The patient’s history, physical, and imaging findings are consistent with the diagnosis of rotator cuff arthropathy with pseudoparalysis in elevation. He had sufficient strength in external and internal rotation. Initial treatment was nonoperative including rehabilitation and NSAIDs. A subacromial injection of lidocaine and cortisone that resulted in immediate diminished pain and persisted for several weeks can be helpful in ruling out other causes of shoulder pain such as cervical radiculopathy and malingering. The usual indication for surgery is symptoms that persist over several months of nonoperative treatments in those with inability to lift the arm against gravity, known as pseudoparalysis in those with glenohumeral joint osteoarthritis and superior positioning of the humeral head on the glenoid. The osteoarthritis in patients with massive rotator cuff tears is usually different than those with an intact rotator cuff. The osteoarthritis usually involves the superior humeral head and the superior glenoid as a result of the superior position of the humeral head. Alternatively there can be concentric glenoid wear. Loss of glenoid bone can make positioning of the glenoid component more difficult and sometimes can make it impossible if it is severe. Inferior humeral head osteophytes, common in those with an intact rotator cuff, are uncommon in rotator cuff arthropathy. It is important to assess the acromion as it is sometimes thin, making it more prone to fracture after reverse TSA and the patient should be aware of this risk. When there is AC joint osteoarthritis as with this patient, I have been successful using the absence of tenderness at the AC joint and the absence of pain at the AC joint with cross-body motion, as is usually the case, in leaving the AC joint alone. When I started doing reverse TSA in the early 2000s I did it only in those more than 70 years of age. I now do the surgery in younger patients with good results [1, 2] but the vast majority of my patients with reverse TSA continue to be elderly.

With the patient in the Fowler’s position and the arm on a Mayo Stand a deltopectoral incision was made (Fig. 14.2). The deltopectoral interval was split and the clavipectoral fascia was incised lateral to the conjoined tendon. A self-retaining retractor was placed with one side under the conjoined tendon and the other under the deltoid tendon. With the arm at the side, the shoulder was then externally rotated to expose the subscapularis tendon insertion on the lesser tuberosity. As the superior two-thirds of the subscapularis tendon was not torn it was peeled off the lesser tuberosity (Fig. 14.3). A Fukuda retractor was placed into the joint and the shoulder was extended and externally rotated to expose the humeral head. The shoulder was pushed superior to expose more of the humeral head. The long head of the biceps tendon is almost always torn and in the rare instance when it is not a tenodesis can be done just below the bicipital groove. I used an intramedullary guide and a proximal humeral cutting guide to aid in the humeral osteotomy. The intramedullary guide was placed into the lateral humeral head 9 mm posterior to the bicipital groove which is the average. It is sometimes difficult to find the junction of the lateral humeral head and the greater tuberosity when the cuff arthropathy is chronic and the proximal humerus has a “bald eagle” appearance. Then a starting point is made so that the intramedullary guide will go straight down the humeral shaft. I positioned the guide in about 20° of retroversion and an osteotomy was made from anterior to posterior at a 155° angle with the shaft (Fig. 14.4). The humeral head was removed and the diameter was measured. Inferior humeral head osteophytes, although unusual, were removed with a rongeur. I then directed my attention to the glenoid. With the shoulder in abduction and external rotation the humeral osteotomy was placed posterior to the glenoid so that the surface of the humeral osteotomy was 90° to the surface of the glenoid (Fig. 14.5). If it will not stay, I retract the humerus with a Sonnabend or a Fukuda retractor. As the anterior capsule was intact it was incised at its insertion onto the labrum with a scalpel. An axillary nerve tug test was done by placing a finger along the inferior margin of the subscapularis muscle about 6 cm medial to the conjoined tendon while the deltoid muscle was tugged laterally. The axillary nerve was felt to tighten. It can be exposed and tagged. The inferior capsule was incised with a Bovie. The Bovie is used so that if it is close there will be stimulation of the axillary nerve alerting me. If there is any question of axillary nerve injury the tug test is repeated. A forked retractor was placed on the anterior glenoid neck. A finger was placed inferior to the glenoid to palpate the lateral border of the scapula. Sometimes a 1/2 in. periosteal elevator is used to remove the origin of the long head of the triceps tendon from the inferior glenoid if it prevents palpation of the lateral border of the scapula. A forked retractor was also placed on the lateral border of the scapula. A starting point was made so that the glenoid component will overhand the inferior glenoid by about 7 mm. This is usually a bit inferior to the center of the glenoid but varies depending on the glenoid size. The glenoid was reamed (Fig. 14.6) as was the hole for the center peg. The glenoid baseplate was impacted in place with its inferior hole aligned with the lateral border of the scapula. The inferior screw is the most important and I like it to be a locking screw at least 30–42 mm in length with good cortical fixation at its tip. I direct it posterior so that it exits the scapula at the infraspinatus fossa and not the subscapularis fossa. Anterior and posterior screws were placed. If needed, the superior screw is also a locking screw that I try to place into the base of the coaracoid so it does not exit the posterior glenoid where it may weaken the base of the scapula spine and contribute to scapula spine fracture. A trial glenosphere was placed on the glenoid baseplate. The humerus was then again dislocated in extension and external rotation. The humeral reaming guide was placed down the humeral canal in line with the humeral shaft (Fig. 14.7) and the proximal humerus was reamed. I prefer to cement the humeral prosthesis so I then placed a cement restrictor down the canal so that it was a centimeter or two below the prosthesis. For an uncemented humeral stem the surgeon can place the humeral trial so that there is good fixation in the humeral bone. I prefer to place as large a humeral cup as will fit to diminish impingement and to maximize range of motion. The prosthesis is reduced tightly so that there is only a millimeter or two of inferior shuck. As the shoulder had good passive range of motion and the arm could be placed at the side without the humeral prosthesis lifting off the trial glenosphere, the trial components were removed and the glenosphere impacted onto the glenoid baseplate so that it overhung the inferior glenoid bone by about 7 mm. A mildly lateral-based glenoid design may be more effective than prosthesis positioning in diminishing scapula notching [3]. A glenosphere with the Morse taper hole eccentric to its center can also aid in its placement. The humeral prosthesis was then cemented in place (Fig. 14.8). If the humeral component is not cemented, morselized cancellous bone from the humeral head can be placed in the canal of the humerus before the prosthesis to aid in fixation if needed. The subscapularis tendon was sewn back to the lesser tuberosity with sutures through bone and running, locking sutures in the tendon. The deltopectoral interval was closed followed by the skin.