One of the primary functions of the rotator cuff is to balance the force couples at the glenohumeral joint. Inman et al.¹ were the first to describe the coronal plane force couple around the shoulder joint. The coronal plane force couple represents the balance of moments created by the deltoid muscle and by the inferior portions of the rotator cuff (infraspinatus, teres minor, and subscapularis muscles). A second, and clinically more important, force couple is the transverse plane force couple.^2,3,4,5,6,7 In this force couple, the posterior infraspinatus and teres minor muscles balance the anterior subscapularis muscle. Balancing the transverse and coronal plane force couples is essential in providing a stable fulcrum for glenohumeral joint motion. Thus, a substantial tear of the subscapularis tendon can potentially lead to unbalanced force couples and concomitant dysfunction of glenohumeral joint motion.

Despite the fact that the subscapularis is the largest of the rotator cuff tendons, surprisingly little has been written on isolated or combined subscapularis tears.^{8,9,10,11,12,13,14,15,16,17,18,19,20,21} As advancements in technique and understanding have enabled surgeons to arthroscopically repair even massive rotator cuff tears, greater attention has been placed on identifying and arthroscopically repairing complete and partial thickness articular sided subscapularis tendon tears.

It is our belief that nonretracted tears of the subscapularis tendon are often not identified. Careful arthroscopic examination of the subscapularis tendon often reveals partial thickness articular surface tears or full thickness tears of the upper subscapularis that could not have been seen with open surgery.

In a cadaveric study, Sakurai et al.²² demonstrated that out of 20 shoulders with supraspinatus tears, 17 had concomitant partial thickness articular sided subscapularis tendon tears. All of these subscapularis tears appeared to have begun at the superior portion of the tendon where the greatest degree of tendon degeneration was noted. Bennett,²³ in a clinical study, reported that up to 27% of patients undergoing arthroscopy for rotator cuff, labrum, or ligament disorder have subscapularis tears.

The senior author (SSB) et al.²⁴ have recently studied the subscapularis tendon insertion footprint onto the lesser tuberosity of the proximal humerus in 19 cadavers. The superior to inferior length of the footprint was found to average 2.5 cm (range, 1.5 to 3.0 cm). The subscapularis footprint was noted to be trapezoidal in morphology, with its widest portion spanning the superior aspect of the insertion. The average width of the insertional footprint superiorly averaged 1.83 cm (range, 1.5 to 2.6 cm), whereas inferiorly the footprint narrowed to an average of 0.31 cm (range, 0.1 to 0.7 cm). The width of the subscapularis insertion remained constant until 1.48 cm below the superior margin of the footprint (i.e., the upper 60.4% of the subscapularis footprint was a constant width). We believe that repair of complete or partial tears of the superior portion of the subscapularis is important because, as this cadaver study has confirmed, the superior portion of the footprint is the largest and most substantial portion of the footprint, suggesting that the upper part of the subscapularis is the strongest part of the tendon. Therefore, we repair even small tears of the upper subscapularis.

In our clinical experience, there are no contraindications to arthroscopic subscapularis tendon repair in medically stable patients. Because the subscapularis tendon represents an
essential component of the transverse plane force couple, most patients complain of significant pain and weakness. Typically, patients present with internal rotation weakness and an increase in passive external rotation.

Several tests can be used to identify subscapularis tears in patients. The liftoff test is performed by having the patient place the hand of the affected arm on the back and asking the patient to lift the arm posteriorly off the back.^11,12,24a If the patient is unable to lift the arm posteriorly off the back, the test is considered positive. However, patients are often unable to perform the liftoff test due to pain or internal rotation contracture.

Another test that we use is the Napoleon (“belly-press”) test.^25,25a We perform this test by placing the hand on the belly similar to the position in which Napoleon held his hand for portraits. We grade the Napoleon test as negative (or normal) if the patient can push the hand against the stomach with the wrist straight; positive if the wrist must flex to 90 degrees to push against the stomach; and intermediate if the wrist is flexed from 30 degrees to 60 degrees to accomplish a belly press. In patients with weak internal rotation, the wrist flexes and the elbow drops back behind the trunk as the patient recruits the posterior deltoid to press the hand into the belly. We have found this test very useful in predicting the degree of subscapularis tearing.²⁶ Patients with positive Napoleon tests have tears of the entire subscapularis tendon; those with tears involving more than 50% of the tendon but not the entire tendon have intermediate Napoleon tests; and those with intact tendons or tears less than 50% may have a negative Napoleon test.