Acromion and Scapula Fractures After Reverse Total Shoulder Arthroplasty

Acromion and Scapula Fractures After Reverse Total Shoulder Arthroplasty

Ryan Colley, DO

Jonathan C. Levy, MD


Reverse total shoulder arthroplasty (RTSA) continues to have expanded indications for various shoulder pathology. Although originally RTSA was intended for rotator cuff arthropathy, it is now commonly used for acute fractures, massive irreparable rotator cuff tears without arthritis, severe arthritis with glenoid bone loss, and revision shoulder arthroplasty.1,2 In fact, RTSA is now being performed on younger patients with greater confidence.3,4 The increased utilization of RTSA has brought about a greater appreciation of complications and a focused effort to define strategies for prevention and management.5,6,7,8,9,10,11,12 Postoperative fractures of the acromion and scapula have surfaced as a complication uniquely more common to RTSA than other forms of shoulder arthroplasty.13 One of the difficulties with these fractures involves the seemingly elusive diagnosis. Although often times they are associated with sudden and dramatic impacts on pain and function, the presentation may be more subtle. Acromion fractures in the setting of RTSA have traditionally been treated without surgery as operative management is unpredictable. In this chapter, we hope to guide the shoulder surgeon to have a better understanding of how to identify and manage acromial and scapula fractures after RTSA.


Deltoid activity is critical to functional recovery following RTSA. The deltoid muscle has a broad origin across the lateral third of the clavicle, acromion, and scapular spine.11 It consists of three heads: anterior, middle, and posterior, which function as a shoulder flexor, abductor, and extensor, respectively. Following RTSA, the deltoid has a critical role in restoration of shoulder elevation14 and likely plays an impactful role in restoration of external rotation especially with increasing lateralization of the humerus.14,15,16,17,18 With all RTSA systems, the arm is typically lengthened,7,11,19 which increases the abductor moment arm of the deltoid,11,20 as the center of rotation of the shoulder is moved medially. The increased abductor moment improves the mechanical advantage of the deltoid to facilitate shoulder function. In addition, the increased arm length and deltoid tension provide a compression between the components enhancing stability of the semiconstrained RTSA prosthesis.11,21 When postoperative acromion fractures occur, the length-tension relationship of the deltoid is altered, as the acromion often tilts inferiorly (FIGURE 38.1). This displaces the deltoid origin, which will not only impact strength but also may create impingement between the acromion and greater tuberosity during attempted elevation. In addition, the loss of deltoid tension (and associated compression) has been reported to be associated with prosthetic instability (FIGURE 38.2).22 With acromion fractures being observed in various areas along the length of the acromion, the impact on function seems to be related to the amount of deltoid origin involved, with more medial fractures demonstrating worse functional outcomes.19


In order to have a better understanding of postoperative acromion fractures following RTSA, two classification systems have been described (FIGURE 38.3). Both of these classification systems are descriptive and based on anatomic location of the fracture. Crosby et al23 performed a retrospective review of 22 postoperative fractures and developed a system based on location of the fracture relative to the acromioclavicular (AC) joint. Type I fractures were described as small fractures along the anterior acromion, which includes the origin of the coracoacromial ligament. Type II fractures occur posterior to the AC joint. Finally, type III fractures are considered posterior acromion or base of the scapular spine. The hypothesis behind type III fractures involved a stress reaction from the tip of superior metaglene screw. The reproducibility of the Crosby classification has not been evaluated.

Levy et al19 described postoperative acromion fractures based on the location of the fracture relative to the deltoid origin (see FIGURE 38.3). This classification system was based on the principle that greater amounts of deltoid origin are impacted by more medial acromion fractures, with the greatest involvement of deltoid origin being involved in the base of the acromion fractures. Thus, the more medial the acromion fracture, the
higher the fracture type; the higher the fracture type, the more the deltoid origin was involved, resulting in worse outcomes. In the study that defined the classification system, a consecutive series of 18 patients with postoperative acromion or scapular spine pain was evaluated. Of the 18 patients, 7 had negative results on radiographs and required a computed tomography (CT) scan for further evaluation. Type I fractures were described as those through the midpart of the acromion involving the acromion origin of the anterior deltoid and a portion of the middle deltoid. Type II fractures involve the entire origin of the middle deltoid together with the acromion origin of the anterior deltoid. Type III fractures involve the anterior, middle, and posterior deltoid origin. The interobserver reliability and agreement were excellent in this study, helping to validate this classification system.

Despite the availability of two classification systems, many articles simply describe fracture location and determine whether it is related to trauma, stress fatigue, or spontaneous. Mayne et al24 rationalized that the bony pathology comprises a spectrum for which the earliest presentation is a stress reaction. If the process is able to
continue, a nondisplaced fracture may occur. Finally, in the worst situation, a complete displaced fracture may result. Clinical results deteriorate based on the location and displacement of the fracture.


Acromion fractures are often seen in patients prior to RTSA. With rotator cuff dysfunction, the superior directed forces created by the unopposed deltoid create stress across the acromion, often creating erosive changes, fractures, and fragmentation of the acromion. Preoperative acromion fractures have been reported in patients treated with RTSA.25 Among 457 consecutive patients treated with RTSA, Walch et al. reported a 9% incidence (41/457 patients) of preoperative acromion fractures, with 23 patients having an os acromiale, 17 with fragmentation of the acromion, and 1 having a nonunion of the scapular spine. Of interest, the impact on postoperative outcomes was minimal as Constant scores and subjective results were no different from those of patients without acromion pathology. Similar findings were reported by Aibinder et al26 who reported no impact on postoperative outcomes in patients with preoperative os acromiale. Although postoperative radiographs demonstrated acromion tilt, this did not have an impact on outcome when compared with patients without inferior tilt of the os acromiale. Thus, it is commonly accepted that preoperative acromion fractures can essentially be ignored and no change in operative planning is necessary when performing RTSA in these patients.

The true etiology of postoperative acromion or scapular spine fractures is not known. A multitude of factors have been implicated as contributing factors including age, osteoporosis, acromion wear, previous acromioplasty, humeral lengthening, glenoid component screw location and trajectory, retractor placement, release of the coracoacromial ligament, acromion length, delta angle, preoperative glenoid inclination, and greater tuberosity impingement. However, to date, few definitive risk factors have been identified. In a case-control study comparing patients with postoperative acromion fractures with controls, Otto et al27 examined the clinical risk factors in 53 patients with postoperative scapular fractures. The only significant risk factor identified was osteoporosis, which was present in 30.8% of patients with an acromion fracture. Screw position, smoking, body mass index, endocrine disease, chronic steroids, excessive alcohol use, and autoimmune disease were found to be no different between control and fracture cohorts. This observation was subsequently confirmed in a more recent study.28

Surgical indication also likely plays a role, especially considering that patients with rotator cuff tear arthropathy often have acromion wear and fragmentation. In a systematic review of the RTSA literature, acromion fractures were reported to be most common after RTSA
for inflammatory arthritis (10.9%) and massive rotator cuff tears (3.8%) and lowest for degenerative joint disease (2.7%), posttraumatic arthritis (2.1%), and acute fractures.29

Excessive humeral lengthening has been implicated as a potential contributor to acromial stress. Although evidence suggests that lengthening of the humerus is associated with neurologic symptoms following RTSA,30 it would make sense that excessive lengthening could contribute to postoperative acromion fractures as stress is transmitted to the acromion.7,9,19,24,27 Humeral lengthening is compounded by inferiorly translating the glenosphere, utilizing humeral components with valgus neck-shaft angles, and adding additional humeral-sided augmentations to the polyethylene insert. Furthermore, in an on-lay design implant when the polyethylene is placed on top of the anatomic neck osteotomy, higher rates of fractures of the base of the acromion spine were reported.31 In contrast, when the humeral shell was placed within the metaphysis of the proximal humerus (in-lay), the incidence of acromion fractures decreased by more than 50% (11%-4%) when compared with previously published work using an earlier-generation implant system (FIGURE 38.4).32

Use of a lateralized center of rotation has been thought to play a role in acromial stress fractures as well. Wong et al33 were the first to suggest this theory utilizing a fine element analysis (FEA) of 10 RTSA reconstructed cadaveric shoulders with a 38-mm glenosphere and a 155° neck-shaft angle. By varying glenoid inferior translation, lateral offset, and humeral lateralization at a 155° angle, this FEA model suggested that glenosphere lateralization increased stress by 17.2%, whereas humeral lateralization increased stress by only 1.7%. Conversely, humeral medialization decreased stress by 1.4%. The concentration of this stress occurred along the lateral aspect of the acromion in this FEA model. A recent systematic review of 90 articles noted that lateralized glenosphere designs had a significantly higher rate of acromion fractures when compared with medial glenosphere designs.23