid and pectoralis major origins were released from the acromion and clavicle, and the conjoined tendon was cut along with the lesser pectoral muscle at the level of the coracoid. After total exposure of the subscapularis muscle was achieved, the upper and lower subscapularis nerves were identified along with their branches. Identification marks were placed at the points where the nerves and their branches penetrated into the muscle. The subscapularis muscle was then opened following the direction of its fibers, along with the joint capsule, to expose the anterior border of the joint where another mark was placed at the 3 o'clock position (Figure 1). Measurements were taken with a millimetric ruler from the branches to the border of the glenoid with the shoulder in internal rotation (arm folded over the chest), neutral, and maximum external rotation (average 48 ~, range 40 ~ to 70~ Statistical analysis of the measurements was carried out with the Student t test to determine with a 95% confidence interval where the nerve would be located. Afterwards, the chi-squared test was used to correlate the distance with sex, weight, and height of the cadavers. This test was also used to determine whether the branches had significant differences in their distance from the anterior border of the glenoid according to the degree of rotation of the joint. RESULTS
In all shoulders at least two branches, the upper and the lower subscapularis nerves, were found. The upper branch was single in 14 shoulders, double in 9, and triple in 2. The lower branch was single in 21 shoulders and double in 4. The mean distance between the single or main upper subscapularis nerve and the border of the glenoid was 3.88 cm when the shoulder was in internal rotation fiR) with an SD of 0.75 cm, in neutral rotation (NR) the distance was 3.30 cm with an SD of 0.70 cm,
216
Checchia et al.
J. Shoulder Elbow Surg. May/June 1996
Figure 1 Motor enervation of subscapularis muscle and point at border of glenoid used for taking measurements IA). and in external rotation (ER) the distance was 2.54 cm with an SD of 0.74 cm. The secondary upper subscapularis nerve was found 4.16 cm from the glenoid border in IR (SD 1.18 cm), at 3.75 cm in NR (SD 1.08 cm), and at 3.22 cm in ER (SD 1.26 cm). The single or main lower subscapularis nerve was found 4.56 cm from the glenoid border in IR (SD 1.37 cm), at 4.06 cm in NR (SD 1.32 cm), and at 3.41 cm in ER (SD 1.27 cm). The secondary lower subscapularis nerve had distances from the glenoid border of 5.22 cm in IR, 4.60 cm in NR, and 3.88 cm in ER with corresponding SDs of 1.83 cm, 1.61 cm, and 1.26 cm (Table I; Figure 2). The statistical analysis of results was divided into three different stages. First, data such as height and weight were compared with the distance in measurements of the branches with the chi-squared test. These factors were not statistically significant (p > O. 1). Therefore the size of the patient does not appreciably alter the distance of the nerve branches from the border of the glenoid. During a second stage of analysis measurements were studied with the Student t test with an interval of 95%, that is, the distance where the nerve branch would be found with 95% of prob-
Figure 2 Rightshoulder x-ray film of cadaver with arm in external rotation. Note metal markers placed on branches of upper subscapularis nerve (upper marks) and on lower subscapularis nerve (lower mark). Also note how one of the branches is in close proximity to border of glenoid.
ability. This distance ranged from 2.41 to 5.35 cm for the main or single subscapularis nerve when in IR, from 1.93 to 4.67 cm in NR, and from 1.09 to 3.99 cm in ER. When the nerve was found to be double, the distances for the secondary upper subscapularis nerve ranged from 1.85 to 6.47 cm in IR, from 1.63 to 5.87 cm in NR, and from 0.75 to 5.69 cm in ER. For the lower subscapularis nerve the normal intervals were as follows: from 1.87 to 7.25 cm in IR, from 1.47 to 6.65 cm in NR, and from 0.92 to 5.90 cm in ER. Upper nerve tertiary branches were found in two cadavers, and secondary lower branches were found in four specimens. This small number did not allow statistical analysis (Table II). The third stage of statistical analysis of results was carried out by comparing the results of the Student t test with the distance of the nerves from
J. Shoulder Elbow Surg. Volume 5, Number 3
Checchia et al.
217
Table I Number of subscapularis nerve branches and the mean difference between the points of entrance of the nerves into the muscle mass and the border of the glenoid Mean distance (cm) Nerve Upper Main or single Secondary Lower Main or single Secondary
No. of branches
Internal rotation (SD)
Neutral rotation (SD)
External rotation (SD)
14 9
3.88 (0.75) 4.16 (1.18)
3.30 (0.70) 3.75 (1.08)
2.54 (0.74) 3.22 (1.26)
21 4
4.56 (1.37) 5.22 (1.83)
4.06 (1.32) 4.60 (1.61)
3.41 (1,27) 3.88 (1.26)
Table II Distance where the nerve branch would be found with 95% probability (Student t test) Interval (cm)-95% probability Nerve Upper Main or single Secondary Lower Main or single Secondary
Branches
Internal rotation
Neutral rotation
External rotation
2,41-5.35 1,85-6.47
1.93-4.67 1.63-5.87
1.09-3.99 0.75-5.69
1,87-7.25 --
1.47-6.65 --
0.92-5.90 --
the glenoid rim for the different positions of the arm (IR, NR, ER). This comparison was done with the chi-squared test. Statistical significance was found for the distances of the main upper subscapularis nerve when changing from IR to NR (p < 0.005) and from NR to ER (p< 0.01). No statistically significant difference was found for the distances of the secondary upper subscapularis nerve and the lower subscapularis nerves. DISCUSSION The main objective of this article was not that of a detailed anatomic description of the motor nerve branches of the subscapularis muscle. If such descriptions are sought, we recommend the papers of Kerr 19 and Kato. 18 Nonetheless these studies do not describe the points of entry of the nerves into the muscle in relation to a fixed anatomic point, one accessible to the surgeon. In 1994 McCann et al. 23 related these nerve branches to three points of the scapula, but only one of these points can be easily accessible through a standard deltopectoral approach. This point is the anterior border of the glenoid. McCann et al. were mainly concerned with placement of electromyography needles into the sub-
scapularis muscle and did not mention the position of the arm at the time measurements were taken from the specimens. It is known that the anatomic structures of the shoulder come closer to or farther away from the surgical field according to the position of the arm, especially in abduction/external rotation74 A lengthening or retraction of approximately 1 cm of the subscapularis tendon corresponds to 20 ~ rotation of the shoulder. 22 This rotation could bring the motor branches to the subscapularis nerves closer to or more distant from the operative field. In light of such findings the main objective of this study was to identify the factor or factors that could enhance the risk of surgically induced injury to the subscapularis nerves. One would think that the larger the patient is, the greater is the distance of these branches from the border of the glenoid. This association, however, proved not to be statistically significant. However, it was found that the main upper subscapularis branch penetrates the muscle as close as 1 cm from the medial border of the glenoid. When the nerve is double, the secondary branch may be found at less than 1 cm from this point. Therefore these nerve branches could be
218
Checchia et al.
highly susceptible to injury during surgery. Because the upper branch or branches are responsible for a great part of the motor enervation of the muscle, a lesion at this point would jeopardize the surgery's functional outcome. This problem can also occur with the lower subscapularis nerve, which was found to be as close as 1 cm from the anterior border of the glenoid. Statistical analysis correlates the points of entry of the nerve branches with the degree of rotation of the arm (internal, neutral, or external). All the nerve branches were closest to the anterior border of the glenoid when the arm was changed from internal to external rotation. The objective of this analysis was to verify whether this apparent approximation would be directly responsible for the increase in the risk of injury to the nerve. This was only found to be true for the main or single branch of the upper subscapularis, which presented a p < 0.005 for changes from internal to neutral rotation and a p < 0.01 for changes from neutral to external rotation. No statistically significant changes in position were found in the secondary branch of the upper subscapularis nerve or in its lower branch during rotation of the arm. Therefore the position of the arm during surgery probably does not substantially increase or decrease the risk of injury of these ~ o nerves. In conclusion, the motor nerve branches of the subscapularis muscle may be damaged during surgical procedures that demand extensive muscle release. The distance of the nerve branches to the anterior border of the glenoid may be as little as 1 cm. The position of the arm during surgery is responsible for changes in the distances from the glenoid rim of the points of entry of the nerve into the muscle. Thus an increased risk of injury, especially of the upper branch of the subscapularis nerve is presentwhen the arm is in external rotation. REFERENCES 1. Basmajian JV. Granl's method of anatomy: by regions de scriptive and deductive. Baltimore: The Williams & Wilkins Company, 1971. 2. Bateman JE. The shoulder and neck. Philadelphia: WB Saunders Company, 1972. 3. Chirona RL, HechtJS. Subscapular motor point block for the painful hemiplegic shoulder. Arch PhysMed Rehabil 1990; 71:4289. 4. Cofield RH. Subscapular muscle transposition for repair of chronic rotator cuff tears. Surg Gynecol Obstet 1982;154: 667 72. 5. Cole Aj, Kabada MP, Wootlen M, Reid MD. Eleclromyo graphic study of the subscapularis. Arch Phys Med Rehabil [abstract]. 1990;71:790.
J. Shoulder Elbow Surg. May/June 1996 6. DangelojG, FattiniCA. Anatomia humana e segmentar.S~o Paulo: LivrariaAtheneu, 1987. 7. DePalma AF. Cirugia dd hombro. Buenos Aires: Editorial M6dica Panamerica, 1985:242. 8. Didio LJA. Sinopse de anatomia. Rio deJaneiro: Guanabara Koogan, 1974. 9. Ellis H. Clinical anatomy. Oxford and Edinburg: Blackwell Scientific Publications, 1969. 10. Erhart E. Elementos de anatomia humana. S~o Paulo: Ath eneu, 1976. 11. Gardner E, Gray DJ, Q'Rahilly R. A regional study of human structure. Philadelphia: WB Saunders Company, 1963. 12. Gardner WD, Osburn WA. Structure of the human body. Philadelphia: WB Saunders Company, 1973:135. 13. Gray HFRS. Anatomia. Rio de Janeiro: Editora Guanabara Koogan SA, 1988:791. 14. Green JH, Silver PHS. An introduction to human anatomy. Oxford UniversiV Press, 1981:83. 15. HechtJS. Subscapular nerve block in lhe painful hemiplegic shoulder. Arch PhysMed Rehabil 1992;73:10369. 16. Hollinshead H. Sinopse de anatomia. Rio de Janeiro: Gua nabara Koogan, 1974:195. 17. Kadaba MP, Cole A, Wootten ME, Reid MD, April EW, Bighani UL. Intramuscularwire electromyography of the subscapularis. J Orthop Res 1992;10:394-7. 18. Kato K. Enervation of the scapular muscles and its morphoJogical significance in man. Anat AnzJena 1989;168:15568. 19. Kerr AT. The brachid plexus of the nerves in man: the variations in its formation and branches. Am J Anat 1918; 23:285 395. 20. Llorca FO. Anatomia humana. Barceloma: Editorial CienF fico-M6dica, 1959. 21. Lockhart RD, Hamilton GF, Fyfe FW. Anatomia humana. Rio de Janeiro: Editorial InteramericanaSA, 1959:287. 22. Malsen III FA, Thomas SC, Rockwood Jr CA. Anlerior glenohumeral instabili~/. In: Rockwood Jr CA, Matsen III EA, editors. The shoulder. Philadelphia: WB SaundersCompany, 1990:526 622. 23. McCann PD, Cordasco FA, TickerJB, et al. An anatomic study of the subscapularis nerves: a guide for electromyographic analysis of the subscapularis muscle. J SHOULDr:R EtBow SuRo 1994;3:94 9. 24. Neer CS. Shoulder reconstruction. Philadelphia: WB Saunders Company, 1990:116. 25. Rockwood Jr CA. Personal communication. San Antonio, Texas: 1988. 26. Romanes G. Cunningham's textbook of anatomy. Oxford UniversiV Press, 1964. 27. Rouvi@re H, Cordier G. Anatomia humana descriptiva y topografica. Madrid: Casa Editorial BaiNyBailliere SA, 1959. 28. Spence AP. Anatomia humana. Barcelona: Editorial Cienti ficoM6dica, 1959. 29. Testul L. Tratado de anatomia humana. Barcelona: Salvat Editores. SA, 1944:534. 30. Thompson E, Hayden A. Anatomy. Berwyn, Illinois: Physi cians Record Company, 1969:417. 31. Watterson PA, Taylor GI, CrockJG. The venous territoriesof muscles: analomical study and clinical implications. BrJ Hast Surg 1988;41:569 85.
214
performed by fractional lengthening of the tendon 24 or by medial transfer of the tendon's insertion7 s These procedures have been associated with extensive and careful release of the anterior, posterior, superior, and inferior surfaces of the muscle. This release becomes very important in reoperations, where adherence of the subscapularis muscle to the base of the coracoid process and the conjoined tendon and to the neck of the scapula can be found. These attachments must be released to avoid postoperative restriction of motion. Recent literature 4' 24 also describes transfer of part of the subscapularis tendon, especially superiorly and laterally to close massive rotator cuff lesions. All these aggressive procedures manipulating the subscapularis muscle may cause injury to its enervation, provided by branches of the posterior fascicule of the brachial plexus, that is, roots C5 and C6, called the subscapularis nerves. In 1918 Kerr 19 studied 157 brachial plexuses. In all specimens ~ o subscapularis nerves were found. They were called the upper and lower subscapularis nerves. The upper nerve was found to be single in 53.5%, double in 40.76%, and triple in 5.73%. The lower subscapularis nerve varied greatly in its origin; it was found to be single in approximately 50% and double in the re-
J. Shoulder Elbow Surg. Volume 5, Number 3
maining specimens. Kerr's study showed that the lower nerve enervates only the axillary part of the muscle. Thus the upper subscapularis nerve regardless of the number of its branches is responsible for motor enervation of the subscapularis muscle. That study does not mention, however, the distance of these branches from any anatomic reference point. An extensive descriptive and topographic anatomic literature review was carried out, but most textbooks do not mention the subscapularis nerves.* Some authors describe the presence of ~ o subscapularis nerves.t Gardner et al. 11 were the only ones to mention the variation in the number of branches. None of the anatomic literature reviewed mentioned reference points for entry of these branches into the muscle. The recent interest in dynamic shoulder electromyography and subscapular nerve block has increased the importance of the anatomic position of the nerve branches in the subscapularis muscle.3, s. 15 In 1989 Kato 18 studied the nerve branches of the subscapularis muscle and had findings similar to those of Kerr. 19 In 1992 Kadaba et al. 17 described a technique for electromyographic study of this muscle by introducing two electrodes into the muscle, thus separating the upper and lower nerve branches. In 1994 McCann et al. 23 continued the previous study by dissecting 50 shoulders of 36 cadavers. They found three nerve branches in approximately 82% of shoulders, four nerve branches in 16%, and two nerve branches in 2% of the specimens studied. They also measured the mean distance between the branches and three bony points: the base of the coracoid process, the medial border of scapula (at the base of the scapula), and the anterior border of the glenoid. Nonetheless they did not specify the degree of rotation of the joint at the time measurements were taken. The objective of this article was to answer the following two questions. What is the distance between the point of entry of these nerves into the muscle mass and the border of the glenoid? Does this distance change according to the position of the shoulder, especially external rotation? METHODS
Twenty-five shoulders of 25 fresh cadavers without morbid rigidity were studied. The group's age *References 1, 6, 8-10, 12, 20, 26, 28. TReferences2, 7, 13, 14, 16, 21, 27, 29-31.
Checchia et al.
215
at death ranged from 35 to 88 years with a mean of 61.3 years; 64% were male, and 36% were female. Nineteen were white, and six were not white. The mean weight of the cadavers was 69.5 kg (range 54 to 100 kg). Their mean height was 171 cm (range 160to 180cm t. The dissection was performed alternating right and left shoulders regardless of the reported predominance with the cadavers in the supine position. An extended deltopectoral approach was used. After the deltopectoral sulcus was opened, part of the deltoid and pectoralis major origins were released from the acromion and clavicle, and the conjoined tendon was cut along with the lesser pectoral muscle at the level of the coracoid. After total exposure of the subscapularis muscle was achieved, the upper and lower subscapularis nerves were identified along with their branches. Identification marks were placed at the points where the nerves and their branches penetrated into the muscle. The subscapularis muscle was then opened following the direction of its fibers, along with the joint capsule, to expose the anterior border of the joint where another mark was placed at the 3 o'clock position (Figure 1). Measurements were taken with a millimetric ruler from the branches to the border of the glenoid with the shoulder in internal rotation (arm folded over the chest), neutral, and maximum external rotation (average 48 ~, range 40 ~ to 70~ Statistical analysis of the measurements was carried out with the Student t test to determine with a 95% confidence interval where the nerve would be located. Afterwards, the chi-squared test was used to correlate the distance with sex, weight, and height of the cadavers. This test was also used to determine whether the branches had significant differences in their distance from the anterior border of the glenoid according to the degree of rotation of the joint. RESULTS
In all shoulders at least two branches, the upper and the lower subscapularis nerves, were found. The upper branch was single in 14 shoulders, double in 9, and triple in 2. The lower branch was single in 21 shoulders and double in 4. The mean distance between the single or main upper subscapularis nerve and the border of the glenoid was 3.88 cm when the shoulder was in internal rotation fiR) with an SD of 0.75 cm, in neutral rotation (NR) the distance was 3.30 cm with an SD of 0.70 cm,
216
Checchia et al.
J. Shoulder Elbow Surg. May/June 1996
Figure 1 Motor enervation of subscapularis muscle and point at border of glenoid used for taking measurements IA). and in external rotation (ER) the distance was 2.54 cm with an SD of 0.74 cm. The secondary upper subscapularis nerve was found 4.16 cm from the glenoid border in IR (SD 1.18 cm), at 3.75 cm in NR (SD 1.08 cm), and at 3.22 cm in ER (SD 1.26 cm). The single or main lower subscapularis nerve was found 4.56 cm from the glenoid border in IR (SD 1.37 cm), at 4.06 cm in NR (SD 1.32 cm), and at 3.41 cm in ER (SD 1.27 cm). The secondary lower subscapularis nerve had distances from the glenoid border of 5.22 cm in IR, 4.60 cm in NR, and 3.88 cm in ER with corresponding SDs of 1.83 cm, 1.61 cm, and 1.26 cm (Table I; Figure 2). The statistical analysis of results was divided into three different stages. First, data such as height and weight were compared with the distance in measurements of the branches with the chi-squared test. These factors were not statistically significant (p > O. 1). Therefore the size of the patient does not appreciably alter the distance of the nerve branches from the border of the glenoid. During a second stage of analysis measurements were studied with the Student t test with an interval of 95%, that is, the distance where the nerve branch would be found with 95% of prob-
Figure 2 Rightshoulder x-ray film of cadaver with arm in external rotation. Note metal markers placed on branches of upper subscapularis nerve (upper marks) and on lower subscapularis nerve (lower mark). Also note how one of the branches is in close proximity to border of glenoid.
ability. This distance ranged from 2.41 to 5.35 cm for the main or single subscapularis nerve when in IR, from 1.93 to 4.67 cm in NR, and from 1.09 to 3.99 cm in ER. When the nerve was found to be double, the distances for the secondary upper subscapularis nerve ranged from 1.85 to 6.47 cm in IR, from 1.63 to 5.87 cm in NR, and from 0.75 to 5.69 cm in ER. For the lower subscapularis nerve the normal intervals were as follows: from 1.87 to 7.25 cm in IR, from 1.47 to 6.65 cm in NR, and from 0.92 to 5.90 cm in ER. Upper nerve tertiary branches were found in two cadavers, and secondary lower branches were found in four specimens. This small number did not allow statistical analysis (Table II). The third stage of statistical analysis of results was carried out by comparing the results of the Student t test with the distance of the nerves from
J. Shoulder Elbow Surg. Volume 5, Number 3
Checchia et al.
217
Table I Number of subscapularis nerve branches and the mean difference between the points of entrance of the nerves into the muscle mass and the border of the glenoid Mean distance (cm) Nerve Upper Main or single Secondary Lower Main or single Secondary
No. of branches
Internal rotation (SD)
Neutral rotation (SD)
External rotation (SD)
14 9
3.88 (0.75) 4.16 (1.18)
3.30 (0.70) 3.75 (1.08)
2.54 (0.74) 3.22 (1.26)
21 4
4.56 (1.37) 5.22 (1.83)
4.06 (1.32) 4.60 (1.61)
3.41 (1,27) 3.88 (1.26)
Table II Distance where the nerve branch would be found with 95% probability (Student t test) Interval (cm)-95% probability Nerve Upper Main or single Secondary Lower Main or single Secondary
Branches
Internal rotation
Neutral rotation
External rotation
2,41-5.35 1,85-6.47
1.93-4.67 1.63-5.87
1.09-3.99 0.75-5.69
1,87-7.25 --
1.47-6.65 --
0.92-5.90 --
the glenoid rim for the different positions of the arm (IR, NR, ER). This comparison was done with the chi-squared test. Statistical significance was found for the distances of the main upper subscapularis nerve when changing from IR to NR (p < 0.005) and from NR to ER (p< 0.01). No statistically significant difference was found for the distances of the secondary upper subscapularis nerve and the lower subscapularis nerves. DISCUSSION The main objective of this article was not that of a detailed anatomic description of the motor nerve branches of the subscapularis muscle. If such descriptions are sought, we recommend the papers of Kerr 19 and Kato. 18 Nonetheless these studies do not describe the points of entry of the nerves into the muscle in relation to a fixed anatomic point, one accessible to the surgeon. In 1994 McCann et al. 23 related these nerve branches to three points of the scapula, but only one of these points can be easily accessible through a standard deltopectoral approach. This point is the anterior border of the glenoid. McCann et al. were mainly concerned with placement of electromyography needles into the sub-
scapularis muscle and did not mention the position of the arm at the time measurements were taken from the specimens. It is known that the anatomic structures of the shoulder come closer to or farther away from the surgical field according to the position of the arm, especially in abduction/external rotation74 A lengthening or retraction of approximately 1 cm of the subscapularis tendon corresponds to 20 ~ rotation of the shoulder. 22 This rotation could bring the motor branches to the subscapularis nerves closer to or more distant from the operative field. In light of such findings the main objective of this study was to identify the factor or factors that could enhance the risk of surgically induced injury to the subscapularis nerves. One would think that the larger the patient is, the greater is the distance of these branches from the border of the glenoid. This association, however, proved not to be statistically significant. However, it was found that the main upper subscapularis branch penetrates the muscle as close as 1 cm from the medial border of the glenoid. When the nerve is double, the secondary branch may be found at less than 1 cm from this point. Therefore these nerve branches could be
218
Checchia et al.
highly susceptible to injury during surgery. Because the upper branch or branches are responsible for a great part of the motor enervation of the muscle, a lesion at this point would jeopardize the surgery's functional outcome. This problem can also occur with the lower subscapularis nerve, which was found to be as close as 1 cm from the anterior border of the glenoid. Statistical analysis correlates the points of entry of the nerve branches with the degree of rotation of the arm (internal, neutral, or external). All the nerve branches were closest to the anterior border of the glenoid when the arm was changed from internal to external rotation. The objective of this analysis was to verify whether this apparent approximation would be directly responsible for the increase in the risk of injury to the nerve. This was only found to be true for the main or single branch of the upper subscapularis, which presented a p < 0.005 for changes from internal to neutral rotation and a p < 0.01 for changes from neutral to external rotation. No statistically significant changes in position were found in the secondary branch of the upper subscapularis nerve or in its lower branch during rotation of the arm. Therefore the position of the arm during surgery probably does not substantially increase or decrease the risk of injury of these ~ o nerves. In conclusion, the motor nerve branches of the subscapularis muscle may be damaged during surgical procedures that demand extensive muscle release. The distance of the nerve branches to the anterior border of the glenoid may be as little as 1 cm. The position of the arm during surgery is responsible for changes in the distances from the glenoid rim of the points of entry of the nerve into the muscle. Thus an increased risk of injury, especially of the upper branch of the subscapularis nerve is presentwhen the arm is in external rotation. REFERENCES 1. Basmajian JV. Granl's method of anatomy: by regions de scriptive and deductive. Baltimore: The Williams & Wilkins Company, 1971. 2. Bateman JE. The shoulder and neck. Philadelphia: WB Saunders Company, 1972. 3. Chirona RL, HechtJS. Subscapular motor point block for the painful hemiplegic shoulder. Arch PhysMed Rehabil 1990; 71:4289. 4. Cofield RH. Subscapular muscle transposition for repair of chronic rotator cuff tears. Surg Gynecol Obstet 1982;154: 667 72. 5. Cole Aj, Kabada MP, Wootlen M, Reid MD. Eleclromyo graphic study of the subscapularis. Arch Phys Med Rehabil [abstract]. 1990;71:790.
J. Shoulder Elbow Surg. May/June 1996 6. DangelojG, FattiniCA. Anatomia humana e segmentar.S~o Paulo: LivrariaAtheneu, 1987. 7. DePalma AF. Cirugia dd hombro. Buenos Aires: Editorial M6dica Panamerica, 1985:242. 8. Didio LJA. Sinopse de anatomia. Rio deJaneiro: Guanabara Koogan, 1974. 9. Ellis H. Clinical anatomy. Oxford and Edinburg: Blackwell Scientific Publications, 1969. 10. Erhart E. Elementos de anatomia humana. S~o Paulo: Ath eneu, 1976. 11. Gardner E, Gray DJ, Q'Rahilly R. A regional study of human structure. Philadelphia: WB Saunders Company, 1963. 12. Gardner WD, Osburn WA. Structure of the human body. Philadelphia: WB Saunders Company, 1973:135. 13. Gray HFRS. Anatomia. Rio de Janeiro: Editora Guanabara Koogan SA, 1988:791. 14. Green JH, Silver PHS. An introduction to human anatomy. Oxford UniversiV Press, 1981:83. 15. HechtJS. Subscapular nerve block in lhe painful hemiplegic shoulder. Arch PhysMed Rehabil 1992;73:10369. 16. Hollinshead H. Sinopse de anatomia. Rio de Janeiro: Gua nabara Koogan, 1974:195. 17. Kadaba MP, Cole A, Wootten ME, Reid MD, April EW, Bighani UL. Intramuscularwire electromyography of the subscapularis. J Orthop Res 1992;10:394-7. 18. Kato K. Enervation of the scapular muscles and its morphoJogical significance in man. Anat AnzJena 1989;168:15568. 19. Kerr AT. The brachid plexus of the nerves in man: the variations in its formation and branches. Am J Anat 1918; 23:285 395. 20. Llorca FO. Anatomia humana. Barceloma: Editorial CienF fico-M6dica, 1959. 21. Lockhart RD, Hamilton GF, Fyfe FW. Anatomia humana. Rio de Janeiro: Editorial InteramericanaSA, 1959:287. 22. Malsen III FA, Thomas SC, Rockwood Jr CA. Anlerior glenohumeral instabili~/. In: Rockwood Jr CA, Matsen III EA, editors. The shoulder. Philadelphia: WB SaundersCompany, 1990:526 622. 23. McCann PD, Cordasco FA, TickerJB, et al. An anatomic study of the subscapularis nerves: a guide for electromyographic analysis of the subscapularis muscle. J SHOULDr:R EtBow SuRo 1994;3:94 9. 24. Neer CS. Shoulder reconstruction. Philadelphia: WB Saunders Company, 1990:116. 25. Rockwood Jr CA. Personal communication. San Antonio, Texas: 1988. 26. Romanes G. Cunningham's textbook of anatomy. Oxford UniversiV Press, 1964. 27. Rouvi@re H, Cordier G. Anatomia humana descriptiva y topografica. Madrid: Casa Editorial BaiNyBailliere SA, 1959. 28. Spence AP. Anatomia humana. Barcelona: Editorial Cienti ficoM6dica, 1959. 29. Testul L. Tratado de anatomia humana. Barcelona: Salvat Editores. SA, 1944:534. 30. Thompson E, Hayden A. Anatomy. Berwyn, Illinois: Physi cians Record Company, 1969:417. 31. Watterson PA, Taylor GI, CrockJG. The venous territoriesof muscles: analomical study and clinical implications. BrJ Hast Surg 1988;41:569 85.