- Email: [email protected]
The Belly-Off Sign: A New Clinical Diagnostic Sign for Subscapularis Lesions
The Belly-Off Sign: A New Clinical Diagnostic Sign for Subscapularis Lesions Markus Scheibel, M.D., Petra Magosch, M.D., Maria Pritsch, Ph.D., Sven Lichtenberg, M.D., and Peter Habermeyer, M.D.
Purpose: We describe a new clinical sign for subscapularis lesions that we call the belly-off sign and compare its diagnostic value with standard subscapularis tests (the lift-off test, internal rotation lag sign, and belly-press test/Napoleon sign) in cases of different types of subscapularis lesions. The belly-off sign represents the inability of the patient to maintain the palm of the hand attached to the abdomen with the arm passively brought into flexion and internal rotation. Type of Study: Case series. Methods: This descriptive and explorative study included 60 patients (mean age, 55.3 years) with clinical, surgical, and/or magnetic resonance imaging– confirmed evidence of isolated or combined lesions of the subscapularis musculotendinous unit. In all patients, a complete physical shoulder examination was performed and the results were compared with the intraoperative and magnetic resonance imaging findings. Results: The belly-off sign appeared to be more reliable than any other diagnostic test or sign in detecting isolated partial tears of the subscapularis tendon (group I), combined partial subscapularis and complete supraspinatus tendon tears (group II), and postoperative subscapularis insufficiency with mild atrophy of the upper aspect of the subscapularis muscle (group VII). In cases of complete/near complete subscapularis tears with or without supraspinatus and infraspinatus tendon tears (groups IV, V, and VI), the belly-off sign appeared to be as reliable as the belly-press test, internal rotation lag sign, and lift-off test. For assessing partial subscapularis tendon tears and insufficient external rotators (group III), the belly-off sign was less reliable than any other diagnostic test or sign. Conclusions: The belly-off sign represents a promising new clinical diagnostic sign for subscapularis lesions. In particular, subtle lesions of the upper subscapularis tendon and postoperative subscapularis insufficiencies can be detected by the belly-off sign in cases of intact external rotators. In cases of an advanced lack of external rotators, the belly-off sign becomes negative and loses its diagnostic value. We therefore conclude that the belly-off sign is attributable to an unbalanced transverse force couple with overwhelming of the external rotators of the shoulder. Level of Evidence: Level IV, case series with no, or historical, control group. Key Words: Physical examination—Shoulder—Rotator cuff—Subscapularis tendon—Belly-off sign.
D
ifferent clinical signs and diagnostic tests have been published in the literature to evaluate the integrity of the musculotendinous unit of the subscap-
From the Center for Musculoskeletal Surgery, Campus VirchowClinic, Charité, University Medicine Berlin (M. S.), Department of Shoulder and Elbow Surgery, ATOS–Clinic Heidelberg (P.M., S.L., P.H.); and the Institute of Biometry and Medical Informatics, University of Heidelberg (M.P.), Heidelberg, Germany. Address correspondence and reprint requests to Markus Scheibel, M.D., Augustenburgerplatz 1, 13353 Berlin, Germany. E-mail: [email protected] © 2005 by the Arthroscopy Association of North America 0749-8063/05/2110-4192$30.00/0 doi:10.1016/j.arthro.2005.06.021
ularis. With the lift-off test, described by Gerber and Krushell,1 the affected arm of the patient is internally rotated and extended, placing the hand on the lumbar region. The test is considered positive if the patient is unable to raise the arm posteriorly off the back. The presence of an internal rotation lag sign (IRLS) as described by Hertel et al.2 is evaluated from the same starting position. The affected arm of the patient is held by the examiner at almost maximal internal rotation. The dorsum of the hand is passively lifted away from the body until almost full internal rotation is reached. The patient is than asked to actively maintain this position. The sign is considered positive when lag occurs and the magnitude can be judged in degrees.
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 21, No 10 (October), 2005: pp 1229-1235
1229
1230
M. SCHEIBEL ET AL.
If the lift-off test or the IRLS cannot be tested reliably because of pain or restricted range of motion, the belly-press test is used in addition to assess the function of the subscapularis musculotendinous unit.3 With the belly-press test, the patient presses the abdomen with the hand flat and attempts to keep the arm in maximum internal rotation. The test is considered positive when the elbow drops in a posterior direction, internal rotation is lost, and pressure is exerted by extension of the shoulder and flexion of the wrist. This position is often considered as the Napoleon sign and can be used in predicting not only the presence of a subscapularis tear, but also its general size.4 We describe a new clinical diagnostic sign that we have termed the belly-off sign that adds to the clinical testing of the subscapularis musculotendinous unit. We compared its diagnostic value with that of the lift-off test, IRLS, and belly-press test/Napoleon sign in cases of different types of subscapularis lesions. METHODS Sixty patients (44 male and 16 female) with a mean age of 55.3 years (range, 26 to 83 years) and clinical, surgical, and/or magnetic resonance imaging (MRI)– confirmed evidence of a lesion of the subscapularis musculotendinous unit were included in this descriptive and explorative study. In all patients, a complete physical examination of both shoulders, standard radiographic examination, and MRI were performed. On physical examination, special attention was paid to the integrity of the rotator cuff and any restriction of active or passive range of motion. The clinical integrity of the external rotators of the shoulder (infraspinatus and teres minor) was assessed using the dropping and the hornblower sign.5 The dropping sign is performed with the shoulder placed in 0° of abduction and 45° of external rotation. The examiner holds the patient’s forearm in this position and instructs the patient to maintain this position. If the patient is unable to maintain this position, the arm drops back into the neutral position, and the dropping sign is considered positive. The hornblower sign is performed with the arm at 90° of abduction in the scapular plane. The elbow is than flexed to 90°, and the patient is than asked to rotate the arm externally. If this is not possible, the hornblower sign is said to be positive. The dropping sign has a 100% sensitivity and 100% specificity for irreparable tears of the infraspinatus, and the hornblower sign has a 100% sensitivity and 93% specificity for irreparable tears of the teres minor.5 To assess the function of the
FIGURE 1. Starting position for the evaluation of the belly-off sign: The affected arm of the patient is passively brought into flexion and maximum internal rotation with the elbow flexed to 90°. The elbow of the patient is supported by 1 hand of the examiner while the other hand places the palm on the abdomen.
subscapularis musculotendinous unit, the lift-off test, IRLS, and belly-press test/Napoleon sign were performed as described above. The lift-off test was graded positive if the patient was unable to raise the dorsum of the hand to the lumbar region. The IRLS was considered positive when a complete lag occurred; it was judged as intermediate when lag occurred but the patient was able to keep the hand off the lumbar region. The Napoleon sign was graded according to Burkhart and Tehrany with slight modification.4 The sign is considered negative if the patient can push the hand against the stomach with the wrist straight; intermediate if full extension cannot be achieved, but the angle of the wrist is less than 60°; and positive if the wrist is flexed 60° to 90° while pushing the hand against the abdomen. Finally, the presence of the belly-off sign was evaluated. The arm of the patient was passively brought into flexion and maximum internal rotation with the elbow 90° flexed. The elbow of the patient is supported by 1 hand of the examiner while the other hand brings the arm into maximum internal rotation placing the palm of the hand on the abdomen (Fig 1). The patient is then asked to keep the wrist straight and actively maintain the position of internal rotation as the examiner releases the wrist. If the patient cannot maintain this position, lag occurs and the hand lifts off the abdomen resulting in the belly-off sign (Fig 2).
THE BELLY-OFF SIGN
1231 RESULTS
Group I: Isolated Partial Subscapularis Tendon Tears Nine patients had an arthroscopically confirmed isolated partial tear of the upper subscapularis tendon. In all patients, the lift-off test, the IRLS, the Napoleon sign, and the belly-off sign could be evaluated. The lift-off test was found to be positive in 2 patients (22%). The IRLS was graded as intermediate in 3 patients (33%) and in the remaining patients it was negative. The Napoleon sign was graded as intermediate in 5 (56%) patients and negative in 4. All patients in this group had a positive belly-off sign (100%).
FIGURE 2. The patient is then asked to keep the wrist straight and actively maintain the position of internal rotation as the examiner releases the wrist while maintaining support at the elbow. If the patient cannot maintain this position, the hand lifts off the abdomen resulting in the belly-off sign.
To evaluate results, the patients were divided into 7 groups: group I, isolated partial tears of the subscapularis tendon; group II, combined partial subscapularis tendon tears and complete supraspinatus tendon tears; group III, combined partial subscapularis tendon tears and supraspinatus, infraspinatus, and teres minor tendon tears; group IV, isolated complete/near complete subscapularis tendon tears; group V, combined complete/near subscapularis tendon tears and supraspinatus tendon tears; group VI, combined complete/near complete subscapularis tendon tears and supraspinatus, infraspinatus, and teres minor tendon tears; and group VII, postoperative subscapularis insufficiency. Group VII consisted of patients who had previously undergone open instability surgery and had clinical signs of subscapularis insufficiency. In most of the patients in groups I to VI, the ultimate diagnosis was established during arthroscopic or open rotator cuff exploration. Patients in group VII had MRI performed in order to evaluate the musculotendinous defect of the subscapularis. MRI was performed in paracoronal, transaxial, and parasagittal planes in T1-weighted spin-echo sequences and T2-weighted fat-suppressed sequences. The contralateral side of all patients served as a control group. In cases of symptoms or clinical signs of a rotator cuff tear on the contralateral side, MRI was obtained for further evaluation.
Group II: Combined Partial Subscapularis Tendon Tears and Complete Supraspinatus Tendon Tears Eight patients had an arthroscopically confirmed partial subscapularis tendon tear in combination with a complete full-thickness supraspinatus tendon tear. The lift-off test and the IRLS could be evaluated in 6 patients. Two patients were unable to place the arm into the required starting position because of pain. In all patients, the Napoleon sign and the belly-off sign could be evaluated. The lift-off test was found to be positive in 2 patients (33%). The IRLS was graded as intermediate in 2 patients (33%) and positive in 1 patient (17%). In the remaining patients it was negative. The Napoleon sign was graded as intermediate in 3 patients (38%) and negative in 5 patients. All patients in this group had a positive belly-off sign (100%). Group III: Combined Partial Subscapularis Tendon Tears and Complete Supraspinatus, Infraspinatus, and Teres Minor Tendon Tears Nine patients had an arthroscopically confirmed partial subscapularis tendon tear with a posterosuperior massive cuff tear. Seven patients had an additional complete supraspinatus and infraspinatus tendon tear. Five patients showed an external rotation lag sign. The lift-off test was found to be positive in 2 patients (29%), and the IRLS was graded as intermediate in 4 patients (58%). The Napoleon sign was graded as intermediate in 4 patients (58%). The bellyoff sign was found to be positive in 1 patient (14%). Two patients had an additional teres minor insufficiency demonstrated by the hornblower sign. The liftoff test and the IRLS could not be evaluated. The
1232
M. SCHEIBEL ET AL.
Napoleon sign was graded as intermediate in 1 patient, whereas the belly-off sign was found to be negative. Group IV: Isolated Complete/Near Complete Subscapularis Tendon Tears Five patients had an isolated complete tear of the subscapularis tendon. Three patients had an isolated near complete tear with involvement of at least the upper two thirds of the subscapularis tendon. In 6 cases, the diagnosis had been established intraoperatively, in 2 cases by MRI. The lift-off test and the IRLS could be evaluated in 6 cases and were considered as positive in all of them (100%). In 2 acute cases, the required starting position could not be achieved because of a painful restricted range of motion. The Napoleon sign was graded as positive in 5 patients (63%) and intermediate in 3 patients (38%). All patients had a positive belly-off sign (100%). Group V: Combined Complete/Near Complete Subscapularis Tendon Tears and Complete Supraspinatus Tendon Tears Nine patients were intraoperatively diagnosed with a combined subscapularis and supraspinatus tendon tear. Three patients had a near complete and 6 patients had a complete subscapularis tendon tear. The lift-off test and the IRLS could be evaluated in 6 patients and were considered as positive in all of them (100%); in 3 patients, neither test could be performed because of pain. The Napoleon sign was graded as intermediate in 2 patients (22%) and positive in 7 patients (78%). The belly-off sign was positive in all patients (100%). Group VI: Complete Subscapularis Tendon Tears and Supraspinatus, Infraspinatus, and Teres Minor Tendon Tears Seven patients were intraoperatively diagnosed with a 3 to 4 tendon tear including subscapularis, supraspinatus, and infraspinatus, and in 1 case also the teres minor. The lift-off test and the IRLS could be evaluated in 5 patients and were considered as positive in all of them (100%). The Napoleon sign was graded as intermediate in 2 cases (29%) and positive in 5 cases (71%). The belly-off sign was positive in 5 cases (71%). In 2 patients with an additional lesion of the teres minor, it was negative. Group VII: Postoperative Subscapularis Insufficiency This group consisted of 10 patients who had undergone previous open instability surgery and had clinical
evidence of a subscapularis deficit. In all patients, MRI was performed to evaluate the musculotendinous unit of the subscapularis. Although all of them appeared to have an intact tendon on axial images, parasagittal views showed atrophy of the subscapularis muscle. In 7 cases, there was only mild atrophy; in 3 cases, marked atrophy with fatty degeneration of the upper part of the subscapularis muscle was found. In patients with mild atrophy, the lift-off test was positive in 2 patients (29%) and the IRLS positive in 3 patients (43%). The Napoleon sign was graded as intermediate in 4 patients (57%) and negative in 3 patients. The belly-off sign was positive in all patients (100%). In patients with complete atrophy and fatty degeneration of the upper part of the subscapularis muscle, the lift-off test was positive in all patients (100%). The IRLS was positive in 2 patients and the Napoleon sign in 1 patient. All patients had a positive belly-off sign (100%). Results for all groups are listed in Table 1. Control Group The contralateral side of each patient served as a control. Four patients were excluded from the control group because of impaired passive range of motion. None of the remaining patients in groups I, II, IV, and VII showed any clinical evidence of a subscapularis deficit. Two patients in group III had an intermediate Napoleon sign and a positive belly-off sign. MRI revealed an asymptomatic partial subscapularis and a complete supraspinatus tendon tear. One patient in group V and 2 patients in group VI had signs positive for a subscapularis and supraspinatus deficit, and MRI revealed a subscapularis and supraspinatus tendon tear. One patient in group IV had been treated with an inverse prosthesis after nonunion of a subcapital humeral fracture. Another patient had been treated with a fracture prosthesis due to a head-splitting fracture of the humeral head. Standard radiographs showed a complete resorption of the greater and lesser tuberosity, suggesting a complete lack of rotator cuff function. Both patients had a positive lift-off test, IRLS, and Napoleon sign. However, the belly-off sign was negative in both patients. DISCUSSION Different clinical tests and diagnostic signs have been described in the literature to evaluate the function and integrity of the subscapularis musculotendinous unit. The lift-off test has been reported to be very
THE BELLY-OFF SIGN TABLE 1.
1233
Results of the Lift-Off Test, IRLS, Napoleon Sign, and Belly-Off Sign in Patients With Different Types of Subscapularis Lesions Lift-Off Test
Group I: partial SSC tears (n ⫽ 9) Group II: partial SSC and complete SSP tears (n ⫽ 8)
Group III: partial SSC and complete SSP, ISP, and TM tears (n ⫽ 9)
Group IV: complete/near complete SSC tears (n ⫽ 8)
Group V: complete/near complete SSC and complete SSP tears (n ⫽ 9)
Group VI: complete/near complete SSC tears and complete SSP, ISP, and TM tears (n ⫽ 7)
Group VII: postoperative SSC insufficiency (n ⫽ 10)
IRLS
Napoleon Sign
Belly-Off Sign
Positive 2/9 Negative 7/9
Positive 0/9 Intermediate 3/9 Negative 6/9
Positive 0/9 Intermediate 5/9 Negative 4/9
Positive 9/9 Negative 0/9
Positive 2/6 Negative 4/6 Unable to perform 2/8
Positive 1/6 Intermediate 2/6 Negative 3/6 Unable to perform 2/8
Positive 0/8 Intermediate 3/8 Negative 5/8
Positive 9/9 Negative 0/9
Positive 2/7 Negative 5/7 Unable to perform 2/9
Positive 0/7 Intermediate 4/7 Negative 3/7 Unable to perform 2/9
Positive 0/9 Intermediate 4/9 Negative 5/9
Positive 1/9 Negative 8/9
Positive 6/6 Negative 0/6 Unable to perform 2/8
Positive 6/6 Intermediate 0/6 Negative 0/6 Unable to perform 2/8
Positive 5/8 Intermediate 3/8 Negative 0/8
Positive 8/8 Negative 0/8
Positive 6/6 Negative 0/6 Unable to perform 3/9
Positive 6/6 Intermediate 0/6 Negative 0/6 Unable to perform 3/9
Positive 7/9 Intermediate 2/9 Negative 0/9
Positive 9/9 Negative 0/9
Positive 5/5 Negative 0/5 Unable to perform 2/7
Positive 5/5 Intermediate 0/5 Negative 0/5 Unable to perform 2/7
Positive 5/7 Intermediate 2/7 Negative 0/7
Positive 5/7 Negative 2/7
Positive 5/10 Negative 5/10
Positive 5/10 Intermediate 0/10 Negative 5/10
Positive 1/10 Intermediate 4/10 Negative 5/10
Positive 10/10 Negative 0/10
Abbreviations: SSC, Subscapularis; SSP, supraspinatus; ISP, infraspinatus; TM, teres minor.
reliable in patients with complete subscapularis tears who were able to perform the test.1,3 Our study confirmed previous observations that the lift-off test is of limited value in patients who cannot bring the affected arm into the starting position required because of pain and/or restricted range of motion.4,6 Our study has also shown that, in cases of complete/near complete tears of the subscapularis tendon or complete subscapularis tendon tears including different types of external rotator deficits, the lift-off test remains a highly sensitive diagnostic tool. However, in 14 of our patients with partial tears of the subscapularis tendon and intact external rotators, the lift-off test was negative. Mondori et al.7 recently reported on their clinical findings in patients with partial or complete subscapularis tears. In the group with a complete tendon rupture, the lift-off test was found to be positive in all
cases, whereas in the partial-tear group, the lift-off test was found to be positive in only 40% of cases. Results of a published electromyographic study in healthy subjects have shown that the lift-off test was significantly superior in activating the lower part of the subscapularis compared with the belly-press test.8 Stefko et al.9 have found significant electromyographic activity in the potentially confounding shoulder girdle muscles (latissimus dorsi, teres major, pectoralis major, posterior deltoid, and triceps) during the performance of the lift-off test. These electromyographic data may explain the findings in our study and previously reported results that partial tears of the subscapularis tendon can be missed by the lift-off test.2,7 It seems that in cases of lesions of the upper subscapularis tendon, other potential internal rotators can compensate the structural defect.
1234
M. SCHEIBEL ET AL.
The IRLS was also designed to evaluate the integrity of the subscapularis musculotendinous unit. Hertel et al.2 found that the IRLS was as specific but more sensitive than the lift-off test for assessing the subscapularis. Partial ruptures of the subscapularis could be missed by the lift-off test but were detected by the IRLS. The authors concluded when the cranial part of the subscapularis starts to tear, the IRLS becomes positive. The lift-off test finally becomes positive when the subscapularis has completely failed. Our study could confirm these findings that partial tears of the subscapularis tendon can be detected by the IRLS but were missed by the lift-off test. However, in the majority of cases, the IRLS failed to detect these lesions in our evaluation. The belly-press test represents an excellent additional test for evaluating the function and integrity of the subscapularis musculotendinous unit. With the belly-press test, the extension and maximum internal rotation position that is often found to be painful in acute cases but is required for the lift-off test, and the evaluation of the IRLS can be avoided. This position is often called the Napoleon sign and has been graded into positive, intermediate, and negative test results by Burkhart and Tehrany.4 The authors found that the Napoleon sign correlated with the degree of subscapularis tear in that the tears involving the entire subscapularis tended to have a positive Napoleon sign, whereas tears involving the less than the upper half of the tended to have a negative Napoleon sign. Tears involving more than the upper two thirds of the tendon tended to have an intermediate result. According to these findings, the Napoleon sign has its limitation in detecting tears of the upper half of the subscapularis because these patients tended to have a negative result. This is in accordance with our findings that the Napoleon sign failed to detect partial subscapularis tendon tears in approximately half of the cases. Tokish et al.8 found the belly-press test to be superior to the lift-off test for activating the upper subscapularis muscle. These electromyographic data may explain the findings that the Napoleon sign was more reliable than the lift-off test or IRLS in detecting lesions of the upper subscapularis lesions. To our knowledge, the belly-off sign has not been described in the literature. We believe that it adds to the clinical testing of the integrity of the subscapularis musculotendinous unit. In patients with partial tears of the subscapularis tendon with or without supraspinatus tendon tears (groups I and II) including intact external rotators, as well as in cases of postoperative subscapularis insufficiency (group VII) with mild atrophy of the upper subscapularis muscle, the belly-off sign was found
to be more reliable than any other diagnostic test or sign. In cases of complete/near complete subscapularis tendon tears with or without supraspinatus and infraspinatus involvement (groups IV, V and VI), the belly-off sign appeared to be as reliable as the lift-off test, IRLS, or Napoleon sign. In patients with documented partial subscapularis tendon tears and a lack of external rotators (group III), in particular infraspinatus, the belly-off sign was found to be negative in most of the cases and thus failed to detect these lesions. Our results suggest that the inability of the patient to actively maintain the palm of the hand attached to the abdomen with the arm brought into flexion and internal rotation is due to an unbalanced transverse force coupled with overwhelming of the external rotators of the shoulder.10 This study suggests that there is a correlation between the magnitude of the bellyoff sign and the extent of the rotator cuff tear, but we were unable to perform an objective grading system. The external rotators of the shoulder are the posterior parts of the deltoid, the infraspinatus, and the teres minor. The infraspinatus and teres minor together contribute to more than 80% and the teres minor alone up 45% of the external rotation torque.11,12 Walch et al.5 showed that the teres minor alone can contribute enough power to external rotation to avoid the hornblower sign. Our study has shown that, in cases of complete subscapularis tendon tears, the teres minor alone can contribute enough power to lift the hand off the abdomen (group VI). Chang et al.11 mentioned that the supraspinatus, infraspinatus, and teres minor have remarkable activity during internal shoulder rotation. These so-called cocontractions are thought to be physiologic, providing dynamic stability to the glenohumeral joint, and they may also contribute to the belly-off sign. The role of the deltoid muscle in the belly-off phenomenon remains unclear. However, these data need to be determined in an electromyographic study. The subscapularis muscle originates from the subscapular fossa and inserts at the lesser tubercle of the humerus. Clinical, cadaveric, and electromyographic studies show the subscapularis muscle to have at least 2 separate innervations and functions.8,13-15 In a position of flexion and internal rotation, the origin and insertion points of the subscapularis are approximated. Hertel et al.2 already mentioned that any weakness of a musculotendinous unit under observation is emphasized the closer it is tested to the position of maximum possible shortening, because at this length the muscle acts at the greatest biologic and mechanical disadvantage. Therefore, the belly-off sign is similar to the IRLS. However, we believe that the position of flexion and internal rotation has some advantages compared with the extension
THE BELLY-OFF SIGN and internal rotation position required for the lift-off test and IRLS. First, it seems that during the flexion and internal rotation position required for the belly-off sign, other potential internal rotators have a limited role in compensating this position when the subscapularis starts to tear. Stefko et al.8 have shown remarkable activity of other potential internal rotators (latissimus dorsi, teres major, and pectoralis major) and extensors (triceps and posterior deltoid) during the lift-off test. These findings may also account for the IRLS. In addition, less activity of other internal rotators may be associated with the belly-press test and the belly-off sign. However, further electromyographic studies are needed to confirm these findings. Second, it seems that in the position of flexion and internal rotation the length-tension relationship of the subscapularis is of greatest disadvantage compared with the IRLS. Third, the position of flexion and internal rotation seems to be more tolerable for the patient compared with the starting position required for the lift-off test or IRLS. The evaluation of the belly-off sign is limited in patients with severe abdominal obesity when the arm cannot be brought into the required starting position. In these cases, it is helpful to flex the elbow more than 90° and to put the palm of the hand on the chest. Because the position of the elbow does not affect the rotation of the shoulder, a lift-off of the chest can be observed in patients with abdominal obesity. Another limitation represents an impaired passive range of motion in particular due to adhesive capsulitis. Although these patients were excluded from this study, we have observed patients with a positive Napoleon sign and a positive belly-off sign in cases of posterior capsular contractures without any evidence of subscapularis tendon tears. This has to be taken into account to avoid false-positive results. CONCLUSIONS This report represents a descriptive and explorative study of a new diagnostic sign, the belly-off sign, and describes its relative diagnostic value in comparison with the lift-off test, IRLS, and belly-press/Napoleon sign in cases of different types of subscapularis lesions. A valid statistical estimation of specificity, sensitivity, and positive or negative predictive values was not possible in this study because the different diagnostic tests and signs were only evaluated in patients with documented subscapularis lesions. These data are curently being determined in a blinded trial in patients presenting with shoulder pain. In summary, the belly-off sign represents an effi-
1235
cient and reliable new clinical diagnostic sign for detecting subscapularis lesions. In particular, subtle partial tears of the subscapularis tendon and postoperative subscapularis insufficiencies can be detected by the belly-off sign in cases of intact external rotators. In cases of an advanced lack of external rotators, the belly-off sign becomes negative and loses its diagnostic value to detect lesions of the subscapularis musculotendinous unit. Acknowledgment: The authors thank Rüdiger Himmelhan for the illustrations.
REFERENCES 1. Gerber C, Krushell RJ. Isolated tears of the subscapularis muscle. Clinical features in sixteen cases. J Bone Joint Surg Br 1991;73:389-394. 2. Hertel R, Ballmer FT, Lambert SM, Gerber C. Lag signs in the diagnosis of rotator cuff rupture. J Shoulder Elbow Surg 1996; 5:307-313. 3. Gerber C, Hersche O, Farron A. Isolated rupture of the subscapularis tendon. Results of operative repair. J Bone Joint Surg Am 1996;78:1015-1023. 4. Burkhart SS, Tehrany AM. Arthroscopic subscapularis tendon repair: Technique and preliminary results. Arthroscopy 2002; 17:454-463. 5. Walch G, Boulahia A, Calderone S, Robinson AHN. The “dropping” and “hornblower’s” signs in evaluation of rotator cuff tears. J Bone Joint Surg Br 1998;73:624-628. 6. Deutsch A, Altcheck D, Veltri D, Potter H, Warren RF. Traumatic tears of the subscapularis tendon. Clinical diagnosis, magnetic resonance imaging findings and operative treatment. Am J Sports Med 1997;25:13-22. 7. Mondori T, Nakagawa Y, Oshima M. An isolated rupture of the subscapularis tendon. Presented at the 9th International Congress on Surgery of the Shoulder, Washington, DC, May 2-5, 2004. 8. Tokish JM, Ellis HB, Decker MJ, Torry MR, Hawkins RJ. The belly-press test for physical examination of the subscapularis muscle: Electromyographic validation and comparison with the lift-off test. J Shoulder Elbow Surg 2003;12:427-430. 9. Stefko JM, Jobe FW, VanderWilde RS, Carden E, Pink M. Electromyographic and nerve block analysis of the subscapularis liftoff test. J Shoulder Elbow Surg 1997;6:347-355. 10. Burkhart SS. Arthroscopic treatment of massive rotator cuff tears. Clinical results and biomechanical rationale. Clin Orthop 1991;267:45-56. 11. Chang YW, Hughes RE, Su FC, Itoi E, An KN. Prediction of muscle force involved in shoulder internal rotation. J Shoulder Elbow Surg 2000;9:188-195. 12. Colachis SC, Strohm BR. Effects of suprascapular and axillary nerve blocks on muscle force in upper extremity. Arch Phys Med Rehabil 1971;52:22-29. 13. Kadaba MP, Cole A, Wootten ME, McCann PD, Reid M, Mulford G, April EW, Bigliani LU. Intramuscular wire electromyography of the subscapularis. J Orthop Res 1992;10:394-397. 14. Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coordination in the normal shoulder. Clin Orthop 1990;257: 76-85. 15. McCann PD, Cordasco FA, Ticker JB, et al. An anatomic study of the subscapular nerves: A guide for electromyographic analysis of the subscapularis muscle. J Shoulder Elbow Surg 1994;3:94-99.
Purpose: We describe a new clinical sign for subscapularis lesions that we call the belly-off sign and compare its diagnostic value with standard subscapularis tests (the lift-off test, internal rotation lag sign, and belly-press test/Napoleon sign) in cases of different types of subscapularis lesions. The belly-off sign represents the inability of the patient to maintain the palm of the hand attached to the abdomen with the arm passively brought into flexion and internal rotation. Type of Study: Case series. Methods: This descriptive and explorative study included 60 patients (mean age, 55.3 years) with clinical, surgical, and/or magnetic resonance imaging– confirmed evidence of isolated or combined lesions of the subscapularis musculotendinous unit. In all patients, a complete physical shoulder examination was performed and the results were compared with the intraoperative and magnetic resonance imaging findings. Results: The belly-off sign appeared to be more reliable than any other diagnostic test or sign in detecting isolated partial tears of the subscapularis tendon (group I), combined partial subscapularis and complete supraspinatus tendon tears (group II), and postoperative subscapularis insufficiency with mild atrophy of the upper aspect of the subscapularis muscle (group VII). In cases of complete/near complete subscapularis tears with or without supraspinatus and infraspinatus tendon tears (groups IV, V, and VI), the belly-off sign appeared to be as reliable as the belly-press test, internal rotation lag sign, and lift-off test. For assessing partial subscapularis tendon tears and insufficient external rotators (group III), the belly-off sign was less reliable than any other diagnostic test or sign. Conclusions: The belly-off sign represents a promising new clinical diagnostic sign for subscapularis lesions. In particular, subtle lesions of the upper subscapularis tendon and postoperative subscapularis insufficiencies can be detected by the belly-off sign in cases of intact external rotators. In cases of an advanced lack of external rotators, the belly-off sign becomes negative and loses its diagnostic value. We therefore conclude that the belly-off sign is attributable to an unbalanced transverse force couple with overwhelming of the external rotators of the shoulder. Level of Evidence: Level IV, case series with no, or historical, control group. Key Words: Physical examination—Shoulder—Rotator cuff—Subscapularis tendon—Belly-off sign.
D
ifferent clinical signs and diagnostic tests have been published in the literature to evaluate the integrity of the musculotendinous unit of the subscap-
From the Center for Musculoskeletal Surgery, Campus VirchowClinic, Charité, University Medicine Berlin (M. S.), Department of Shoulder and Elbow Surgery, ATOS–Clinic Heidelberg (P.M., S.L., P.H.); and the Institute of Biometry and Medical Informatics, University of Heidelberg (M.P.), Heidelberg, Germany. Address correspondence and reprint requests to Markus Scheibel, M.D., Augustenburgerplatz 1, 13353 Berlin, Germany. E-mail: [email protected] © 2005 by the Arthroscopy Association of North America 0749-8063/05/2110-4192$30.00/0 doi:10.1016/j.arthro.2005.06.021
ularis. With the lift-off test, described by Gerber and Krushell,1 the affected arm of the patient is internally rotated and extended, placing the hand on the lumbar region. The test is considered positive if the patient is unable to raise the arm posteriorly off the back. The presence of an internal rotation lag sign (IRLS) as described by Hertel et al.2 is evaluated from the same starting position. The affected arm of the patient is held by the examiner at almost maximal internal rotation. The dorsum of the hand is passively lifted away from the body until almost full internal rotation is reached. The patient is than asked to actively maintain this position. The sign is considered positive when lag occurs and the magnitude can be judged in degrees.
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 21, No 10 (October), 2005: pp 1229-1235
1229
1230
M. SCHEIBEL ET AL.
If the lift-off test or the IRLS cannot be tested reliably because of pain or restricted range of motion, the belly-press test is used in addition to assess the function of the subscapularis musculotendinous unit.3 With the belly-press test, the patient presses the abdomen with the hand flat and attempts to keep the arm in maximum internal rotation. The test is considered positive when the elbow drops in a posterior direction, internal rotation is lost, and pressure is exerted by extension of the shoulder and flexion of the wrist. This position is often considered as the Napoleon sign and can be used in predicting not only the presence of a subscapularis tear, but also its general size.4 We describe a new clinical diagnostic sign that we have termed the belly-off sign that adds to the clinical testing of the subscapularis musculotendinous unit. We compared its diagnostic value with that of the lift-off test, IRLS, and belly-press test/Napoleon sign in cases of different types of subscapularis lesions. METHODS Sixty patients (44 male and 16 female) with a mean age of 55.3 years (range, 26 to 83 years) and clinical, surgical, and/or magnetic resonance imaging (MRI)– confirmed evidence of a lesion of the subscapularis musculotendinous unit were included in this descriptive and explorative study. In all patients, a complete physical examination of both shoulders, standard radiographic examination, and MRI were performed. On physical examination, special attention was paid to the integrity of the rotator cuff and any restriction of active or passive range of motion. The clinical integrity of the external rotators of the shoulder (infraspinatus and teres minor) was assessed using the dropping and the hornblower sign.5 The dropping sign is performed with the shoulder placed in 0° of abduction and 45° of external rotation. The examiner holds the patient’s forearm in this position and instructs the patient to maintain this position. If the patient is unable to maintain this position, the arm drops back into the neutral position, and the dropping sign is considered positive. The hornblower sign is performed with the arm at 90° of abduction in the scapular plane. The elbow is than flexed to 90°, and the patient is than asked to rotate the arm externally. If this is not possible, the hornblower sign is said to be positive. The dropping sign has a 100% sensitivity and 100% specificity for irreparable tears of the infraspinatus, and the hornblower sign has a 100% sensitivity and 93% specificity for irreparable tears of the teres minor.5 To assess the function of the
FIGURE 1. Starting position for the evaluation of the belly-off sign: The affected arm of the patient is passively brought into flexion and maximum internal rotation with the elbow flexed to 90°. The elbow of the patient is supported by 1 hand of the examiner while the other hand places the palm on the abdomen.
subscapularis musculotendinous unit, the lift-off test, IRLS, and belly-press test/Napoleon sign were performed as described above. The lift-off test was graded positive if the patient was unable to raise the dorsum of the hand to the lumbar region. The IRLS was considered positive when a complete lag occurred; it was judged as intermediate when lag occurred but the patient was able to keep the hand off the lumbar region. The Napoleon sign was graded according to Burkhart and Tehrany with slight modification.4 The sign is considered negative if the patient can push the hand against the stomach with the wrist straight; intermediate if full extension cannot be achieved, but the angle of the wrist is less than 60°; and positive if the wrist is flexed 60° to 90° while pushing the hand against the abdomen. Finally, the presence of the belly-off sign was evaluated. The arm of the patient was passively brought into flexion and maximum internal rotation with the elbow 90° flexed. The elbow of the patient is supported by 1 hand of the examiner while the other hand brings the arm into maximum internal rotation placing the palm of the hand on the abdomen (Fig 1). The patient is then asked to keep the wrist straight and actively maintain the position of internal rotation as the examiner releases the wrist. If the patient cannot maintain this position, lag occurs and the hand lifts off the abdomen resulting in the belly-off sign (Fig 2).
THE BELLY-OFF SIGN
1231 RESULTS
Group I: Isolated Partial Subscapularis Tendon Tears Nine patients had an arthroscopically confirmed isolated partial tear of the upper subscapularis tendon. In all patients, the lift-off test, the IRLS, the Napoleon sign, and the belly-off sign could be evaluated. The lift-off test was found to be positive in 2 patients (22%). The IRLS was graded as intermediate in 3 patients (33%) and in the remaining patients it was negative. The Napoleon sign was graded as intermediate in 5 (56%) patients and negative in 4. All patients in this group had a positive belly-off sign (100%).
FIGURE 2. The patient is then asked to keep the wrist straight and actively maintain the position of internal rotation as the examiner releases the wrist while maintaining support at the elbow. If the patient cannot maintain this position, the hand lifts off the abdomen resulting in the belly-off sign.
To evaluate results, the patients were divided into 7 groups: group I, isolated partial tears of the subscapularis tendon; group II, combined partial subscapularis tendon tears and complete supraspinatus tendon tears; group III, combined partial subscapularis tendon tears and supraspinatus, infraspinatus, and teres minor tendon tears; group IV, isolated complete/near complete subscapularis tendon tears; group V, combined complete/near subscapularis tendon tears and supraspinatus tendon tears; group VI, combined complete/near complete subscapularis tendon tears and supraspinatus, infraspinatus, and teres minor tendon tears; and group VII, postoperative subscapularis insufficiency. Group VII consisted of patients who had previously undergone open instability surgery and had clinical signs of subscapularis insufficiency. In most of the patients in groups I to VI, the ultimate diagnosis was established during arthroscopic or open rotator cuff exploration. Patients in group VII had MRI performed in order to evaluate the musculotendinous defect of the subscapularis. MRI was performed in paracoronal, transaxial, and parasagittal planes in T1-weighted spin-echo sequences and T2-weighted fat-suppressed sequences. The contralateral side of all patients served as a control group. In cases of symptoms or clinical signs of a rotator cuff tear on the contralateral side, MRI was obtained for further evaluation.
Group II: Combined Partial Subscapularis Tendon Tears and Complete Supraspinatus Tendon Tears Eight patients had an arthroscopically confirmed partial subscapularis tendon tear in combination with a complete full-thickness supraspinatus tendon tear. The lift-off test and the IRLS could be evaluated in 6 patients. Two patients were unable to place the arm into the required starting position because of pain. In all patients, the Napoleon sign and the belly-off sign could be evaluated. The lift-off test was found to be positive in 2 patients (33%). The IRLS was graded as intermediate in 2 patients (33%) and positive in 1 patient (17%). In the remaining patients it was negative. The Napoleon sign was graded as intermediate in 3 patients (38%) and negative in 5 patients. All patients in this group had a positive belly-off sign (100%). Group III: Combined Partial Subscapularis Tendon Tears and Complete Supraspinatus, Infraspinatus, and Teres Minor Tendon Tears Nine patients had an arthroscopically confirmed partial subscapularis tendon tear with a posterosuperior massive cuff tear. Seven patients had an additional complete supraspinatus and infraspinatus tendon tear. Five patients showed an external rotation lag sign. The lift-off test was found to be positive in 2 patients (29%), and the IRLS was graded as intermediate in 4 patients (58%). The Napoleon sign was graded as intermediate in 4 patients (58%). The bellyoff sign was found to be positive in 1 patient (14%). Two patients had an additional teres minor insufficiency demonstrated by the hornblower sign. The liftoff test and the IRLS could not be evaluated. The
1232
M. SCHEIBEL ET AL.
Napoleon sign was graded as intermediate in 1 patient, whereas the belly-off sign was found to be negative. Group IV: Isolated Complete/Near Complete Subscapularis Tendon Tears Five patients had an isolated complete tear of the subscapularis tendon. Three patients had an isolated near complete tear with involvement of at least the upper two thirds of the subscapularis tendon. In 6 cases, the diagnosis had been established intraoperatively, in 2 cases by MRI. The lift-off test and the IRLS could be evaluated in 6 cases and were considered as positive in all of them (100%). In 2 acute cases, the required starting position could not be achieved because of a painful restricted range of motion. The Napoleon sign was graded as positive in 5 patients (63%) and intermediate in 3 patients (38%). All patients had a positive belly-off sign (100%). Group V: Combined Complete/Near Complete Subscapularis Tendon Tears and Complete Supraspinatus Tendon Tears Nine patients were intraoperatively diagnosed with a combined subscapularis and supraspinatus tendon tear. Three patients had a near complete and 6 patients had a complete subscapularis tendon tear. The lift-off test and the IRLS could be evaluated in 6 patients and were considered as positive in all of them (100%); in 3 patients, neither test could be performed because of pain. The Napoleon sign was graded as intermediate in 2 patients (22%) and positive in 7 patients (78%). The belly-off sign was positive in all patients (100%). Group VI: Complete Subscapularis Tendon Tears and Supraspinatus, Infraspinatus, and Teres Minor Tendon Tears Seven patients were intraoperatively diagnosed with a 3 to 4 tendon tear including subscapularis, supraspinatus, and infraspinatus, and in 1 case also the teres minor. The lift-off test and the IRLS could be evaluated in 5 patients and were considered as positive in all of them (100%). The Napoleon sign was graded as intermediate in 2 cases (29%) and positive in 5 cases (71%). The belly-off sign was positive in 5 cases (71%). In 2 patients with an additional lesion of the teres minor, it was negative. Group VII: Postoperative Subscapularis Insufficiency This group consisted of 10 patients who had undergone previous open instability surgery and had clinical
evidence of a subscapularis deficit. In all patients, MRI was performed to evaluate the musculotendinous unit of the subscapularis. Although all of them appeared to have an intact tendon on axial images, parasagittal views showed atrophy of the subscapularis muscle. In 7 cases, there was only mild atrophy; in 3 cases, marked atrophy with fatty degeneration of the upper part of the subscapularis muscle was found. In patients with mild atrophy, the lift-off test was positive in 2 patients (29%) and the IRLS positive in 3 patients (43%). The Napoleon sign was graded as intermediate in 4 patients (57%) and negative in 3 patients. The belly-off sign was positive in all patients (100%). In patients with complete atrophy and fatty degeneration of the upper part of the subscapularis muscle, the lift-off test was positive in all patients (100%). The IRLS was positive in 2 patients and the Napoleon sign in 1 patient. All patients had a positive belly-off sign (100%). Results for all groups are listed in Table 1. Control Group The contralateral side of each patient served as a control. Four patients were excluded from the control group because of impaired passive range of motion. None of the remaining patients in groups I, II, IV, and VII showed any clinical evidence of a subscapularis deficit. Two patients in group III had an intermediate Napoleon sign and a positive belly-off sign. MRI revealed an asymptomatic partial subscapularis and a complete supraspinatus tendon tear. One patient in group V and 2 patients in group VI had signs positive for a subscapularis and supraspinatus deficit, and MRI revealed a subscapularis and supraspinatus tendon tear. One patient in group IV had been treated with an inverse prosthesis after nonunion of a subcapital humeral fracture. Another patient had been treated with a fracture prosthesis due to a head-splitting fracture of the humeral head. Standard radiographs showed a complete resorption of the greater and lesser tuberosity, suggesting a complete lack of rotator cuff function. Both patients had a positive lift-off test, IRLS, and Napoleon sign. However, the belly-off sign was negative in both patients. DISCUSSION Different clinical tests and diagnostic signs have been described in the literature to evaluate the function and integrity of the subscapularis musculotendinous unit. The lift-off test has been reported to be very
THE BELLY-OFF SIGN TABLE 1.
1233
Results of the Lift-Off Test, IRLS, Napoleon Sign, and Belly-Off Sign in Patients With Different Types of Subscapularis Lesions Lift-Off Test
Group I: partial SSC tears (n ⫽ 9) Group II: partial SSC and complete SSP tears (n ⫽ 8)
Group III: partial SSC and complete SSP, ISP, and TM tears (n ⫽ 9)
Group IV: complete/near complete SSC tears (n ⫽ 8)
Group V: complete/near complete SSC and complete SSP tears (n ⫽ 9)
Group VI: complete/near complete SSC tears and complete SSP, ISP, and TM tears (n ⫽ 7)
Group VII: postoperative SSC insufficiency (n ⫽ 10)
IRLS
Napoleon Sign
Belly-Off Sign
Positive 2/9 Negative 7/9
Positive 0/9 Intermediate 3/9 Negative 6/9
Positive 0/9 Intermediate 5/9 Negative 4/9
Positive 9/9 Negative 0/9
Positive 2/6 Negative 4/6 Unable to perform 2/8
Positive 1/6 Intermediate 2/6 Negative 3/6 Unable to perform 2/8
Positive 0/8 Intermediate 3/8 Negative 5/8
Positive 9/9 Negative 0/9
Positive 2/7 Negative 5/7 Unable to perform 2/9
Positive 0/7 Intermediate 4/7 Negative 3/7 Unable to perform 2/9
Positive 0/9 Intermediate 4/9 Negative 5/9
Positive 1/9 Negative 8/9
Positive 6/6 Negative 0/6 Unable to perform 2/8
Positive 6/6 Intermediate 0/6 Negative 0/6 Unable to perform 2/8
Positive 5/8 Intermediate 3/8 Negative 0/8
Positive 8/8 Negative 0/8
Positive 6/6 Negative 0/6 Unable to perform 3/9
Positive 6/6 Intermediate 0/6 Negative 0/6 Unable to perform 3/9
Positive 7/9 Intermediate 2/9 Negative 0/9
Positive 9/9 Negative 0/9
Positive 5/5 Negative 0/5 Unable to perform 2/7
Positive 5/5 Intermediate 0/5 Negative 0/5 Unable to perform 2/7
Positive 5/7 Intermediate 2/7 Negative 0/7
Positive 5/7 Negative 2/7
Positive 5/10 Negative 5/10
Positive 5/10 Intermediate 0/10 Negative 5/10
Positive 1/10 Intermediate 4/10 Negative 5/10
Positive 10/10 Negative 0/10
Abbreviations: SSC, Subscapularis; SSP, supraspinatus; ISP, infraspinatus; TM, teres minor.
reliable in patients with complete subscapularis tears who were able to perform the test.1,3 Our study confirmed previous observations that the lift-off test is of limited value in patients who cannot bring the affected arm into the starting position required because of pain and/or restricted range of motion.4,6 Our study has also shown that, in cases of complete/near complete tears of the subscapularis tendon or complete subscapularis tendon tears including different types of external rotator deficits, the lift-off test remains a highly sensitive diagnostic tool. However, in 14 of our patients with partial tears of the subscapularis tendon and intact external rotators, the lift-off test was negative. Mondori et al.7 recently reported on their clinical findings in patients with partial or complete subscapularis tears. In the group with a complete tendon rupture, the lift-off test was found to be positive in all
cases, whereas in the partial-tear group, the lift-off test was found to be positive in only 40% of cases. Results of a published electromyographic study in healthy subjects have shown that the lift-off test was significantly superior in activating the lower part of the subscapularis compared with the belly-press test.8 Stefko et al.9 have found significant electromyographic activity in the potentially confounding shoulder girdle muscles (latissimus dorsi, teres major, pectoralis major, posterior deltoid, and triceps) during the performance of the lift-off test. These electromyographic data may explain the findings in our study and previously reported results that partial tears of the subscapularis tendon can be missed by the lift-off test.2,7 It seems that in cases of lesions of the upper subscapularis tendon, other potential internal rotators can compensate the structural defect.
1234
M. SCHEIBEL ET AL.
The IRLS was also designed to evaluate the integrity of the subscapularis musculotendinous unit. Hertel et al.2 found that the IRLS was as specific but more sensitive than the lift-off test for assessing the subscapularis. Partial ruptures of the subscapularis could be missed by the lift-off test but were detected by the IRLS. The authors concluded when the cranial part of the subscapularis starts to tear, the IRLS becomes positive. The lift-off test finally becomes positive when the subscapularis has completely failed. Our study could confirm these findings that partial tears of the subscapularis tendon can be detected by the IRLS but were missed by the lift-off test. However, in the majority of cases, the IRLS failed to detect these lesions in our evaluation. The belly-press test represents an excellent additional test for evaluating the function and integrity of the subscapularis musculotendinous unit. With the belly-press test, the extension and maximum internal rotation position that is often found to be painful in acute cases but is required for the lift-off test, and the evaluation of the IRLS can be avoided. This position is often called the Napoleon sign and has been graded into positive, intermediate, and negative test results by Burkhart and Tehrany.4 The authors found that the Napoleon sign correlated with the degree of subscapularis tear in that the tears involving the entire subscapularis tended to have a positive Napoleon sign, whereas tears involving the less than the upper half of the tended to have a negative Napoleon sign. Tears involving more than the upper two thirds of the tendon tended to have an intermediate result. According to these findings, the Napoleon sign has its limitation in detecting tears of the upper half of the subscapularis because these patients tended to have a negative result. This is in accordance with our findings that the Napoleon sign failed to detect partial subscapularis tendon tears in approximately half of the cases. Tokish et al.8 found the belly-press test to be superior to the lift-off test for activating the upper subscapularis muscle. These electromyographic data may explain the findings that the Napoleon sign was more reliable than the lift-off test or IRLS in detecting lesions of the upper subscapularis lesions. To our knowledge, the belly-off sign has not been described in the literature. We believe that it adds to the clinical testing of the integrity of the subscapularis musculotendinous unit. In patients with partial tears of the subscapularis tendon with or without supraspinatus tendon tears (groups I and II) including intact external rotators, as well as in cases of postoperative subscapularis insufficiency (group VII) with mild atrophy of the upper subscapularis muscle, the belly-off sign was found
to be more reliable than any other diagnostic test or sign. In cases of complete/near complete subscapularis tendon tears with or without supraspinatus and infraspinatus involvement (groups IV, V and VI), the belly-off sign appeared to be as reliable as the lift-off test, IRLS, or Napoleon sign. In patients with documented partial subscapularis tendon tears and a lack of external rotators (group III), in particular infraspinatus, the belly-off sign was found to be negative in most of the cases and thus failed to detect these lesions. Our results suggest that the inability of the patient to actively maintain the palm of the hand attached to the abdomen with the arm brought into flexion and internal rotation is due to an unbalanced transverse force coupled with overwhelming of the external rotators of the shoulder.10 This study suggests that there is a correlation between the magnitude of the bellyoff sign and the extent of the rotator cuff tear, but we were unable to perform an objective grading system. The external rotators of the shoulder are the posterior parts of the deltoid, the infraspinatus, and the teres minor. The infraspinatus and teres minor together contribute to more than 80% and the teres minor alone up 45% of the external rotation torque.11,12 Walch et al.5 showed that the teres minor alone can contribute enough power to external rotation to avoid the hornblower sign. Our study has shown that, in cases of complete subscapularis tendon tears, the teres minor alone can contribute enough power to lift the hand off the abdomen (group VI). Chang et al.11 mentioned that the supraspinatus, infraspinatus, and teres minor have remarkable activity during internal shoulder rotation. These so-called cocontractions are thought to be physiologic, providing dynamic stability to the glenohumeral joint, and they may also contribute to the belly-off sign. The role of the deltoid muscle in the belly-off phenomenon remains unclear. However, these data need to be determined in an electromyographic study. The subscapularis muscle originates from the subscapular fossa and inserts at the lesser tubercle of the humerus. Clinical, cadaveric, and electromyographic studies show the subscapularis muscle to have at least 2 separate innervations and functions.8,13-15 In a position of flexion and internal rotation, the origin and insertion points of the subscapularis are approximated. Hertel et al.2 already mentioned that any weakness of a musculotendinous unit under observation is emphasized the closer it is tested to the position of maximum possible shortening, because at this length the muscle acts at the greatest biologic and mechanical disadvantage. Therefore, the belly-off sign is similar to the IRLS. However, we believe that the position of flexion and internal rotation has some advantages compared with the extension
THE BELLY-OFF SIGN and internal rotation position required for the lift-off test and IRLS. First, it seems that during the flexion and internal rotation position required for the belly-off sign, other potential internal rotators have a limited role in compensating this position when the subscapularis starts to tear. Stefko et al.8 have shown remarkable activity of other potential internal rotators (latissimus dorsi, teres major, and pectoralis major) and extensors (triceps and posterior deltoid) during the lift-off test. These findings may also account for the IRLS. In addition, less activity of other internal rotators may be associated with the belly-press test and the belly-off sign. However, further electromyographic studies are needed to confirm these findings. Second, it seems that in the position of flexion and internal rotation the length-tension relationship of the subscapularis is of greatest disadvantage compared with the IRLS. Third, the position of flexion and internal rotation seems to be more tolerable for the patient compared with the starting position required for the lift-off test or IRLS. The evaluation of the belly-off sign is limited in patients with severe abdominal obesity when the arm cannot be brought into the required starting position. In these cases, it is helpful to flex the elbow more than 90° and to put the palm of the hand on the chest. Because the position of the elbow does not affect the rotation of the shoulder, a lift-off of the chest can be observed in patients with abdominal obesity. Another limitation represents an impaired passive range of motion in particular due to adhesive capsulitis. Although these patients were excluded from this study, we have observed patients with a positive Napoleon sign and a positive belly-off sign in cases of posterior capsular contractures without any evidence of subscapularis tendon tears. This has to be taken into account to avoid false-positive results. CONCLUSIONS This report represents a descriptive and explorative study of a new diagnostic sign, the belly-off sign, and describes its relative diagnostic value in comparison with the lift-off test, IRLS, and belly-press/Napoleon sign in cases of different types of subscapularis lesions. A valid statistical estimation of specificity, sensitivity, and positive or negative predictive values was not possible in this study because the different diagnostic tests and signs were only evaluated in patients with documented subscapularis lesions. These data are curently being determined in a blinded trial in patients presenting with shoulder pain. In summary, the belly-off sign represents an effi-
1235
cient and reliable new clinical diagnostic sign for detecting subscapularis lesions. In particular, subtle partial tears of the subscapularis tendon and postoperative subscapularis insufficiencies can be detected by the belly-off sign in cases of intact external rotators. In cases of an advanced lack of external rotators, the belly-off sign becomes negative and loses its diagnostic value to detect lesions of the subscapularis musculotendinous unit. Acknowledgment: The authors thank Rüdiger Himmelhan for the illustrations.
REFERENCES 1. Gerber C, Krushell RJ. Isolated tears of the subscapularis muscle. Clinical features in sixteen cases. J Bone Joint Surg Br 1991;73:389-394. 2. Hertel R, Ballmer FT, Lambert SM, Gerber C. Lag signs in the diagnosis of rotator cuff rupture. J Shoulder Elbow Surg 1996; 5:307-313. 3. Gerber C, Hersche O, Farron A. Isolated rupture of the subscapularis tendon. Results of operative repair. J Bone Joint Surg Am 1996;78:1015-1023. 4. Burkhart SS, Tehrany AM. Arthroscopic subscapularis tendon repair: Technique and preliminary results. Arthroscopy 2002; 17:454-463. 5. Walch G, Boulahia A, Calderone S, Robinson AHN. The “dropping” and “hornblower’s” signs in evaluation of rotator cuff tears. J Bone Joint Surg Br 1998;73:624-628. 6. Deutsch A, Altcheck D, Veltri D, Potter H, Warren RF. Traumatic tears of the subscapularis tendon. Clinical diagnosis, magnetic resonance imaging findings and operative treatment. Am J Sports Med 1997;25:13-22. 7. Mondori T, Nakagawa Y, Oshima M. An isolated rupture of the subscapularis tendon. Presented at the 9th International Congress on Surgery of the Shoulder, Washington, DC, May 2-5, 2004. 8. Tokish JM, Ellis HB, Decker MJ, Torry MR, Hawkins RJ. The belly-press test for physical examination of the subscapularis muscle: Electromyographic validation and comparison with the lift-off test. J Shoulder Elbow Surg 2003;12:427-430. 9. Stefko JM, Jobe FW, VanderWilde RS, Carden E, Pink M. Electromyographic and nerve block analysis of the subscapularis liftoff test. J Shoulder Elbow Surg 1997;6:347-355. 10. Burkhart SS. Arthroscopic treatment of massive rotator cuff tears. Clinical results and biomechanical rationale. Clin Orthop 1991;267:45-56. 11. Chang YW, Hughes RE, Su FC, Itoi E, An KN. Prediction of muscle force involved in shoulder internal rotation. J Shoulder Elbow Surg 2000;9:188-195. 12. Colachis SC, Strohm BR. Effects of suprascapular and axillary nerve blocks on muscle force in upper extremity. Arch Phys Med Rehabil 1971;52:22-29. 13. Kadaba MP, Cole A, Wootten ME, McCann PD, Reid M, Mulford G, April EW, Bigliani LU. Intramuscular wire electromyography of the subscapularis. J Orthop Res 1992;10:394-397. 14. Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coordination in the normal shoulder. Clin Orthop 1990;257: 76-85. 15. McCann PD, Cordasco FA, Ticker JB, et al. An anatomic study of the subscapular nerves: A guide for electromyographic analysis of the subscapularis muscle. J Shoulder Elbow Surg 1994;3:94-99.