Clavicle

In the human fetus, as the sixth postconceptional week approaches, the clavicle develops two centers of ossification for its body. They join together approximately 1 week later and become the first bone to ossify in humans. In contrast to the other bones in the body, which form from cartilaginous precursors, clavicle ossification evolves from a mesenchymal or precartilaginous stage, sometimes termed membranous or dermal bone. Medial and lateral cartilaginous masses develop thereafter, resulting in linear growth of the clavicle, predominantly from the medial mass ( Fig. 9-1 ). A secondary center of ossification not present at birth appears in the medial end of the clavicle at 18 to 20 years, with union to the remaining clavicle before 25 years. A similar ossification center both appearing and uniting in the 20th year at the lateral end of the clavicle was described by Todd and D’Errico. Duplication of the clavicle, especially distally with the formation of two distinct acromioclavicular joints (ACJs), is an extremely rare phenomenon uniquely captured with computed tomography (CT) scan surface reconstruction ( Fig. 9-2 ).

Transverse section of an 11-week-old human fetus clavicle. Note the developing diaphysis and the lateral (l) and medial (m) cartilaginous masses. The shape at this early stage of development resembles the adult clavicle.

(From Ogata S, Uhthoff H. The early development and ossification of the human clavicle: an embryologic study. Acta Orthop Scand . 1990;61[4]:330-334.)

Computed tomography scan images of duplication of the acromioclavicular joint.

(From Viard B, Karp J-S, Tremlet J, et al. Unilateral duplication of the acromioclavicular joint: case report and literature review. Surg Radiol Anat. 2013;35[10]:973-977.)

Acromion and Coracoid Processes

The acromion and coracoid processes are cartilaginous at birth. As many as three secondary ossification centers have been reported for the coracoid process, which phylogenetically is a separate bone. These centers and two for the acromion appear between 13 and 16 years and typically unite to form a single bond between 14 and 20 years. Nonunions are possible at all sites but are most commonly seen in the acromion (os acromiale).

Overview

The human torso bears the upper limb by means of suspensory muscles to the scapula, clavicle, and humerus, maximizing the opportunity for movement. The muscles also serve as support for the sternoclavicular joint (SCJ), the limb’s only true articulation to the axial skeleton. The clavicle braces the upper limb at a fixed distance from the axial skeleton that permits optimal movement and power ( Fig. 9-3 ). This is possible only through the clavicle’s attachment to the scapula at the ACJ and at the coracoid process via the coracoclavicular (CC) ligaments. So linked to the axial skeleton, motion of the scapula on the thorax is possible only through the combined motions occurring via the sternoclavicular (SC) and ACJs.

The clavicle functions as an intercalary strut between the axial skeleton and the upper limb via its interconnecting ligaments with the scapula.

The ACJ is served by sensory branches from the suprascapular, axillary, and pectoral nerves. Its blood supply is derived from the acromial branch of the thoracoacromial artery, posterior humeral circumflex artery, and suprascapular (transverse scapular) artery. These vessels converge to form a smaller network (acromial rete) that overlies the acromion and penetrates the joint.

Skin

The innervation of the skin overlying the distal clavicle, ACJ, and acromion process is predominantly via the supraclavicular nerves, deriving from cervical roots C3 and C4. From anterolateral to posterolateral, the coverage can overlap with the upper lateral brachial cutaneous nerve, a branch of the axillary nerve derived from cervical root C5. Branches of the supraclavicular nerve are spared with skin incisions directly over the ACJ, while more medial skin incisions and dissection are likely to encounter the lateral branch consistently and sometimes the intermediate branch, if present.

Muscle Attachments

The anterior head of the deltoid muscle originates from the anterosuperior (AS), anterior, and anteroinferior (AI) aspects of the lateral clavicle; ACJ ligaments; and anterior acromion process. The trapezius inserts onto the posterosuperior (PS) aspect of the lateral clavicle, ACJ ligaments, and medial aspect of the acromion process. The actions of these two muscles contribute to the stability of the ACJ and aid the suspension of the upper limb from the clavicle and the axial skeleton. The coracobrachialis and the short head of the biceps brachii originate from the tip of the coracoid process. Proximal and inferomedial to the tip is the insertion of the pectoralis minor. Although not in its immediate vicinity, other muscles, including the pectoralis major, sternocleidomastoid, levator scapulae, rhomboid major and minor, and serratus anterior, have significant action on the ACJ.

Pertinent Osteology

Distal Clavicle

The S -shaped clavicle tapers from being a cylindrical bone medially to a more flattened bone laterally, where superior and inferior surfaces are distinguishable. The anterior portion of the lateral clavicle articulates with the acromion process of the scapula, and hence, the ACJ. On its inferior surface, points of attachment for the CC ligaments are delineated by a slightly prominent tubercle and a more subtle line for the conoid and trapezoid ligaments, respectively ( Fig. 9-4 ). Bone mineral density progressively increases from the most lateral aspect of the clavicle to an area of optimal bone density corresponding to the insertion area of the CC ligaments.

Osteology of the clavicle, from below. A, The attachments for the trapezoid (6) and the conoid (8) ligaments are outlined. 1, Sternohyoid; 2, costal tuberosity; 3, pectoralis major; 4, subclavius; 5, deltoid; 7, trapezius muscle. B, The inferior aspect of the clavicle.

(Modified from McMinn RMH, Hutchings RT. Color Atlas of Human Anatomy. Chicago, Year–Book Medical Publishers; 1985:93-94.)

Acromioclavicular Joint Architecture

Prior to the last 10 to 12 years, investigations of the morphology of the ACJ had been limited. The hyaline cartilage–covered convex lateral clavicle and the concave medial acromion articulate as a true diarthrodial joint. Whether it remains so throughout life is uncertain. Transitional changes of hyaline cartilage to fibrocartilage have been reported to occur on both sides of the joint between the second and third decades of life. In a cadaver study with the mean age at death of 81 years (range, 69 to 91 years), Hatta et al. documented that the articular cartilage in the lower half of both the clavicle and the acromion was significantly thinner than that in the upper half accompanied by more articular cartilage degeneration. From anterior to posterior, the changes were consistent.

The joint normally varies in size and shape with respect to adult age, gender, and skeletal morphology. Likewise, the orientation of the coronal plane of the joint is variable, ranging from nearly vertical to, less frequently, angulations from superior lateral to inferior medial approaching 50 degrees, a phenomenon that results in increasingly greater overriding of the lateral clavicle on the medial acromion ( Fig. 9-5 ). Although less common, angulations from superior medial to inferior lateral leading to underriding of the clavicle with respect to the acromion have also been reported. It is not unusual for some degree of vertical incongruence of the ACJ to exist in the complete absence of pathologic events. In a cadaver study that utilized dissections and CT scanning, Colegate-Stone et al. identified three main 3D shapes of the ACJ: flat, oblique, and curved ( Fig. 9-6 ). On the basis of the intra-articular disk, Emura et al. identified three primary types of ACJs and their incidence in a cadaver study (mean age, 82.9 years): type I, complete disk (3.8%); type II, incomplete disk (25%); and type III, absent disk (71.2%).

The sagittal plane orientation of the acromioclavicular joint shows great variability.

A and B, Morphology of the acromioclavicular joint.

( A, From Colegate-Stone T, Allom R, Singh R, et al. Classification of the morphology of the acromioclavicular joint using cadaveric and radiologic analysis. J Bone Joint Surg Br . 2010;92[5]:743-746. B, Courtesy Steven Lippitt, MD.)

The relatively lax joint capsule enables essential movement while resisting displacement of the acromion over the clavicle.

Ligaments

A relatively weak, thin capsule encloses the ACJ. Distinct acromioclavicular (AC) ligaments, named by their location—anterior, posterior, superior, and inferior—augment the strength of the capsule.

The CC ligament is actually two ligaments, the trapezoid and conoid ligaments ( Fig. 9-7 ). Despite their morphologic differences, the trapezoid and conoid ligaments have similar dimensional, viscoelastic behavioral, and structural properties.

The ligamentous anatomy around the acromioclavicular joint.

Both ligaments attach to the superior aspect of the angular region of the coracoid process. From a more anterior and lateral site that begins at the angle of the coracoid, courses in the direction of its tip, and ends posterior to the pectoralis minor tendon attachment, the trapezoid ligament is oriented toward the clavicle in a superior, anterior, and slightly lateral direction to attach at the trapezoid line on the inferior clavicle. From a more posterior and medial site on the angle, the conoid ligament expands in circumference as it runs superiorly and slightly medially toward its point of insertion, the conoid tubercle. Here, the lateral one-third and medial two-thirds of the clavicle join to form the posterior curve of the clavicle; at the curve’s apex, the conoid tubercle can be identified on the posteroinferior clavicle, just posterior and medial to the most medial aspect of the trapezoid line (see Fig. 9-4 ).

Distal to the angle, on the lateral aspect of the more horizontal part of the coracoid, is the attachment of the CAL ( Fig. 9-8 ). It joins onto nearly the entire width of the acromion process at its AI edge.

The coracoacromial ligament attachment to the superolateral aspect of the coracoid process.

(Modified from Fealy S, April EW, Khazzam M, et al. The coracoacromial ligament: morphology and study of acromial enthesopathy. J Shoulder Elbow Surg . 2005;14[5]:542-548.)

Rios et al. showed that regardless of gender or race, the length of the clavicle can serve as a gauge for the determination of the anatomic attachments of the trapezoid and conoid ligaments. Reliable anatomic points in osteology specimens and fresh specimens were used to measure to the center of the trapezoid ligament and the medial edge of the conoid ligament. The distance from the lateral edge of the clavicle to the respective ligament point of reference was measured and used to express a ratio of the measured distance to the clavicle length. This value was 0.17 for the trapezoid and 0.31 for the conoid. In another cadaver study, Renfree et al. determined that there were no gender differences for the measurement from the center of the articular surface to the lateral-most point of the trapezoid ligament, but the lateral-most point of the conoid ligament was more medial in men than in women. The only significant gender variation identified by Takase was in the coronal dimension of the undersurface of the clavicle and the sagittal dimension of the trapezoid ligament attachment.

Salzmann et al. demonstrated coracoid anatomic landmarks that facilitated identification of the dimensions and orientation of the insertional footprint of the CC ligaments ( Fig. 9-9 ). In a similar anatomic study, Xue et al. determined that the location of the center points of the footprints could be accurately expressed by the ratios of the conoid and trapezoid centers to the dimensions of the clavicle and coracoid process. Similarly, Takase precisely defined the anatomic parameters of the two CC ligaments. Knowledge of these landmarks may prove useful intraoperatively when anatomic reconstructions are being considered.

A and B, The coracoclavicular ligaments with their footprint on the superior left coracoid process. cla, clavicle; con, conoid ligament; cor, coracoid process; tra, trapezoid ligament.

(From Salzmann G, Paul J, Sandmann G, et al. The coracoid insertion of the coracoclavicular ligaments: an anatomic study. Am J Sports Med . 2008;36[12]:2392-2397.)

By way of its attachments to the clavicle and scapula, and their respective relationships to the axial skeleton and torso, the CC ligament helps to link scapulohumeral motion and scapulothoracic motion. These two ligaments create a much stronger union of the clavicle to the scapula than the AC ligaments. This ensures that the clavicle and scapula will move simultaneously as the ligaments reach maximum tension, signifying the limits of the natural excursion of the ACJ. The functional strength of the ACJ increases in the presence of intact CC ligaments. Thus the effective suspension of the upper limb from the torso via the SCJ and supporting musculature is assured.

Although it is tempting to simplify the function of the AC and CC ligaments as anterior-to-posterior and superior-to-inferior stabilizers, respectively, displacement resistance of the lateral clavicle is significantly more complex, as borne out in the results of experimental cadaver investigations. Not only are the ligaments distinguishable as discrete anatomic structures, the individual roles they play biomechanically are becoming more clear. Collectively, in the midst of a wide range of load, displacement, and restraint variables, these studies have shown that anterior restraint is afforded by the inferior AC ligaments and conoid ligament, posterior restraint predominantly by the superior and posterior AC ligaments and the trapezoid ligament, superior restraint by the conoid ligament, and ACJ, compression restraint by the trapezoid ligament. The results of the in vitro studies of Dawson et al. suggested the importance of the CC ligaments for equally restraining both anterior-posterior (AP) and superior-inferior movements, while the restraint of the AC ligaments was threefold greater in the AP plane than the superior-inferior plane. In their study, joint compression contributed to joint stability in a reverse linear fashion, signifying the importance, at least experimentally, of the distal clavicle articulation with the acromion.

Considerably more attention is being paid to the investigation of horizontal instability of the ACJ. Recent research that has included the serial ligament section and tracking measurement has brought this concept to light by demonstrating serially increased translations of the clavicle posteriorly or the acromion anteriorly with sectioning of the AC ligaments and further translations with sectioning of the CC ligaments. The clinical implications of the resultant excessive translations and altered articulation between the medial clavicle and lateral acromion cannot be overlooked, and may help to explain patterns of pain and dysfunction after such injuries.

Experimental sectioning of either the conoid or trapezoid ligaments had no effect on the overall strength of the C ligament. Complete dislocation of the lateral clavicle superior to the acromion process requires the interruption of the integrity of the conoid and trapezoid ligaments. In an experimental study loading shoulders to failure, ligament disruptions were more common than fractures. In another, the conoid was the first to fail. Age, height, and weight showed no significant relationship with the ligament cross-sectional area or the mechanical properties of the shoulder. It has been shown in finite element analysis based on human CT scan information that during progressive abduction, the conoid ligament progressively lengthens while the trapezoid ligament remains relatively unchanged but ultimately becomes lax at full abduction.

Coracoclavicular Articulation

Although lesser bursae can interdigitate between the layers of the conoid and trapezoid ligaments, a more prominent bursa can exist between the ligaments, sometimes sufficiently expanding to create a CC joint between the coracoid process and the clavicle. Twenty cases of such an articulation were documented by dissection or with radiographs in 1939. The gross and histologic appearance is that of a diarthrodial joint with cartilage-covered articulating surfaces and a synovial-lined capsule, and correlates with its radiographic properties. The incidence of this unusual shoulder articulation is probably around 0.5% but, depending on geographic region, can vary widely from 0.04% to 27%, and it is often bilateral ( Fig. 9-10 ). The incidence appears to increase with age. It can develop in response to restricted movement of the scapula caused by variations in the shapes of the scapulae, clavicles, and first ribs.

Coracoclavicular joint. A, Three-dimensional computed tomographic scan of shoulder girdles showing bilateral coracoclavicular joints. B, Anteroposterior radiograph showing the coracoclavicular joint.

(From Cheung TF, Boerboom AL, Wolf RF, Diercks RL. Asymptomatic coracoclavicular joint. J Bone Joint Surg Br . 2006;88[11]:1519-1520.)

Symptomatic CC joints have been reported in the settings of osteoarthritis (OA) and rheumatoid arthritis (RA), trauma, and thoracic outlet syndrome, but the overall clinical significance of the CC joint is uncertain ( Fig. 9-11 ). Troublesome peripheral ipsilateral limb pain, paresthesias, and weakness may evolve.

The coracoclavicular joint in a male patient with ankylosing spondylitis. Note the ipsilateral acromioclavicular joint.

Lehtinen et al. recognized resorption of the undersurface of the clavicle and believed it to be an atypical manifestation of RA involvement of the CC joint. Singh et al. conducted an extensive review of the features of the CC joint, leading to a proposed algorithm for treatment. Corticosteroid injection into the joint has been successfully used to treat a painful CC articulation. Surgical excision of the joint to resolve pain has been reported.

Complete Excision

Surgical removal of the lateral clavicle is the definitive treatment for a multitude of disorders of the ACJ. The method of excision initially described was the contemporarily termed “open” technique. Arthroscopic excision of the distal clavicle was first described by Johnson in 1986. Snyder and Esch et al. separately reported arthroscopic excision of the distal clavicle in 1988.

Excision of the lateral portion of the clavicle is indicated when the cause of shoulder pain can be attributed solely to the ACJ or a component of a shoulder pain–generating syndrome. The diagnosis is confirmed by the response to intra-articular injection of local anesthetic. Suspicion of additional or alternate diagnoses requires thorough diagnostic investigation, occasionally including arthroscopy of the shoulder performed in conjunction with arthroscopic distal clavicle resection (DCR).

It seems that either Facassini or Morestin first performed excision of the distal clavicle about a century ago. In 1941, Gurd and Mumford independently gave the first published accounts of the procedure. Subsequently, it has been the subject of many investigations seeking to define the essential yet safe extent of excision. Nearly all recommendations that have been proposed, whether the technique is performed in the traditional open manner or with an arthroscope, take into account the protection of the investing ligamentous support for the ACJ and the CC ligaments. By doing so, postresection instability that might contribute to discomfort and dysfunction is potentially minimized.

The recommended resection length varies from 0.5 to 2.5 cm. Although the precise amount to be excised in each instance is unknown, it would not be erroneous to suggest removing approximately 1 cm of the lateral clavicle.

However, this recommendation lacks a high level of evidence support from the literature. Branch et al. suggested that 5 mm is adequate to prevent bony abutment in both rotationally and axially loaded shoulders if the CC and AC ligaments are intact. This recommendation was confirmed in a study by Edwards et al. They further showed that if the resection was 2.5 mm or less, abutment was likely to occur between the acromion and the clavicle, particularly inferiorly and posteriorly, and that joint stiffness diminished with the amount of bone resected. Boehm et al. advised a resection of no greater than 0.8 cm in women and 1.0 cm in men based on cadaver studies using magnification and microcalipers. Exceeding the recommended amount consistently resulted in detachment of some portion of the trapezoid ligament. The anatomic study by Renfree et al. measured the distance from the center of the articular surface of the distal clavicle to the resected edge of the clavicle. They concluded that resection of less than 11.0 mm should never violate any portion of the trapezoid ligament in 98% of men or women, and resection of less than 24.0 mm should never violate any of the conoid ligament. Resection of more than 7.6 mm of the distal clavicle in men and 5.2 mm in women, performed by an arthroscopic approach, can violate the superior AC ligament. Eskola showed that patients with a resection of more than 1.0 cm have more pain than patients with less than 1 cm of resection. The findings of their cadaveric dissections led Stine and Vangsness to advise that to preserve the integrity of the investing AC ligaments and CAL, ACJ resections should be 5 to 7 mm by removing 2 to 3 mm from the medial acromion and 3 to 4 mm from the lateral clavicle. This recommendation is supported by the results of an in vitro biomechanical investigation by Pandhi et al., whose investigation compared peak load to failure of specimens undergoing 1 cm DCR to specimens undergoing mm resection of the medial acromion and distal clavicle, and showed significantly higher resistance to failure with symmetric mm resection of bone from the ACJ.

In a biomechanical cadaver study, Matthews et al. determined no significant differences in residual ACJ stiffness and compressive contact between the acromion and the lateral clavicle when comparing mock 5-mm arthroscopic resection to 10-mm open resection. Beitzel et al., when examining horizontal translations of the clavicle in cadavers undergoing sequential excision of the distal clavicle, documented progressively increasing AP excursions compared with the native joint with inferior capsule excision and 10-mm bone excisions, even when the superior and posterior AC capsular ligaments and the CC ligaments remained intact. Further excursion was observed with complete AC capsule excision, suggesting the significant importance of protection or reconstruction of the joint capsule.

In a cadaver study by Harris et al., 15.3 mm was the mean distance between the end of the clavicle and the most lateral fibers of the trapezoid ligament. Rockwood et al. observed cases with persistent pain after lateral clavicle excision that they attributed to inadequate resection and confirmed radiographically. He recommended excision of at least 2 to 2.5 cm of the clavicle, acknowledging that resection to this extent sacrifices the majority of the trapezoid ligament but spares the conoid ligament so that the lateral clavicle remains stabilized. With open techniques for lateral clavicle excision, meticulous closure of the superior capsule and investing AC ligaments is extremely important.

Under experimental conditions, it was predicted that when the distal clavicle is resected, higher loads may be borne by the adjacent soft tissues. In the same model, the intact surrounding soft tissues exerted a dampening effect, with a reduction in the compressive loads across the ACJ. In another study using cadavers, posterior translation of the clavicle was significantly reduced when the CAL was incorporated as an augment to the capsule repair performed at the time of open excision of the distal 1.0 cm of the distal clavicle.

In an attempt to further spare the investing stabilizing ligaments, the plane of resection should mimic the plane of the angular orientation of the existing joint when possible. To prevent the abutment of the acromion and clavicle during adduction of the arm, Neer advocated the removal of 1.5 to 2.0 cm, resecting more superiorly and posteriorly than inferiorly and anteriorly. Intraoperative ultrasonography has been used for verification of appropriate arthroscopic DCR. On the basis of their investigation and to prevent further supraspinatus outlet–encroaching osteophytes, Roidis et al. recommended DCR at the time of surgery for impingement with or without rotator cuff repair (RCR), even when no symptoms could be attributed to the ACJ.

Partial Excision

Concern exists for the fate of acromial stability at the time of open or arthroscopic subacromial decompression operations, especially when performed with coplaning of the ACJ and distal clavicle. The inferior AC ligament can extend as far as 10 mm lateral to the joint line. In addition, the fibers of the medial-most CAL can be found intimate to the inferior AC ligaments and capsule. It has been shown experimentally that it is impossible to perform standard subacromial decompression without violating the integrity of the inferior AC ligaments on the acromion process. Although the immediate effect in the experimental cadaver model is AP and superior compliance increased as much as 32%, in vivo factors such as ligament healing and muscle recovery can dampen the effect of acromioplasty and coplaning. In a cadaveric study by Edwards, coplaning the lateral clavicle together with acromioplasty contributed to more joint laxity and translations of the lateral clavicle than did acromioplasty alone.

According to studies with 4- to 7-year follow-up data, coplaning of the ACJ at the time of arthroscopic subacromial decompression did not contribute to instability, pain, or further surgery at the ACJ. Contrary to these reports, Fischer et al. reported increased OA and instances of symptomatic ACJ instability, some necessitating reoperation, and expressed their concerns with coplaning at the time of arthroscopic subacromial decompression. This experience led to their recommendation for either resection or preservation of the entire lateral clavicle. Other authors have observed that arthroscopic subacromial decompression accompanied by coplaning or partial excision of the lateral clavicle can result in AP laxity and instability symptoms arising from the modified ACJ. Kuster et al. reported that arthroscopic subacromial decompression did not accelerate the development of osteoarthritic changes in the ACJ. In the report by Kharrazi et al. in 2007, nearly 50% of patients who underwent reoperation for ACJ symptoms following arthroscopic subacromial decompression and ACJ surgery had persistent symptoms. Contrary to their earlier report, they identified no reason to perform concomitant coplaning or complete excision of the distal clavicle. The evidence from the literature is conflicting, but suggests that coplaning can create or intensify symptoms arising from the ACJ.

Results of Excision

Rabalais and McCarty performed a systematic review confirming that the best level of evidence for the excision of the lateral clavicle is level III or level IV. From their careful analysis and identification of low levels of evidence, they concluded that the method with the best results or fewest complications could not be determined. That stated, there is little argument that excision of the lateral clavicle successfully addresses the majority of the pathologic entities encountered at the ACJ. However, there are unsatisfactory outcomes because of cosmesis, pain, and dysfunction that have at their root mechanical more so than nonmechanical factors. The factors are often interrelated, and include improper resection, instability, loss of soft tissue integrity, heterotopic ossification, scars, and neuromas. The vast majority of clinical scenarios for which excision of the lateral clavicle is the only reasonable treatment to consider offer absolutely no technical challenges. Therefore it is the execution of the technique that results in unsatisfactory outcomes.

Open

In the series by Chronopoulos et al., a complication rate of 64% was reported. They found not only the usual complications but also several not reported in the literature such as ACJ tenderness (55%), incisional scar sensitivity (55%), hypertrophic incisional scar (14%), and infection (10%). The incidence of postoperative stiffness is reportedly as high as 29%.

Arthroscopic

The incidence of reoperation for inadequate excision ranges from 0% to 6.2%. Fracture of the distal clavicle attributable to misidentification of the ACJ was reported by Ghodadra et al. Iatrogenic excision of the CC ligaments can result in a type III AC dislocation. A secondary operation to treat heterotopic ossification has been reported.

Author’s Preferred Method

The patient is placed in the beach chair position with the head supported on an adjustable well-padded headrest to allow free access to the superior aspect of the shoulder. The skin incision is made in the relaxed skin tension lines approximately 1 cm medial to the ACJ, extending for 3 to 4 cm ( Fig. 9-14 ). A small, straight Gelpi retractor is placed into the subcutaneous tissue, which is gently elevated from the underlying fascia.

A superior incision in the relaxed skin tension lines is made slightly medial to the acromioclavicular joint for a distance of 3 to 4 cm.

(Courtesy Steven Lippitt, MD.)

Beginning 2.0 to 2.5 cm medial to the ACJ, electrocautery is used to incise the soft tissue overlying the superior aspect of the clavicle ( Fig. 9-15 ). The incision should be carefully placed to avoid the trapezius and deltoid muscles from their respective insertions. The incision is continued laterally through the capsule of the ACJ onto the acromion process for a distance of approximately 0.5 to 1 cm. Subperiosteal dissection of the distal clavicle is carried out with electrocautery to create anterior and posterior soft tissue flaps that include the capsule, the superior ligaments of the ACJ, and the adjacent muscles. Complete soft tissue dissection is confirmed with a 0.25-inch key elevator enabling the placement of two small Chandler retractors bent to 90 degrees anterior and posterior to the distal clavicle. A length of 1.25 cm of the distal clavicle is measured, marked, and resected using a microsagittal saw ( Fig. 9-16 ). The resected fragment is grasped with a towel clip or pointed bone reduction forceps and manipulated to allow access for the release of the remaining soft tissue attachments inferiorly.

In line with the clavicle, the deltotrapezius fascia and underlying acromioclavicular joint capsule are divided and reflected from the distal clavicle.

(Courtesy Steven Lippitt, MD.)

The surrounding soft tissues are protected as the lateral clavicle is excised with a microsagittal saw.

(Courtesy Steven Lippitt, MD.)

Upon removal of the completely resected lateral clavicle, the wound is thoroughly irrigated with normal saline by the bulb syringe technique. The resected clavicular surface is inspected, palpated, and smoothed, if needed, using a pinecone bur, rasp, or file. Proliferative synovium and abnormal intra-articular disk material is excised. By placing a fingertip into the area of resection and performing extremity adduction, contact between the acromion and the residual distal clavicle is recognized and corrected by additional resection.

The surrounding soft tissues are infiltrated with a local anesthetic solution consisting of equal parts of 0.5% Marcaine with epinephrine and 1% lidocaine. With the medial subcutaneous flap retracted with a bent Army Navy retractor, a 6-D Mayo trocar needle delivers No. 2 nonabsorbable sutures through the anterior and posterior soft tissue flaps ( Fig. 9-17 ). No more than three sutures placed with a figure-of-eight technique are usually necessary. The sutures are tied with four rows to limit the likelihood of palpable and potentially painful knots beneath the skin at the surgical site. Leaving a small tag further reduces the possibility of palpable sutures. For extremely thin patients, absorbable sutures are preferred. The subcutaneous tissue is closed with 2-0 absorbable sutures using a buried technique, capturing the dermal layer to ensure excellent superficial soft tissue apposition. Strips of 0.5-inch adhesive-backed paper tape cut in half are applied to the skin. A sterile dressing covers the wound, and the arm is placed in a sling.

The residual defect is closed by careful repair of the deltotrapezius fascia and acromioclavicular joint capsule. No more than three sutures placed with a figure-of-eight technique are usually necessary. The sutures are tied with four throws to limit the likelihood of palpable and potentially painful knots beneath the skin at the surgical site. Leaving a small tag further reduces the possibility of palpable suture.

(Courtesy Steven Lippitt, MD.)

Patients are discouraged from active arm elevation without the assistance of the opposite limb for 3 to 4 weeks to protect the deep soft tissue repair. Heavy use of the extremity is not recommended for 6 to 8 weeks to ensure that sufficient preliminary healing of the AC ligaments has taken place and that the residual void created at the time of DCR has been filled with scar. Hydrotherapy to obtain range of motion is suitable as soon as the wound is dry. Return of full strength, range of motion, and full function can be expected in most cases by 12 weeks.

Technical Pearls and Pitfalls

Cosmetically unappealing scars may be avoided by carefully placing the incision in the relaxed skin tension lines.

Placing the skin incision approximately 1 cm medial to the palpable prominence of the distal clavicle enables optimum and safe placement of the saw. The angle of resection favors removing slightly more bone superiorly and posteriorly than inferiorly and anteriorly ( Fig. 9-18 ). This orientation of the plane of resection prevents contact between the remaining clavicle and the acromion when the arm is placed in cross-body adduction. A saw is preferred to the use of either a bur, which creates more bone debris, or an osteotome, which can shatter the bone.

The plane of resection favors slightly more bone removal superiorly and posteriorly to prevent abutment of the acromion process on the clavicle during extremity adduction.

(Courtesy Steven Lippitt, MD.)

The soft tissue sleeve that includes the ACJ capsule and its supporting ligaments is carefully closed side to side without an attempt to obliterate the dead space formed by the absence of the lateral clavicle. It is important that the repair suture engage the entirety of the reflected soft tissue, not just its most superior portion. This effectively reconstructs the superior AC ligaments and capsule, which reduces the possibility of iatrogenic AC instability. The temptation to create an interposition arthroplasty using the soft tissues adjacent to the joint should be resisted. Their relative immobility risks a closure under tension that can lead to dehiscence of the tissues at the repair site. Overtightening the trapezius or deltoid muscle at the resection site can contribute to focal discomfort. Likewise, an insufficient closure can be painful, cosmetically unappealing, and functionally compromising.

Heterotopic ossification can develop at the surgical site. Preventive measures include subperiosteal dissection with electrocautery rather than sharp or blunt elevation of the soft tissue from the bone, thorough irrigation of the surgical site to remove residual bone debris, and prophylactic administration of pharmacologic agents such as indomethacin.

Summary

Disorders of the distal clavicle that result from internal derangement of the ACJ or irreversible deterioration of its articular surface can be effectively treated with excision of the distal clavicle through a limited open surgical approach. It is very important to carefully preserve the soft tissues contiguous with the joint capsule. Sufficient resection requires removal of no less than 1 cm and no more than 2.5 cm of the distal clavicle.

An anatomic approximation of the soft tissues at the time of closure will safeguard the horizontal stability of the ACJ. Cosmesis is maintained through the use of proper skin incision orientation and by making no attempt to interpose a large volume of the adjacent soft tissues. Technical ease and swiftness of performance predictably resulting in favorable and durable outcomes are the hallmarks of this operation.

Characteristics of the Painful Acromioclavicular Joint Evaluation

This section focuses on the assessment of the atraumatic causes of ACJ pain and dysfunction.

History

Pain is the most common symptom of an ACJ disorder. Without prompting, nearly all patients point directly at the ACJ or its vicinity when they begin to describe their symptoms. However, the expression of ACJ pain may be altogether different as reported in an interesting study by Gerber et al. Hypertonic saline injected into the ACJ of otherwise asymptomatic volunteers resulted in variable pain patterns that included not only the ACJ but also the area of the anterolateral deltoid, the suprascapular region including the trapezius and the supraspinatus, and the anterolateral aspect of the neck. The results of a pain mapping study enabled Bayam et al. to report that the pain resulting from ACJ pathology was predominantly sharp and stabbing, localized to the anterior shoulder, and seldom radiated to the forearm ( Fig. 9-19 ). It was the most focal and the least severe compared with pain arising from other shoulder disorders.

Mapping for acromioclavicular joint pain.

(Modified from Bayam L, Ahmad MA, Naqui SZ, et al. Pain mapping for common shoulder disorders. Am J Orthop . 2011;40[7]:353-358. Copyright 2011 The American Journal of Orthopedics.)

The positive value of pain diagrams for detecting and diagnosing AC disorders was shown by Walton et al. For the patient exhibiting shoulder pain between the mid-clavicle and the deltoid insertion with a positive Paxinos test (described under “Physical Examination”) and a positive bone scan for uptake at the ACJ, the researchers were convinced that the diagnosis of ACJ pain was “virtually certain.” Patients with a problematic ACJ do not commonly mention swelling or crepitus arising in or about the joint.

Physical Examination

With the patient either sitting or standing and the overlying clothing removed from both shoulders, the examiner looks for asymmetry: shoulder posture at rest, skin color, muscle bulk, and surface topography ( Fig. 9-20 ). Both shoulders are examined for active motion, voluntary provocative maneuvers, skin temperature, tenderness, ligamentous laxity, passive motion, sensation, strength, circulation, and discriminatory maneuvers. The depth of examination of the cervical spine, remaining components of the shoulder girdle, and the upper limb is variable and determined individually.

Patient preparation for physical examination requires unimpeded access to both shoulders.

The patient may be asked to identify as precisely as possible, the location of the pain. A positive “one finger test” is recognized when the patient points directly to the ACJ.

Asymmetric tenderness well localized to the ACJ indicts the ACJ as the source of the symptoms, especially if the pain so produced is essentially the same in character as the pain experienced by the patient. It is sometimes very difficult to confidently identify and successfully palpate the ACJ when the patient is obese or very muscular. In these instances, it might prove helpful to reference the triangle formed by the clavicle, scapular spine, and base of the neck. The ACJ is directly anterior to an examining fingertip placed just medial to the acromion process at the lateral apex of the triangle ( Fig. 9-21 ).

Surface landmarks helping to identify the acromioclavicular joint, which lies directly anterior to the soft spot at the apex of the isosceles triangle formed by the scapular spine, the clavicle, and the base of the neck.

Provocative Testing

Several pain-exacerbating examination maneuvers that suggest ACJ pathology have been described, including the cross-body adduction stress test, the AC resisted extension test, Buchberger test, the active compression test, and the Paxinos test. However, in a systemic review with meta-analysis, Hegedus et al. concluded that the diagnostic accuracy of these orthopedic special tests could not be substantiated. With the cross-body adduction stress test, the examiner passively elevates the upper limb to 90 degrees in the sagittal plane and, with the elbow either extended or flexed slightly, moves the limb medially ( Fig. 9-22 ). The test is considered positive if it reproduces pain in or near the ACJ.

The cross-body adduction stress test.

The passive external rotation test uniquely stresses the ACJ compared with more traditional tests such as cross-body and active compression tests. With the arm at the side maintaining the thoracohumeral relationship with the flexed elbow, the arm is passively externally rotated to achieve the maximum range possible. The anterior deltoid muscle is passively stretched, resulting in pressure on the ACJ as the distance between the clavicle and acromion are reduced.

The AC resisted extension test starts with the upper limb elevated 90 degrees in the sagittal plane, the elbow flexed to 90 degrees, and internal rotation to 90 degrees ( Fig. 9-23 ). With the examiner’s hand fixed in the space against the posterior elbow, the patient extends the shoulder in the transverse plane, meeting the examiner’s resistance. The test is considered positive if pain is experienced in the ACJ.

The acromioclavicular resisted extension test.

The Buchberger test combines inferiorly directed force to the lateral clavicle with passive forward elevation of the slightly adducted and externally rotated upper limb. The test is considered positive if pain is invoked or intensified in or near the ACJ. To perform the Paxinos test, the examiner places a thumb over the posterolateral acromion and the ipsilateral or contralateral index or long finger (or both) over the superior aspect of the mid-part of the clavicle. AS pressure is applied by the thumb while inferior pressure is applied to the acromion. The test is considered positive if pain is produced or intensified in the region of the ACJ ( Fig. 9-24 ).

The Paxinos test. The examiner’s thumb and finger are squeezed together.

(From Walton J, Mahajan S, Paxinos A, et al. Diagnostic values of tests for acromioclavicular joint pain. J Bone Joint Surg Am . 2004;86[4]:807-812.)

The active compression test is begun by requesting the patient to fully extend the elbow and forward elevate the upper limb 90 degrees in the sagittal plane followed by 10 to 15 degrees medial to the sagittal plane ( Fig. 9-25 ). The thumb is pointed downward by full internal rotation of the shoulder and pronation of the forearm. An inferiorly directed force is applied to the upper limb thus positioned. The force is released, the forearm is fully supinated, and the force is reapplied. The test is considered positive if pain is produced with the first maneuver and reduced or eliminated with the second maneuver. Pain in or near the ACJ indicates a problem of the ACJ.

The active compression test. This two-step maneuver can help to distinguish between acromioclavicular joint and labral lesions. A, This maneuver may reproduce pain in the shoulder. Note a downward force being applied to the hyperpronated adducted limb, which has been raised to 90 degrees. B, This maneuver may reduce or eliminate shoulder pain. Note the limb in full supination. The test is considered positive if pain produced with maneuver one is reduced or eliminated with maneuver two. Both acromioclavicular joint and labral tears may be discerned by the location of the pain.

(Modified from O’Brien SJ, Pagnani MJ, Fealy S, et al. The active compression test: a new and effective test for diagnosing labral tears and acromioclavicular joint abnormality. Am J Sports Med . 1998;26:610-613.)

The Bell–Van Riet (BVR) test, described in 2011, is a variation of a cross-body adduction test. The upper limb is passively elevated to 90 degrees, maximally adducted, and fully internally rotated, with the elbow in full extension and the forearm pronated. The examiner facilitates maintenance of the extreme position—pain is considered a “positive cross-arm adduction sign.” A resisted downward force on the forearm that evokes pain and inability to maintain the starting position is considered a positive BVR test. The maneuver can be repeated with the limb in external rotation. Essentially the same test was described in a nearly identical publication in 2013 and designated the sac test. To better understand how the BVR test evoked pain, Alta et al. invited healthy volunteers 23 to 60 years of age with no history of shoulder problems to undergo 4D CT scanning of the shoulder with the arm adducted plus resisted elevation, essentially simulating the BVR test. They observed that the clavicle translated posteriorly and superiorly, with some opening of the superior joint space and maintenance of joint width. Reduced translation and increased age were positively correlated.

I learned the hug test from Douglas Harryman and Frederick Matsen as an additional clinical tool for assessing the ACJ ( Fig. 9-26 ). The examiner stands at a right angle to the patient’s symptomatic side, reaches around the patient’s shoulders, and locks both hands together. The hug is then performed, compressing both shoulders in the coronal plane between the examiner’s chest and hands. The test is considered positive if pain is produced or aggravated in or near the ACJ closest to the examiner.

The hug test elicits pain from the affected acromioclavicular joint.

Maritz and Oosthuizen showed high sensitivity for ACJ line tenderness and cross-body adduction stress testing. O’Brien et al. determined that the active compression test exhibited a high degree of accuracy (100% sensitive and 96.6% specific) for ACJ lesions.

In a study by Walton et al., ACJ tenderness to direct palpation had the greatest sensitivity for the detection of ACJ disorder in patients with shoulder pain, followed by the Paxinos test. They also reported that the most accurate clinical maneuver for detecting an AC abnormality was the Paxinos test. Chronopoulos et al. analyzed patients with isolated chronic disorders of the ACJ to determine the diagnostic value of physical examination provocative testing techniques. The most sensitive test, the cross-body adduction test, was also the least accurate. The active compression test had the greatest specificity, the least sensitivity, and the most accuracy. The AC resisted extension test was between the other two for all diagnostic values. Interestingly, provocative maneuvers described for other shoulder maladies (Neer impingement sign, Hawkins impingement sign, painful arc sign, drop arm sign, and Speed’s test, Jobe’s sign, neck tenderness) are capable of inducing pain in patients with chronic lesions of the ACJ.

Using dynamic ultrasonography of the ACJ in normal volunteers, Park et al. concluded that the cross-body adduction and active compression tests resulted in more stress to the ACJ than the passive external rotation test, highlighting their utility.

The response to local anesthesia instilled into the ACJ can help identify the cause of shoulder pain. This is especially true when provocative testing is performed both before and after the injection. Von Reit and Bell ascertained the sensitivity of provocative clinical tests with local anesthetic and corticosteroid injection testing of the ACJ. The most sensitive test was the BVR performed in internal rotation (98%), followed by the O’Brien test (83%), the cross-arm adduction test (67%), the Jacob’s test (41%), and the Paxinos test (12%). Curiously, the exact same results were reported by Saccomanni, who failed to credit Van Riet and Bell for their work. In addition, the author of this chapter has personally spoken with Shaffer, the author of the article cited by Von Riet and Bell and Saccomanni, and Dr. Shaffer stated “… I’m afraid that I did not find a reference for that test in my manuscript, nor am I aware of this particular test or have any experience with it clinically.” In a primary care setting that utilized local anesthetic injection testing followed by certain traditional ACJ tests, the limited diagnostic value of the tests was realized while combinations of aspects of history and physical examination appeared to provide more diagnostic value for identifying painful ACJ conditions.

Despite the very superficial location of the ACJ, its relative size, variable anatomy, and pathologic changes often confound arthrocentesis. Successful needle entry into the joint often lacks tactile confirmation ( Fig. 9-27 ). Bisbinas et al. reported incorrect placement more than 60% of the time when orientation to the joint was obtained by palpation alone. Accuracy of intra-articular placement confirmed by arthrography was 43.3% in a study by Wasserman et al. In cadaver studies, ACJ injection accuracy by palpation was 57% to 70%. The use of ultrasound guidance improved the accuracy to 90%. When the accuracy of injection in the clinical setting is deemed critical, some authors recommend using image intensification or fluoroscopic guidance to improve accuracy. Others advise the use of sonography, which has been shown to be 98% to 100% accurate compared with the 40% accuracy of injection guided by palpation. Edelson et al. showed the accuracy of ultrasound-guided injection with arthrography. Reiterating their sources, Daley et al. as well as Aly et al. cited that the injection accuracy rate is higher when performed with image-guided assistance versus the landmark-guided approach (palpation) (93.6% vs. 68.2%). Surgical decision-making with respect to the response to injection should take into account all the factors and current data relevant to injections in and around the ACJ.

Even with vast experience, acromioclavicular joint injections are easier said than done.

(Courtesy Steven Lippitt, MD.)

Imaging

Plain Radiographs

Due to its composition of less-dense bone and its location in a region of the shoulder with relatively less soft tissue coverage, standard shoulder techniques must be altered to properly delineate the ACJ. Depending on the patient’s stature or body habitus, a reduction of x-ray kilovoltage by as much as one-half is sometimes necessary in order to gain optimal visualization of the ACJ.

Anteroposterior View

It is preferable to obtain this view with the patient either standing or sitting with the arms unsupported and neutrally positioned in the sagittal plane. Anatomic variations and technical inconsistencies are effectively dealt with by imaging bilateral ACJs simultaneously on a single cassette or on two smaller cassettes for wider patients ( Fig. 9-28 ). A 10- to 15-degree cephalic tilt applied to the x-ray beam optimizes the image of the ACJ by eliminating its superimposition on the posterior acromion and lateral scapular spine ( Fig. 9-29 ).

Positioning and technique for Zanca views. Both shoulders (A) are imaged simultaneously on one or two cassettes (B) , depending on the width of the shoulders.

The Zanca view is taken with the x-ray beam directed 10 to 15 degrees cephalad.

Dimensions and configuration are extremely variable, mandating a cautious approach to interpretation in the setting of injury. AC width ranges from 0.5 mm to 7 mm, and the CC interval ranges from 1.1 cm to 1.3 cm.

Lateral Views

One of two views is selected: either an axillary lateral view or a scapular Y view. The axillary view offers an advantage for determining the anterior or, much more commonly, posterior displacement of the clavicle ( Fig. 9-30 ). The scapular Y view highlights superior displacement of the clavicle, although posterior displacement can be demonstrated ( Fig. 9-31 ).

The axillary lateral view enables assessment for anterior or posterior displacement of the clavicle with respect to the acromion.

Scapular Y view showing superior displacement of the clavicle resulting from an injury.

Other Views

The axial and tangential views of the ACJ have been described in normal volunteers. The views are not in widespread usage, and their role in injury assessment and treatment has not been investigated.

Stress Views

It is sometimes possible to subject the acutely injured shoulder to an additional deforming stress as an aid to diagnosis. A radiograph obtained simultaneously with the stress maneuver often captures the critical information sought by the examiner. Stress views can facilitate the diagnosis and characterization of the severity of an AC dislocation, although some question their value. In a survey of members of ASES, which had a 94% response rate, 81% indicated that they did not recommend the routine use of stress radiographs in the emergency department, and 57% did not use the views. Only 9% changed their treatment plan based on the outcome of the views. The radiographic assessment for nontraumatic conditions of the ACJ does not usually incorporate stress views.

Anteroposterior Stress Views

For standard AP stress views, the technical principles and patient positioning are unchanged from routine AP views of the ACJ. Stress of 10 to 15 pounds is applied to both upper limbs while the patient is invited to relax as much as possible. The method of weight attachment to the upper limb, hand-held versus wrist-suspended, might not be important. It may be possible to ascertain avulsion of the anterior deltoid by other forms of stress application, but the accuracy of a weight-lifting view and its role in the routine assessment of suspected AC dislocations has yet to be determined.

Lateral Stress Views

It is possible to demonstrate pathologic relationships between the acromion and the clavicle by using a shoulder-forward view (Alexander view), obtained like a scapular Y while the patient actively protracts the shoulder ( Figs. 9-32 and 9-33 ). Although it is anticipated that this stress will displace the acromion anterior and inferior to the clavicle, more comprehensive imaging may be necessary to properly define the severity of injury. A similar cross-body adduction view was reported and used by Barnes et al. to help identify a stable or unstable ACJ, and was identified by Beitzel et al. as the Basamania view ( Fig. 9-34 ). Tauber et al. described the advantages of dynamic axillary views in instances of suspected ACJ injury for determining the presence of horizontal instability. With this view, they introduced the concept of gleno-acromio-clavicular angle as a means to quantify the translations of the distal clavicle observed while the shoulder assumed three specific positions for the axillary radiograph ( Fig. 9-35 ).

Lateral stress or Alexander views to assess for injury of the acromioclavicular joint. Left, The shoulders are thrust forward while the radiograph is taken. Center, Alexander view in the relaxed position. The acromioclavicular joint is only minimally displaced. Right, With the shoulders thrust forward, the acromion is displaced anteriorly and inferiorly under the distal end of the clavicle.

Clinical appearance of the lateral stress view.

Basamania view. A, Anteroposterior view showing widening of the acromioclavicular joint and the coracoclavicular interval. B, The patient pulls the elbow of the affected arm to their midline using the unaffected arm. C, Clinical appearance. D, The acromion process of the scapula has rotationally translated internally, medially, and inferiorly beneath the distal clavicle, which is indicative of instability.

(Courtesy Carl Basamania, MD, The Polyclinic, Seattle, WA.)

Functional axillary view used to reveal dynamic horizontal instability. A, Patient positioning. B, Position 0. C, Position 1, detecting posterior dislocation. D, Position 2, detecting anterior dislocation. E, Surface computed tomography of position 0. F, Surface computed tomography of posterior dislocation. GACA, Glenoacromioclavicular angle.

(From Tauber M, Koller H, Hitzl W, et al. Dynamic radiologic evaluation of horizontal instability in acute acromioclavicular joint dislocations. Am J Sports Med . 2010;38[6]:1188-1195.)

Computed Tomography

The primary purpose of CT is to obtain a more accurate assessment of the morphology of the distal clavicle, coracoid process, and acromion process and, to a lesser extent, their relationships with one another. Accounting for differences in patient positioning, reduced gravitational forces may come to bear upon CC and AC relationships. It can aid in the analysis of complex injuries, fractures, neoplastic processes, or infections, especially when postprocessing 3D images are generated (see Fig. 9-10A ). Wide view 4D CT scanning (dynamic) has been described, and provides an opportunity to document multiplanar components of ACJ injuries and better understand their significance ( Fig. 9-36 ).

A to C, Examples of the type of advanced imaging that may enhance acromioclavicular joint classification schemes.

(From Dyer DR, Troupis JM, Kamali MA. Wide field of view CT and acromioclavicular joint instability: a technical innovation. J Med Imaging Radiat Oncol . 2015;59[3]:326-330.)

Nuclear Medicine

When there are no other clues to a diagnosis in the presence of vague shoulder pain, bone scintigraphy is helpful to rule out an occult osseous lesion of the shoulder girdle. It is an excellent imaging tool for accessing semiacute injuries to the ACJ and ACJ arthropathy. The scan is positive when it demonstrates focal increased radioactive tracer uptake when compared with the uninjured contralateral side, especially if the plain radiographs demonstrate bilateral symmetrical AC arthrosis ( Fig. 9-37 ). It confirms the ongoing osseous activity in cases of osteolysis. Walton et al. found that bone scanning had a higher degree of accuracy for determining an AC abnormality than did MRI or plain radiographs. In this study, it had a higher predictive value (positive and negative) than did clinical testing or local anesthetic injections. While usually not necessary, single-photon emission computed tomography (SPECT) scan imaging has the capacity to render the active processes at the ACJ in a dramatic manner ( Fig. 9-38 ). Tc-99m leukocyte imaging has proved useful in instances of septic arthritis of the ACJ.

The 2-hour delayed image technetium-99m bone scan revealing intense uptake in the right acromioclavicular joint.

Single-photon emission computerized tomography images provide remarkable identification of an active process at the acromioclavicular joint on the right. A, Symptomatic joint. B, Asymptomatic joint. C, Axial cut. D, Coronal cut.

Magnetic Resonance Imaging

MRI effectively displays the pathologic changes that result from injuries and nontraumatic disorders that involve the ACJ. New MRI scan orientations for the ACJ and specifically to appropriately image the CC ligaments have been described and their utility confirmed in a report by Schaefer et al. Complete correlation of clinical, plain imaging, MRI and pathologic findings may prove difficult. While currently not particularly practical, stress MRI has been shown to distinguish partial from complete injuries as well as functional integrity of the ligament. However, plain films usually suffice and do so at significantly less expense.

MRI is the best imaging modality for assessing masses or cysts in the region of the ACJ. In that setting, it can delineate the internal characteristics, margins, and surrounding tissue response to the mass. Concomitant massive RC tears are identified in many instances.

The value of the routine use of MRI for the diagnosis and management of arthropathy of the ACJ has been questioned. One reason is because it is exceedingly rare that an MRI report for a patient with shoulder pain fails to mention an abnormality of the ACJ due to its high sensitivity for detecting AC abnormalities. Another is that degenerative findings of the ACJ are just as likely to be present in asymptomatic persons. The importance of the clinical evaluation in the presence of this particular imaging abnormality cannot be overemphasized.

In the study by Walton et al., MRI had a higher predictive value (positive and negative) than did clinical tests or local anesthetic injections. It has been shown to be more sensitive for OA than conventional radiography. Fiorella et al. reported a 12.5% incidence of increased T2 signal in a series of shoulder MRI examinations, and stated that it usually was of no clinical significance except in patients with chronic AC pain and no other imaging abnormalities. Increased T2 signal intensity representing edema is also a common finding in traumatic and arthropathic conditions of the ACJ ( Fig. 9-39 ). MRIs of patients with symptomatic ACJs are more likely to show edema in the distal clavicle and higher grades of OA than in asymptomatic patients. Reactive bone edema in the distal clavicle, medial acromion, or both is a more reliable predictor of AC pathology than are degenerative changes. Fluid in the ACJ is a common MRI finding in patients with shoulder problems ( Fig. 9-40 ). Advancing age, ACJ osteophytes, and GHJ fluid are notable associated findings that suggest a relationship between ACJ fluid and OA.

Magnetic resonance image showing increased T2 signal in the lateral clavicle.

Magnetic resonance image showing fluid in the acromioclavicular joint.

With special techniques, the intra-articular disk may be distinguishable, when present, from the adjacent articular cartilage.

Ultrasonography

Technologic advancements together with the expanding capabilities of ultrasonography have further stimulated an interest in applying this form of imaging to the ACJ. Its reliability for the injured ACJ is well documented. Heers and Hedtmann have demonstrated the advantages of ultrasonography for determining the extent of injury to the deltoid and trapezius muscles and their common fascia. They expressed concern that conventional sagittal plane radiography might have the limitations of either overestimating or underestimating the extent of soft tissue injury. A simple dynamic provocative maneuver performed with ultrasonography can aid in detecting lesser injuries to the ACJ, such as type I injury.

Blankstein et al. favored ultrasonography for assessing patients presenting with atraumatic anterior shoulder pain, citing less pain, negligible ionizing radiation, and the visualization of degenerative changes that can go undetected by conventional radiography.

Ultrasonography has proven utility for diagnosing septic arthritis of the ACJ.

Overview

As many as 40% of shoulder injuries involve the ACJ. Most injuries to the ACJ and its environment are due to minor episodes of trauma that result in nothing more than contusions, minor strains, and sprains, never coming under the scrutiny of a healthcare provider, and from which full recovery is the norm. However, it is not infrequent that the events that result in injury to the ACJ are of sufficient magnitude to damage adjacent soft tissue structures, especially if the conditions of the sustaining force are considered high energy. Under these circumstances, the AC and CC ligaments and the attachments of the deltoid and trapezius muscles are susceptible to partial or complete tearing or avulsion.

ACJ dislocations represent 12% to 15% of all dislocations of the shoulder girdle and 8% of all joint dislocations in the body. Emergency department visits for AC dislocation was 1.8 per 10,000 inhabitants in an Italian epidemiologic investigation, and 8.9 per 10,000 in a Scottish review. In a Swedish urban population of nearly one-quarter million, taking into account both sexes and all ages, AC dislocations represented 4% of all shoulder injuries. Dislocations were eight times more common in male subjects, and the prevalence was highest in the age group younger than 35 to 39 years, a finding also observed in previous investigations. Shoulder injuries often happen during alpine skiing, and around 20% affect the ACJ. In a prospective study of junior hockey players, ACJ injuries were second only to facial lacerations in terms of frequency of injury. A recent study of shoulder injuries in competitive rugby players documented a 32% incidence of ACJ involvement. Injury to the ACJ is the sixth most common among injuries sustained by collegiate-level football players (2.8%). More than 96% are incomplete injuries (see below). The injury is 12 to 14 times more likely to occur during a game than during practice. The incidence of ACJ injuries among all shoulder injuries sustained by players in the national football league during a 12-year period was 29.2%. That review also found that quarterbacks sustained the highest rate of injury, with 21 injuries per 100 players.

Associated Injuries

Potentially high forces risk damage to other structures in proximity to the ACJ. Fractures of the ribs, scapula, humerus, and clavicle have occurred. Rarely, a simple avulsion from the clavicle may occur.

Scapulothoracic dissociations have been reported, and add complexity to the management of ACJ injuries. Isolated dislocation or sprain of the ACJ is infrequently associated with severe head, great vessel, or thoracic injury. Pulmonary injuries have been reported.

Neurologic injuries that can include the brachial plexus are rarely associated with AC injuries.

Accompanying intra-articular GHJ lesions identified arthroscopically in both acute and chronic dislocations have included partial-thickness and complete-thickness tears of the RC, biceps tendon, and rotator interval; osteochondral humeral head defects; labral tears, especially type 2 SLAP lesions; and GHJ dislocation, sometimes treated concomitantly with the ACJ dislocation. Individuals over the age of 45 years are at threefold greater risk of having such a lesion.

Classification

The extent and severity of injury to the interconnecting and surrounding soft tissues of the scapula and the clavicle (the ACJ capsule with its intimate investing AC ligaments and CC ligaments and the deltoid and trapezius muscles) define the current and most widely accepted classification of ACJ injuries. Zaricznyj suggests that the term AC dislocation is an incomplete description of the most severe form of injury to the uniting structures of the two bones and that a complete separation of the two bones should be termed scapuloclavicular dislocation.

Injuries to the AC and CC ligaments may be distinguished by the preservation or loss of the gross structural integrity of the ligaments. Incomplete injuries result in no or only slight elongation of the ligament fibers, and complete injuries tear the ligament fibers into at least two separate pieces. By virtue of strength alone, the AC ligaments are more vulnerable than the CC ligaments. The lowest energy injuries result in isolated damage to the AC ligaments, and higher energy injuries affect both the AC and CC ligaments. The most severe injuries damage the deltoid and trapezius muscles or attachments. This phenomenon of sequential injury, first mentioned by Cadenat, became the fundamental principle upon which the earliest classification schemes were based, and continues to captivate contemporary investigators with respect to pathoanatomy and mechanism of injury.

Incomplete injury to the AC ligaments and no injury to the CC ligaments was termed type I. Complete injury to the AC ligaments with incomplete injury to the CC ligaments was termed type II. Complete injury to both the AC and CC ligaments was termed type III. Bannister et al. and Rockwood astutely recognized that certain injuries with displacement were sufficiently severe to warrant strong consideration for surgery as the preferred method of treatment, and proposed additional classification: type III subcategories A, B, and C by Bannister and separate types IV, V, and VI by Rockwood. Rockwood’s proposal gained widespread acceptance, and together with the original classification, it encompassed injuries to the ACJ in all recognized forms ( Fig. 9-41 ). Type IV injuries are the result of a summation of force vectors that result in the clavicle eventually resting posterior to the acromion, penetrating the trapezius muscle, sometimes completely through its substance. The clavicle in type V injuries is displaced superiorly but because of disruption of the attachments of the deltoid and trapezius muscles, significantly more than in type III. Type VI injuries feature a clavicle displaced inferior to the acromion or the coracoid process. An investigation of observer reliability for the diagnosis of ACJ injuries (limited to types III to V) found that consistency of grading was more common individually than it was among the individuals.

Classification of ligamentous injuries to the acromioclavicular joint. Type I, The acromioclavicular and coracoclavicular ligaments are intact. Type II, The acromioclavicular ligaments are completely torn, but the coracoclavicular ligaments remain intact. Type III, The acromioclavicular and coracoclavicular ligaments are disrupted. Type IV, The acromioclavicular and coracoclavicular ligaments are disrupted, and the distal end of the clavicle is displaced posteriorly into or through the trapezius muscle. Type V, The acromioclavicular and coracoclavicular ligaments are completely torn, and due to the severity of injury, the deltoid and trapezius muscles are detached from the distal clavicle. Type VI, The acromioclavicular and coracoclavicular ligaments are completely torn, and there is inferior displacement of the clavicle beneath the coracoid process.

Rockwood’s classification represents an amalgamation of historical data derived from the analysis of plain and stress radiographs, anatomic dissections, cadaver experiments, and surgical observations infused with extensive, in-depth personal experience. On this basis, and rightly so, the pathoanatomy is often inferred, only to be resolved by direct visualization at the time of operative treatment or by means of more sophisticated imaging such as MRI and ultrasonography. The latter concept was applied to four acutely injured patients, two uninjured patients, and one fresh frozen cadaver. In two of the four injured patients, the type of injury based on classic plain radiographic parameters and Rockwood’s classification did not correlate with the pathoanatomy detected by MRI. However, confirmation of findings was only possible in one of the four patients who elected for operative treatment.

Differences in ligament attachments and orientations support the notion of individual sprain of the conoid and trapezoid ligaments with ACJ injury. Perhaps this is what Neer astutely recognized decades earlier as a subgroup of injuries resulting in less displacement and healing with stability that he identified as “type II .” Selective sequential sectioning of the individual CC ligaments in the presence of complete AC resection (type II “plus”) documented pathologic translations of the clavicle rendering the joint experimentally and radiographically unstable. Beitzel et al. suggested distinguishing two variations of the type III Rockwood injury: type IIIa, a stable injury without a demonstrable overriding clavicle on the cross-body adduction stress view, and type IIIb, an unstable injury with resistant scapular dyskinesia and an overriding clavicle on the cross-body adduction stress view. The reliability of the Rockwood classification of ACJ dislocations as a tool for diagnosis and treatment has been questioned. Cho et al. determined that 3D CT did not add further to the reliability of classification and treatment of these injuries. Kraeutler et al. documented the inconsistency, even among shoulder surgeons, in the use of the Rockwood classification and its application for treatment decision making. In cadavers, Eschler et al. investigated the plain radiographic results of sequential sectioning of the AC ligaments and joint capsule and the CC ligaments. Their results confirmed the inferred pathoanatomic changes designated in the Rockwood classification, and also drew attention to the fact that a type V injury could be produced in cadavers without disturbance of the deltoid or trapezius attachments. The investigations and experiences of other authors have suggested that the instability patterns may be more complex, including medial instability, and may not always be well explained by classifications based on static rather than functional images.

Nemec et al. prospectively imaged 44 patients suspected of having an acute ACJ injury with plain imaging and MRI. Their findings resulted in a classification that adapted the Rockwood classification to the MRI findings ( Table 9-1 ). The classification on plain imaging and MRI were concordant in 52.2%, and the results of the MRI led to reclassification to a less severe injury in 34.2% and to a more severe injury in 11.4%. The MRI identified additional ligamentous lesions in 25% of the patients.

Several authors have suggested subclassification of type VI into type VIa (supracoracoid or subacromial), a lower energy injury reported to be always accompanied by a diaphyseal clavicle fracture, and VIb (subcoracoid), a higher energy injury.

Elamin et al. proposed a type VII injury based on their experience with a single instance of a locked superior AC dislocation. Intraoperatively, the joint was irreducible and complete tearing of the AC and CC ligament with periosteal stripping of the lateral clavicle was observed. Resisted engagement with the acromion was reconciled by DCR, while CC and AC reconstruction was performed with synthetic ligament and direct repair, respectively.

With the evolution of sophisticated imaging, perhaps new classifications of injuries will emerge that take into account the relevance of multiplanar aspects of injury (see Fig. 9-36 ). In a similar manner, it is intriguing to contemplate the impact of similar clinical investigations with expanded databases upon the classification of these injuries. Until then, based on current information gathered from relevant clinical and imaging reports, the author, as suggested by others, proposes a modification of the widely and highly regarded classic Rockwood classification of ligamentous injuries of the ACJ obtained from plain radiographs for traumatic disorders of the ACJ, that encompasses additional patterns of injury recognized and documented by others ( Fig. 9-42 ). Perhaps wider use of sophisticated imaging that may validate the results of in vitro investigations will lead to further refinement of this classification.

Author’s proposed contemporary modified Rockwood classification of acromioclavicular joint injuries. Type I, The acromioclavicular and coracoclavicular ligaments are intact. Type II, The acromioclavicular ligaments are completely torn, but the coracoclavicular ligaments remain intact. Type IIIa, The acromioclavicular and coracoclavicular ligaments are torn. The extent of tearing of the conoid ligament enables stability with adduction stress testing. Type IIIb , The acromioclavicular and coracoclavicular ligaments are torn. The extent of tearing of the conoid ligament prevents stability with adduction stress testing. Type IV, The acromioclavicular and coracoclavicular ligaments are disrupted, and the distal end of the clavicle is displaced posteriorly into or through the trapezius muscle. Type V, The acromioclavicular and coracoclavicular ligaments are completely torn, and due to the severity of injury, the deltoid and trapezius muscles are detached from the distal clavicle. Type VIa, The acromioclavicular and coracoclavicular ligaments are completely torn, and there is inferior displacement of the clavicle beneath the acromion process but above the coracoid process accompanied by a diaphyseal clavicle fracture. Type VIb, The acromioclavicular and coracoclavicular ligaments are completely torn, and there is inferior displacement of the clavicle beneath the coracoid process. Type VII, The acromioclavicular and coracoclavicular ligaments are completely torn, and there is superior displacement of the clavicle above the acromion process.

(Courtesy Steven Lippitt, MD.)

Classification of injury in the immature skeleton is similar to that in adults, and has been offered by Dameron and Rockwood. Type I injuries are the result of a mild sprain without compromise of the investing periosteal tube. Type II injuries result from partial disruption of the superior periosteal tube with slight widening of the lateral clavicular physis. Type III injuries result from a complete disruption of the periosteal tube with less than 100% superior translation of the metaphysis. Type IV injuries result from a complete disruption of the periosteal tube with superior and posterior translation of the metaphysis. Type V injuries result from a complete disruption of the periosteal tube with more than 100% superior translation of the metaphysis. Type VI injuries result from a subcoracoid displacement of the metaphysis.

Treatment Analysis

Results

Comments and citations found in the treatment sections for types III below are generally relevant and applicable to types III, IV, V, and VI because many references include the treatment of more than one type of injury. Some authors have grouped treatment of types IV, V, and VI with the consideration that nonoperative treatment is the method of choice for types I, II, and III. Numerous authors have combined their results for treating type IV and V. Citations that report the results of comparative treatment are very frequent, and may likewise be applicable to more than one type of injury. Many authors report simultaneously the results of treatment of both acute and chronic injuries. The reader is advised to access the specific article as needed for clarification when necessary.

The outcomes of the treatment of ACJ injuries have been expressed in most of the validated scoring systems used for the shoulder. These include the Bother Index of the Short Musculoskeletal Function Assessment (MFA), University of California at Los Angeles (UCLA), UCLA Modified AC Score, SST, ASES, various versions of Disabilities of the Arm, Shoulder, and Hand (DASH) score, L’Insalata Score, Shoulder Pain and Disability Index, Single Assessment Numeric Evaluation (SANE) score, Subjective Shoulder Value (SSV), Constant-Murley (C-M), Japan Orthopaedic Association score, Nottingham Clavicle Score, Oxford Shoulder Score, Neer Score, Poigenfurst Score, Walch Score, Visual Analog Scale (VAS) Pain, Short Form-36, Short Form-12 Physical Component Summary, and cosmesis. In addition, ACJ-specific evaluation tools (Imatani, Taft) have been proposed and used.

Mechanism of Injury

Although any number of physical events may be responsible for an injury, the mechanism of injury is a product of forces characterized by direction, magnitude, and point of application.

Direct

The most common mechanism for injury to the ACJ is a direct force that is applied to the superior aspect of the acromion process such as would occur with a fall onto the outer aspect of the shoulder with the upper limb in an adducted position ( Fig. 9-43 ). A falling object or a deliberate blow striking the superior acromion is a rarer mechanism of direct injury.

Acromioclavicular joint injuries most often result from a fall directly onto the point of the shoulder.

The AC ligaments offer the weakest resistance to the forces that push the acromion inferiorly and medially; stronger resistance is afforded by the intact clavicle and SCJ. Although isolated injuries are most common, AC injuries that occur in conjunction with clavicle fractures or SCJ injuries have been reported. More often, these two structures are spared as the force advances through the ACJ to act upon the CC ligaments. This is an extremely strong buffer, but should its integrity be overcome, any remaining forces are likely to be dissipated at the attachment sites of the deltoid and trapezius muscle.

An inferiorly directed force acting upon the superior lateral clavicle can cause more damage when the upper limb is abducted and the scapula is retracted, and can result in an inferior dislocation of the clavicle beneath the coracoid process.

Indirect

Indirect mechanisms of ACJ injury are exceedingly rare. A fall onto the adducted upper limb is apt to drive the humeral head into the inferior aspect of the acromion, subjecting the ACJ to variable degrees of injury ( Fig. 9-44 ). The integrity of the acromion process and glenohumeral stability is risked when extremely high superiorly directed forces are encountered. The ACJ may be injured by pulling or traction-like forces applied to the upper limb.

An indirect mechanism of injury that might occur from a fall onto the adducted upper limb results in complete tearing of the acromioclavicular ligaments while preserving the integrity of the coracoclavicular ligaments.

(From Rockwood CA, Green DP, eds. Fractures . 2nd ed. Philadelphia: JB Lippincott; 1984.)

Acute Lesions

Type I

Signs and Symptoms

Mild pain is well localized to the ACJ, and may be aggravated by movement or maneuvers that load the upper limb. Although the ACJ may be slightly swollen, the shoulders are not unusually asymmetric in appearance. Active motion through a full range is sometimes hindered due to pain. Tenderness to some degree is appreciated at the ACJ but not overlying the CC ligaments. Palpable deformity of the joint is not present.

Imaging

Sometimes, the plain films demonstrate mild soft tissue swelling that accompanies type I injuries. Otherwise, the ACJ appears normal and exhibits symmetry to the uninjured shoulder.

Pathoanatomy

The AC ligaments sustain a mild-to-moderate sprain, maintaining the ACJ integrity. The CC ligaments and the deltoid and trapezius muscles are normal.

Treatment

Virtually all type I injuries should be treated without surgery. *

* References .

Methods

A sling or immobilizer for the upper limb helps to place the shoulder at rest. During the first 24 hours, cryotherapy can help reduce pain and swelling. Nonnarcotic analgesics are usually adequate for successful pain management. As the pain subsides, motion quickly recovers, and external support is discontinued. Although formal physical therapy is usually not necessary, some authors have advised non–pain-provoking gentle range of motion exercises, closed chain exercises, and muscle rehabilitation. As the symptoms abate, isotonic progressive resistive exercises and open chain exercises are employed. Participation in contact or collision activities is best avoided until motion has fully recovered, tenderness has disappeared, and manual muscle testing is not painful.

Results

Full recovery of comfort and function is expected for type I injuries treated nonoperatively. This was confirmed in a report by Babe et al. Time lost in young athletes, including collegiate football players with type I and II injuries was 10 to 12 days. On occasion, symptoms are reported anywhere from 6 months to 5 years following injury. More than 90% of the time, the symptoms remain insignificant or reasonably well tolerated. Shaw et al. determined that patients receiving nonoperative care for type I injuries could expect symptoms to resolve, in most cases, by 12 months. Patients with symptoms at 6 months correlated with those who were symptomatic beyond 1 year.

Mouhsine et al. found a higher incidence of unfavorable outcomes, and were concerned that adverse outcomes were underestimated. In their study, operative treatment was necessary in 13% of type I injuries.

Complications

It is possible but quite unlikely that posttraumatic arthrosis would appear following a single type I injury. However, degenerative changes of the ACJ were reported in 56% of patients following a type I injury in the study by Mouhsine et al. In some instances, degeneration was accompanied by pain, and necessitated a reduction in activity. Rarely, osteolysis of the distal clavicle develops as a late phenomenon after this type of injury.

Author’s Preferred Method

The patient wears an upper limb sling as long as necessary to provide comfort and protection. Cryotherapy and nonnarcotic analgesics are preferred for pain management. Motion recovery is accelerated according to the level of pain. Resolution of ACJ tenderness and full, painless range of motion signal complete healing and return to unrestricted activities.

Type II

Signs and Symptoms

Significant pain is present both at rest and with attempted movement of the upper limb. Loss of functional range of motion is an accompanying complaint.

The shoulders might have an asymmetric appearance, with slight superior prominence of the distal clavicle. Depending on the interval from injury, ecchymosis in the vicinity of the ACJ may be present. Pain limits both active and passive range of motion as well as strength. The ACJ may be exquisitely tender, inhibiting reliable instability testing that might otherwise be positive for excessive excursion only in the AP plane. This can be demonstrated with a two-hand maneuver: With the clavicle and the acromion in grasp, opposing forces in the AP plane are applied to each bone, resulting in excessive movement of one with respect to the other when compared with the contralateral shoulder ( Fig. 9-45 ). A subtle rebound phenomenon of the lateral clavicle can result from the application of light inferiorly directed pressure to the superior aspect of the lateral clavicle. CC ligament tenderness is present.

Anteroposterior stress test of the acromioclavicular joint. The clavicle is held as stationary as possible ( A ) while straight anterior ( B ) and posterior ( C ) stresses are sequentially applied to the acromion process.

Imaging

Plain films can demonstrate subtle differences from the opposite, uninjured shoulder. The ACJ may be widened and have a slight incongruence resulting from the superior displacement of the lateral clavicle. The CC interval is normal and remains so when stress is applied ( Fig. 9-46 ). The utility of MRI to distinguish type II from type III injuries has been suggested.

Radiographic appearance of a type II injury to the right shoulder. With stress radiographs, the coracoclavicular distance in both shoulders ( A and B ) measures 1.5 cm. However, the injured right shoulder ( A ) has a widened acromioclavicular joint compared with the normal left shoulder ( B ).

Pathoanatomy

The AC ligaments are completely torn, resulting in disruption of the ACJ. The joint widens in the transverse and, less often, coronal planes. The CC ligament sustains a mild or moderate sprain that can result in slight widening of the CC interval. The deltoid and trapezius muscles are susceptible to partial detachment from the lateral clavicle. An extremely rare associated finding is fracture of the coracoid process.

Treatment

Nonoperative treatment is preferred for type II injuries. *

* References .

Methods

The same fundamental methods for treating a type I injury are used for a type II injury. A sling or immobilizer for the upper limb helps to place the shoulder at rest. During the first 24 hours, cryotherapy can help reduce pain and swelling. Nonnarcotic analgesics are usually adequate for successful pain management. As the pain subsides, motion quickly recovers, and external support is discontinued. Although formal physical therapy is usually not neces­sary, some authors have advised muscle rehabilitation, particularly of the scapular stabilizers. Later, as postinjury irritability subsides, more sophisticated exercises are employed. Participation in contact or collision activities is best avoided until motion has fully recovered, tenderness has disappeared, and manual muscle testing is not painful. This is usually a bit longer for type II injuries, and can require 6 weeks or more.

Some authors have been tempted to treat the deformity associated with type II injuries. These methods use the concept of opposing forces applied to the clavicle and the acromion in a manner that theoretically reduces the subluxation and sustains the congruity of the ACJ until healing has been completed. Nearly 40 different methods have been described, all with limitations that include noncompliance, skin breakdown, and loss of reduction. Most authors have abandoned an active form of nonoperative treatment in favor of a simple sling that is worn for a short time. At the risk of further, more serious injury while the ligaments heal, it is generally recommended that forceful maneuvers with the upper limb as well as contact or collision activities be avoided for up to 12 weeks.

Results

Shaw et al. determined that patients receiving nonoperative care for type II injuries could expect symptoms to resolve, in most cases, by 12 months. Patients with symptoms at 6 months correlated with those who were symptomatic beyond 1 year.

Even with slight degrees of incongruence of the ACJ, full recovery with return to their preinjury status is expected in most cases. However, sometimes, despite apparent ligament healing, discomfort and dysfunction linger unexpectedly long after the acute event, which is attributable to residual instability, infolded capsule, articular cartilage injury, meniscal injury, residual joint incongruity, and repetitive residual trauma. A theoretical mechanism for such instability has been demonstrated in a cadaveric study. Occasionally, such posttraumatic sequelae may necessitate operative treatment. Mikek reported a mean follow-up of 10.2 years for type I and II injuries, and found that more than half of patients will experience symptoms and obtain noticeably lower functional scores when compared with the uninjured shoulder. In their series, treatment was not necessary for symptoms. While joint dimensional differences were detected by ultrasonography, radiographic degenerative changes were not observed.

Weakness has been reported. Factors that can contribute to persistent symptoms include posttraumatic lateral clavicle osteolysis, intra-articular soft tissue entrapment, free-floating chondral or osteochondral bodies, and an unstable intra-articular disk.

Complications

Reinjuries that occur before the ligament is completely healed can unexpectedly convert a type II into a type III injury. Skin breakdown beneath immobilizing devices forces abandonment of type II injury treatment while the integument heals.

ACJ incongruence, especially in the setting of extreme use, can precipitate symptomatic arthrosis. Such degenerative changes (arthrosis and osteolysis) of the joint have been reported, with an incidence as high as 70% ( Fig. 9-47 ).

Traumatic osteolysis and remodeling of the distal end of the clavicle.

Author’s Preferred Method

Nonoperative treatment is directed toward effective pain management by means of a sling, cryotherapy, and narcotic analgesics. This can require 1 to 2 weeks. Pool therapy is encouraged, and hastens the recovery of active, pain-free movement. Once motion fully recovers and tenderness abates, strength rehabilitation is initiated. Forceful use and contact and collision activities are avoided for 12 weeks.

Type III

Signs and Symptoms

The patient has moderate to severe shoulder pain. Pain is not only localized to the ACJ but can also emanate from the CC and periscapular regions. Motion, either active or passive, in any direction and to any degree, exacerbates the pain. Neurologic symptoms are very rare.

The patient appears acutely injured, often with the affected upper limb postured in a protective and stress-relieving manner: supported higher than usual by the opposite hand and arm and braced against the torso. The soft tissue overlying the acromion may be abraded or contused. Ecchymosis might or might not be present. The lateral clavicle is prominent beneath the skin, and the remaining shoulder has a drooped appearance ( Fig. 9-48 ). The clavicles remain relatively level, while the weight of the upper limb displaces the clavicle inferiorly ( Fig. 9-49 ).

This type III injury results in an uneven appearance to the shoulders. The injured right shoulder does not droop in this instance secondary to pain that results in splinting by the foreshortened trapezius muscle.

The type III injury results in deformity because of the downward displacement of the acromion process together with the upper limb, not because of superior displacement of the clavicle.

Motion and strength assessment are unreliable due to pain. Tenderness is more diffuse: ACJ, distal clavicle, and CC region. Pain permitting, global instability of the clavicle is demonstrable. With the clavicle stabilized with one hand, it may be possible to reduce or nearly reduce the dislocated ACJ by placing the other hand beneath the flexed elbow and applying a superiorly directed force sufficient to often completely obliterate the visible deformity. The deformity returns when the force is removed. According to some authors, type III may be distinguished from type V if a well-performed shoulder shrug reduces the AC dislocation, implying intact deltotrapezius fascia.

Imaging

The ACJ is incongruent from widening, and there is at least 50% contact loss between the acromion and the clavicle. In the majority of type III injuries, there is no contact between the articular surfaces of the lateral clavicle and the medial acromion. The CC interval is widened 20% to 100% ( Fig. 9-50 ). An optimally positioned scapular Y view is particularly helpful ( Fig. 9-51 ). Stress views are a consideration, and demonstrate similar findings ( Fig. 9-52 ). The utility of a cross-body adduction view has been cited by some authors as a discriminatory imaging tool to highlight ACJ instability persisting beyond the acute injury period of 3 to 6 weeks. In addition to ligamentous injury, joint effusion, bone marrow edema (bone bruise), edema, and detachment of the deltoid and trapezius muscles to some degree are often observed on MRI.

Anteroposterior view of a type III injury. The articular surfaces of the acromioclavicular joint are widely displaced. The coracoclavicular interval is significantly increased.

High-quality scapular Y view.

Anteroposterior stress views demonstrating no appreciable difference between the unweighted and the weighted views. A, Contralateral side. B, Unweighted. C, Weighted.

Pathoanatomy

The lateral end of the clavicle is exposed as a result of stripping of the periosteal sleeve. The AC and CC ligaments are torn completely, resulting in dislocation of the ACJ. Midsubstance tearing and avulsion from the clavicle occur nearly equally. In comparison to the normal shoulder, the CC interval may be widened by up to 100%. There is a high probability that deltoid and trapezius muscles are detached from the distal clavicle.

Treatment

The treatment of type III injuries has been, is, and will remain in the foreseeable future, a subject of significant controversy. About every 10 to 20 years, certain orthopedic surgeons are compelled to solicit opinions regarding the treatment of type III injuries from other orthopedic surgeons, including academic department chairs, by the distribution of survey material. Although the purpose of these surveys varies, treatment styles emerge as beacons that divinely light the way for practitioners of the period. This was true in 1974, and it remained true in 2007, as expressed by the conclusions of Nissen et al.: Nonoperative symptomatic treatment is preferred, and operative treatment should include resection of the lateral clavicle, reconstruction of at least the CC ligaments with either sutures or local ligament graft, and protection of the reconstruction with rigid fixation between the coracoid and the clavicle. Subsequently, the ever-growing body of pertinent literature has been exhaustively perused by several authors, from which have emerged current concepts reviews, systematic reviews, meta-analyses, and summaries of evidence-based publications attempting to reveal the best evidence for the treatment of all AC dislocations, and in particular, acute type III injuries.

Nonoperative Treatment

Nonoperative treatment is recommended by many authors, and it is the conclusion of the extensive reviews by Ceccarelli et al., Spencer, and Phillips et al. * Hootman concluded that evidence current at the time did not support operative treatment to achieve better functional outcomes and patient satisfaction. Tamaoki et al. concluded from their review of randomized controlled trials that substantive evidence supporting the decision or timing of surgical treatment did not exist at the time of their review. Beitzel et al. and Modi et al. reached similar conclusions. From their review, Korsten et al. stated that for these lesions, there was no conclusive evidence for the treatment. For the majority of patients, nonoperative is the preferred form of treatment for type III injuries. This specifically includes patients such as contact and collision athletes and cyclists, who remain at high risk for repeated injuries to the ACJ. There are no contraindications to nonoperative treatment.

* References .

Methods

Taking advantage of the fact that the dislocated joint is easily reducible in the acute setting, some authors have advocated a closed reduction accompanied by external immobilization. Various techniques and methods have been used through decades of experience with the type III injury. None has proved any better than another, and each has its own pitfalls.

Recognizing the difficulties with maintaining a satisfactory reduction, many authors have recommended symptomatic treatment only with sling support for the upper limb. Motion and strengthening exercises are initiated as the resolution of pain permits. The pace of therapeutic exercise parallels the resolution of symptoms. Closed and open chain exercises are integrated following a 6- to 12-week period of appropriate mobilization. With only minimal scapular dyskinesia remaining, release to activities without restriction may be possible. The role of orthotics to facilitate periscapular muscle rehabilitation is uncertain. Return to higher levels of activity demands enthusiastic participation in a rehabilitation program that emphasizes progressive restoration of strength, function, and sport-specific drills. An excellent in-depth review of nonoperative care for not only type III but types I and II injuries forms the basis of best practice guidelines.

Results

Excellent and good results are obtainable with nonoperative treatment. * An earlier return to work and other activities is often possible. Korsten et al. concluded that the price paid for a shorter duration of rehabilitation was a worse cosmetic outcome. Delays to return to sport are sometimes as long as 2 months. At mid- to long-term follow-up, 90% to 100% of patients have very favorable outcomes. At the completion of rehabilitation, strength and endurance approaches that of the opposite upper limb.

* References .

Overall, satisfactory results obtained with nonoperative treatment derived from a meta-analysis review of the literature was 87%. Calvo et al. demonstrated that there were no differences in the outcome of nonoperative and operative treatment. Spencer searched 469 publications to identify nine studies with level II and level III evidence that compared nonoperative and operative treatment. From this in-depth analysis of what he considered limited low evidence, he concluded that nonoperative treatment was more appropriate because the results of operative treatment “were not clearly better and were associated with higher complication rates, longer convalescence, and longer time away from work and sport.” De Carli et al. reached a similar conclusion. They also noted poor satisfaction with the cosmetic results in 50% of the patients.

Many authors have observed that pain, flexibility, range of motion, and strength after nonoperative treatment are at least equal to, if not better than, those after operative treatment. * If differences exist, they are slight or activity specific. A higher rate of complications and a slower return to work has been observed with operative treatment. Although not all authors agree, heavy manual laborers and high-performance overhead athletes such as pitchers might not fare as well with nonoperative treatment because they are less likely to be completely pain-free and have normal function. Pain and weakness have been reported in up to 50% of patients.

* References .

After treatment, the extent of congruity or incongruity, the presence of OA, and the development of heterotopic ossification appear to have no correlation with the outcome, regardless of the treatment method. An exceedingly rare consequence of an apparent type III injury is AC fusion ( Fig. 9-53 ).

High-energy shoulder girdle trauma with resultant fractures and apparent type III injury that went on to spontaneously fuse.

Complications

Although not really a complication, the asymmetry of the shoulder becomes obvious after swelling resolves ( Fig. 9-54 ). The extent of deformity may be unexpected and undesirable, perhaps necessitating treatment. Pressure-related skin breakdown is possible with certain types of external immobilization.

Shoulder asymmetry resulting from type III injury. Note the prominence of the left distal clavicle and the inferior attitude of the shoulder compared with the contralateral shoulder, which had been previously treated for acromioclavicular injury.

Persistent pain and weakness are the most common late sequelae with type III injuries. Ossification is often present in the CC interval at the completion of the period of healing ( Fig. 9-55 ). The significance of the ossification is uncertain, but it does not seem to disturb recovery.

Abundant heterotopic ossification resulting from a type III dislocation.

Osteolysis of the distal clavicle has been reported. The risk of posttraumatic OA has been shown to be lower after nonoperatively treated injuries, as is the incidence of CC ossification.

Author’s Preferred Method

Nonoperative treatment of type III injuries differs only slightly from the treatment of type II injuries. The cornerstone is effective pain management by means of a sling, cryotherapy, and narcotic analgesics. This can require 1 to 2 weeks. Pool therapy is encouraged, and hastens the recovery of active, pain-free movement. Once motion fully recovers and tenderness abates, strength rehabilitation is initiated. Protective measures are more relaxed for type III injuries than for type II injuries.

Operative Treatment

Surgical treatment of ACJ dislocations has been performed since 1861, and appears to continue to play an important role for some clinicians in the care of carefully selected patients with these injuries. * On the basis of the available evidence, firm conclusions are elusive regarding operative versus nonoperative treatment, the timing of surgery, open versus arthroscopic surgery, and choice of surgical procedure with respect to (1) ACJ fixation, (2) fixation between the coracoid and the clavicle, and (3) ligament reconstruction. A meta-analysis that analyzed operative versus nonoperative treatment recognized that, from the dearth of well-designed investigations, the emergence of the optimum treatment was impossible. From their comparative study, De Carli et al. concluded that “… results cannot support the routine use of surgery to treat type III ACJ dislocations.”

* References .

Several investigators analyzed their data to compare and distinguish the results of early (acute) versus delayed (chronic) treatments. Song et al. performed a systemic review of early versus delayed operative treatment. From their review, they opined that early intervention may have an edge for improved functional outcomes and quality of reduction; however, their opinion was supported only by low-level evidence studies.

Indications

Operative treatment offers the only opportunity for structural and cosmetic restoration. Reserving it for heavy manual laborers, patients 25 years and younger, athletes, and frequent overhead users has been suggested. That said, in one review, there appeared to be no evidence to support the claim that the outcome of operative treatment is better than that of nonoperative treatment. Currently, proof is lacking that special groups of patients mandate consideration for operative treatment. Trainer et al. advocated a 3-month period of nonoperative treatment to optimize the recovery of motion, strength, and shoulder coordination. Patients with persistent symptoms were offered surgical treatment.

Contraindications

Extremely low demand and inactive patients might have nothing to gain with operative treatment. Significant skin abrasions can delay operative treatment.

Methods

Sage and Salvatore reported primary AC ligament repair with joint meniscus reinforcement over 50 years ago. Even recently, direct ligament repair backed up with a temporary CC screw has been described ( Fig. 9-56 ). The ligament could be coapted with repair in 38% and aligned and approximated in 62% with the use of autogenous palmaris longus or in its absence, half of the flexor carpi radialis, along with nonabsorbable suture. Rushton et al. stabilized the ACJ via two pairs of transosseous tunnels. Sandmann et al. reported a triple cerclage technique whereby one trans-ACJ and two CC sutures maintained the joint reduced. Sobhy used a single nylon tape around the coracoid, through the clavicle, and across the ACJ.

Acromioclavicular, trapezoid, and conoid ligament suture repair protected with a coracoclavicular screw.

(Modified from Assaghir YM. Outcome of exact anatomic repair and coracoclavicular cortical lag screw in acute acromioclavicular dislocations. J Trauma . 2011;71[3]:E50-E54.)

The ACJ may be reduced by a closed or open technique and repaired with metallic wires or pins that temporarily transgress the articular surfaces of the acromion and the clavicle. * The importance of supplemental techniques performed in conjunction with the method of internal fixations has been emphasized by some authors. These techniques include the additional repair of the AC and CC ligaments, AC and CC ligament repair reinforcement with the intra-articular disk, repair or imbrication of the deltoid and trapezius attachments, adjacent soft tissue transfers (CAL, biceps short head), lateral clavicle excision, autologous fascia lata suspension, CC suture augmentation, and clavicle osteotomy. * The duration of temporary fixation is typically longer than 6 weeks but varies with the technique. Motion restrictions are mandatory to protect the integrity of the pins or wires until their removal. Thereafter, motion and strengthening programs are introduced.

* References .

* References .

Nearly 40 years ago, Balser published in German the first account of the use of a hooked plate. While the appearance of the plates may vary, the principles since inception of the concept have not wavered. The open reduction and internal fixation method with a precontoured hooked plate affords rigid internal fixation and spares the articular surfaces of the joint ( Figs. 9-57 to 9-60 ). † The device has the advantage of simultaneously stabilizing the CC (superior-inferior) and the AC (anterior-posterior) ligaments. With no other means of fixation augmentation, stability is afforded by scar that accompanies healing of the surrounding soft tissues. Several authors have described the use of supplemental fixation including CC fixation with polydioxanone sulfate (PDS) sutures, direct CC ligament repair, suturing the capsuloligamentous sleeve of the ACJ, additional screw fixation through the plate, and transposition of the CAL as per Weaver and Dunn. Some authors have advised early range of motion within a day of the operation. Others permit a limited range after a brief period of protection. Implant removal is reported between 4 and 24 weeks. Infrequently, the duration of implant retention is as long as 7 to 13 months. Some authors do not routinely remove the device. Asymptomatic patients may occasionally refuse a secondary operation to remove the plate. Immediately after plate removal, vigorous efforts toward motion and strength recovery are initiated.

Hook plate for type III injury enabling excellent restoration of anatomy.

AC-Hakenplatte tifix plate.

(From Kienast B, Thietje R, Queitsch C, et al. Mid-term results after operative treatment of Rockwood grade III-V acromioclavicular joint dislocations with an AC-hook-plate. Eur J Med Res . 2011;16[2]:52-56.)

Hook plate application with a supplemental coracoclavicular screw through the plate.

(From Darestani R, Ghaffari A, Hosseinpour M. Acromioclavicular joint fixation using an acroplate combined with a coracoclavicular screw. Arch Trauma Res . 2013;2[1]:36-39.)

Acroplate (aap Implantate AG).

(From Cîrstoiu C, Rădulescu R, Popescu D, et al. Acroplate—a modern solution for the treatment of acromioclavicular joint dislocation. J Med Life . 2009;2[2]:173-175.)

† References .

Liu et al. designed and utilized the articulated micromovable and anatomic AC plate (MAAP; Fig. 9-61 ). Ryhanen et al. have described the use of a CC nitinol C -shaped hook implant (AC-Hakenplatte tifix). Insertion with arthroscopic assist has been described.

Articulated hook plate applied across the acromioclavicular joint to permit some joint motion.

(From Liu Q, Miao J, Lin B, et al. Clinical effect of acute complete acromioclavicular joint dislocation treated with micromovable and anatomic acromioclavicular plate. Int J Med Sci . 2012;9[8]:725-729.)

The clavicle and the coracoid process offer secure points of fixation that can be used to maintain either a closed or open reduction of the ACJ. These stabilization techniques may be used independently or as a supplement to other forms of either temporary or permanent means of fixation. In addition, direct repair of the AC and CC ligaments performed concurrently with the CC stabilization can be considered.

Bosworth was the first to describe the use of a screw between the clavicle and the coracoid process to achieve ACJ stability. Others have used the technique as described by Bosworth or have made modifications in the technique ( Figs. 9-62 and 9-63 ). * Garrigues et al. described an intraoperative fluoroscopically aided method for the proper identification of the base of the coracoid process to facilitate the placement and security of purchase of a percutaneously placed CC fixation device such as a screw. The method is applicable to more recent CC devices that incorporate cortical suture buttons, described below.

Satisfactory reduction with the Bosworth screw maintained by coarse threads in the base of the coracoid process.

(From Rockwood CA, Green DP, eds. Fractures . 2nd ed. Philadelphia: JB Lippincott; 1984.)

Excellent reduction with restoration of a normal coracoclavicular interval with a coracoclavicular screw. A, Anteroposterior stress radiograph of the right shoulder in a patient with a type III acromioclavicular dislocation. B, A stress film of the left shoulder revealed it to be normal. C, Postoperative radiograph showing the acromioclavicular joint reduced and held temporarily in place with a special modified coracoclavicular lag screw.

(From Rockwood CA, Green DP, eds. Fractures . 2nd ed. Philadelphia: JB Lippincott; 1984.)

* References .

The exploration and debridement of the ACJ has been advocated, but some authors do not consider this an essential component of treatment. In the setting of CC stabilization, the necessity for the direct repair of the CC ligaments is uncertain. Bosworth and others have elected to leave the ligaments unrepaired. Some authors have emphasized the importance of supplemental CC ligament repair at the time of CC stabilization. Repair of the trapezius and deltoid attachments can contribute to the overall strength of the repair and has been emphasized by several authors.

There are significant variations in the recommendations for postoperative immobilization, restrictions, rehabilitation, the necessity for implant removal, and the timing of implant removal.

The extraosseous, intraosseous, and transosseous placement of absorbable sutures, nonabsorbable sutures, wires, cables, anchors, and synthetic ligaments have been used for indirect ACJ stabilization ( Fig. 9-64 ). * If the material is passed through drill holes in the clavicle, in the coracoid, or in both, the reduction is potentially more accurate, and the outcome may be improved ( Fig. 9-65 ).

Acute type III injury. A, Preoperative. B, Final postoperative treatment using encircling permanent sutures with distal clavicle resection.

A method of coracoclavicular reconstruction with loops passing through a prepared hole in the clavicle.

(Modified from Dimakopoulos P, Panagopoulos A. Functional coracoclavicular stabilization for acute acromioclavicular joint disruption. Orthopedics 2007;30[2]:103-108.)

* References .

Various types of synthetic ligaments have been used for reconstruction. The ligament is looped beneath the coracoid process and then passed through tunnels in the clavicle and secured.

Reconstructions that utilize CC suture techniques date back to 1996. They are espoused to pose less risk to adjacent neurovascular structures with comparable strength of the native ligaments, and are minimally invasive. Additional advantages are less blood loss, shorter operative time, and no secondary incision or hardware to remove. The relatively simple suspension method of Huang et al. incorporated subcoracoid and transclavicular passage of large, nonabsorbable sutures through tunnels at the sites of the conoid and trapezoid footprints. Concomitant CC ligament repair can be performed, and is believed by Lu et al. to be one of the most important components of their repair technique. The potentially less invasive aspects of some of the newer implants, such as TightRope (Arthrex) has been touted.

Concomitant supplemental AC ligament repair has been described. Zhang et al. modified the technique by placing a stabilizing suture anchor across the ACJ.

Transposing the CAL into the resected distal clavicle was suggested in 1917 by Cadenet, advocated by Nevisier in the early 1950s, and was reported by Weaver and Dunn in 1972, who used the technique for both acute and chronic type III injuries. Other authors used the CAL in the same or a similar manner. Modifications of the Weaver-Dunn method were subsequently described by other authors, and were felt to be especially useful for a more active population. The CAL, with or without an accompanying piece of bone, may be transferred to the clavicle with or without supplementary fixation. Such concepts of CC augmentation in conjunction with the Weaver-Dunn operation have been supported by the results of experimental studies.

A Cadenet procedure was described by Cerciello et al. that has been modified by the addition of CC cerclage, CC ligament remnant suturing, and transarticular dual K-wire fixation of the reduced ACJ.

Another method for incorporating the CAL was originally described by Vukov. It differs in that the ligament is left intact, and the reduced distal clavicle, having undergone a limited distal resection, is “elastically” affixed with sutures passed through transosseous tunnels in the clavicle.

In 1941, Gurd described excision of the distal clavicle as a treatment for type III injuries. In the same year, Mumford reported the same operation for type II injuries, and advocated that treatment of symptomatic type III injuries should also include CC ligament reconstruction as well as DCR. The role for excision of the lateral clavicle at the time of surgical treatment of a type III injury is not clear. There were no differences reported between CC ligament fixation alone and CC fixation with excision of the lateral clavicle. Likewise, when the outcome of AC fixation and CC ligament repair with and without lateral clavicle excision were compared, no differences were observed, with the exception of the more frequent appearance of degenerative changes when the lateral clavicle was retained.

In 1965, muscle transfers were suggested as a reasonable treatment for type III injuries. The tip of the coracoid process, including the coracobrachialis and the short head of the biceps, was osteotomized and attached to the clavicle. Scattered reports appeared later, some with slight modifications of the technique. Transposition of the biceps long head or a slip of the conjoined tendon have been described as substitution methods for the CC ligaments. Jiang et al. described reconstruction with a proximally base-conjoined tendon transfer supplemented with transclavicular clavicular sutures anchored in the coracoid process.

In the last 8 years, there has been an explosive trend to seek a more anatomic and biologic solution for CC ligament reconstruction. The notion is that reliable restoration of function and comfort is dependent on the durable restoration of anatomic parameters, including the congruity and stability of the ACJ. The method often includes substituting synthetic or biologic material for each CC ligament separately entering the clavicle and coracoid as close as possible to the footprints of the native ligaments.

Early biomechanical investigations followed the initial report of Jones et al., who described the surgical technique of using autologous semitendinosus tendon, and applied the method in a revision case. Mazzocca et al. were among the earliest advocates for so-called anatomic CC ligament reconstruction (ACCR) of the CC ligament complex using autograft or allograft tendon ( Fig. 9-66 ). Additional biomechanical work by Lee et al. led to the first clinical application and report of the use of autologous hamstring tendon by Nicholas et al. Subsequently, some form of “anatomic” repair or reconstruction that utilizes autogenous or allogenous tendon is being reported more frequently as a preferred method of treatment. ACJ reconstruction that utilizes the same graft for CC reconstruction has been described. In 1928, a similar method was eloquently described by Bunnell, who incorporated a fascial graft weave that offered stability between the clavicle and the scapula at both sites ( Fig. 9-67 ).

A coracoclavicular reconstruction that is considered anatomic uses autograft or allograft tendon.

(Modified from Mazzocca AD, Conway J, Johnson S, et al. The anatomic coracoclavicular ligament reconstruction. Oper Tech Sports Med . 2004;12:56-61.)

Various perspectives of a cadaver shoulder with rubber tubing depicting Bunnell’s fascial weave technique anatomically stabilizing the clavicle and the acromion.

(From Bunnell S. Fascial graft for dislocation of acromioclavicular joint. Surg Gynecol Obstet . 1928;46:563-564.)

An arthroscopic method for treating type III injuries was first described by Wolf and Pennington. In this study, either sutures or allograft semitendinosus were used for CC stabilization. Lafosse et al. presented the initial description of an arthroscopic technique that did not require autologous tendon. Since then, other arthroscopic methods using the CAL, various types of sutures, or tendon autograft or allograft have been described. *

* References .

Arthroscopic-assisted percutaneous CC screw insertion has been described. Somers and Van der Linden reported transclavicular sutures with coracoid suture anchors performed arthroscopically. Further arthroscopic techniques have been investigated, recommended, and performed. Appropriate visualization with the arthroscopic method is by one of two ways: an intra-articular approach via the rotator interval and a subacromial approach. In general, arthroscopic CC stabilization may enable the safer and more accurate placement of materials used for stabilization. Perhaps an additional advantage is the finding of coexisting lesions, which often require additional treatment; the incidence ranges as high as 25% to 35.3%. These include SLAP tears, disorders of the biceps long head, Bankart lesions, glenoid cartilage fissures, humeral head or glenoid minor chondral defects, and partial- and full-thickness RC tears.

In 2006, the use of a contoured, metallic, buttonlike device designed to rest against the outer cortex of bone and a nonabsorbable reinforced suture system was first described by Hernegger and Kadletz. In 2007, the technique of using such buttons in a flipping manner was presented. Many authors have since reported on the use of one or more pairs of flipping CC buttons spanned by nonabsorbable sutures, both open and arthroscopic assisted. *

* References .

Experimental Investigations.

Impetus to find an effective way to render the injured ACJ stable has resulted in scores of investigations that have attempted to analyze and compare the multitude of methods, old and new.

Thirty years ago, upon comparing techniques for AC fixation, Kiefer et al. found that the most rigid fixation was provided by the Bosworth screw. Restoration of ACJ congruency using seven different methods of fixation between the clavicle and the coracoid were studied by Jerosch et al. They demonstrated satisfactory vertical stability with all of the techniques. Excessive horizontal translations between the acromion and the clavicle were demonstrated with all of the loop techniques and the Weaver-Dunn reconstruction, whereas the joint anatomy was best restored with the Bosworth screw or bone anchor reconstruction. Harris et al. showed that Bosworth screw placement with bicortical purchase provided strength and stiffness as good as or better than the intact CC ligaments. In a cadaver model comparing the biomechanical function of different surgical procedures for ACJ dislocations, the Rockwood screw resulted in decreased primary and coupled translations, and showed increased stiffness. CC sling and CAL transfer resulted in increased primary and coupled translations. In situ forces increased for all three reconstructions.

Additional experimental screw fixation for CC ligament reconstruction has been investigated. In this study, CC fixation comparing stainless steel and bioabsorbable 4.5-mm screws showed no difference in pullout strength and fixation strength that exceeded the intact CC ligaments. In a study by Motamedi et al., 6.5-mm CC screw fixation with one-cortex purchase was significantly weaker and stiffer than the intact CC ligaments.

In a cadaver model, McConnell et al. performed reconstructions with CC loop (No. 5 polyester fiber [Mersilene]), CC screw, and hook plate fixation, and subjected the specimens to stiffness and failure testing. Hook plates reproduced physiologic stiffness more closely. The CC screw failed at the highest load and was the stiffest reconstruction, and the CC loop was the least stiff. In a comparative investigation, Fialka et al. identified the poor mechanical properties of the Bosworth technique, especially at more than 90 degrees of abduction. While transarticular K-wires were rigid and stable, a high rate of migration was observed. The most favorable mechanical stability was with the ligament augmentation and reconstruction system (LARS) device.

In an in vitro model of the ACJ, Nüchtern et al. not only showed the physiologic stiffness of the hook plate but also showed the relevant role of the ACJ ligaments for restoring physiologic properties about the ACJ at the time of reconstruction for complete dislocations. Kim et al. investigated changes in 3D motion taking place at the ACJ with and without the presence of a hook plate. In patients who had undergone hook plate for distal clavicle fracture, they showed diminished internal rotation (mean, 16 degrees) and increased anterior translation (2 mm) of the clavicle with respect to the medial acromion compared with the nonoperative side.

Wickham et al. compared the maximum tensile strength and resistance to deformation of one or two strands of Ethibond No. 5 suture, one or two strands of polyester (Mersilene) tape, size 0 polydioxanone braid, and size 2 poly( l -lactic acid) suture braid in a CC model, and demonstrated the significantly greater tensile strength of poly( l -lactic acid) sutures. Two strands of polyester tape, polydioxanone braid, and poly( l -lactic acid) braid exhibited similar resistance to deformation with cyclic loading but performed significantly better than the other test samples. Failure loads comparable to the intact CC ligament (725 N) were observed in the experimental reconstructions performed with braided polydioxanone and braided polyethylene passed around or through the clavicle; failure loads of a 6.5-mm screw with single-cortex purchase in the coracoid were around one-half those of the intact CC ligaments. The strength and stiffness properties of the braided polyethylene suggested that it could serve effectively as an augmentation for ACJ repairs and reconstructions. Martetschläger et al. concluded that inferior biomechanical properties of PDS sutures used to reconstruct the AC and CC ligaments were responsible for failure to achieve vertical stability.

In experimental repairs performed by Wellmann et al., loop or button fixation of polydioxanone sutures exhibited significantly higher ultimate loads than did coracoid suture anchor repairs with No. 2 high-strength polyethylene (Ultrabraid) sutures. In a cadaver model, ACJ congruity was not restored by any CC loop fixation through drill holes in the clavicle but was more closely restored when the drill hole was placed more anteriorly in the clavicle.

Different configurations of synthetic suture direction and orientation have been investigated. Abat et al. determined that passage via double tunnels in both the clavicle and the coracoid results in a reconstruction that is biomechanically more like the native state compared with a V configuration, where only one tunnel is in the coracoid. Schliemann et al. compared the coracoid side flip button tendon graft with tendon looping around the coracoid with synthetic suture augmentation, and found the biomechanical properties of the reconstruction to be comparable. The benefit of passing the reconstruction material transosseous as opposed to extraosseous is less abrasive changes observed in the material and the adjacent bone in contact with the material. The use of a transosseous TightRope method rather than fiber mesh cerclage for such augmentation resulted in less translational motion. Lädermann et al. demonstrated the advantages of AC and CC cerclage over V -shaped orientation synthetic suture (TightRope) CC reconstruction and hooked plate. Their work suggested that CC cerclage might better reproduce the native ligament stiffness.

Tunneling through the bone as opposed to passage around the bone is seen by many as the most effective option to create the so-called “anatomic” reconstruction that hopes to restore normal kinematics and forces for stabilizing the ACJ. Various ways of creating tunnels in the clavicle and coracoid for passage of synthetic sutures with or without augmentation biologic grafts and securing different holding sutures over different types of cortex-preserving devices (buttons) have been described. Beitzel et al. identified no difference between one and two tunnels in the clavicle. As a result of in-depth biomechanical investigations and early clinical experiences, recommendations are coming forth for tunnel sizing, positioning, and spacing in the clavicle and coracoid process. Rylander et al. showed the benefits of a 4-mm versus a 6-mm tunnel in the clavicle. When loaded to failure, drilled specimens failed by button pullout, and looped, undrilled coracoids failed by fracture. According to Spiegl et al., hamstring through 6-mm tunnels significantly weakened the clavicle compared with cortical button and suture through 2.4-mm tunnels. Dumont et al. determined that clavicle strength or resistance to ultimate failure were not enhanced by the presence of interference screws placed in the tunnel to achieve tenodesis. Ferreira et al. showed that load-to-failure testing of a 6-mm tunnel through the coracoid process was significantly higher when the entry and exit of the tunnel was center and center or medial and center, respectively. The mechanism of failure was predominantly in the repair construct and not a fracture of the coracoid. In a 3D virtual shoulder model, Coale et al. were able to highlight the significant risks when attempting to restore the anatomic footprint of the conoid and trapezoid ligaments. In their model, it was not possible to create reconstruction tunnels that both restored anatomic ligament footprints and maintained, without significant risk of cortical breach and fracture, the integrity of the bone through which the tunnel passes. In a similar model, Xue et al. observed a 90.5% incidence of clavicle or coracoid cortical breach with collinear drilling of a 4-mm tunnel based on anatomic landmarks. Alternately, noncollinear drilling of the coracoid tunnels determined by the ligament footprint on the superior aspect of the coracoid process directed to the inferior surface to a position that respected an adequate bony bridges and the cortical margins did not result in cortical breach. They also calculated the angulation of the coracoid tunnels. The advantages on non–image-guided navigation over drill guide technique in cadavers for the accuracy of tunnel placement has been demonstrated in a cadaver investigation. In a matched-pairs cadaver study of coracoid tunnels, Campbell et al. determined that cortical button fixation strength was greater with a 4.5-mm tunnel compared with a 6-mm tunnel. In their study, central and proximal fixation was stronger than eccentric and distal fixation. Experimental reconstructions with semitendinosus graft have shown the ability to reproduce peak loads equivalent to that of the native CC ligaments. Compared with the native ligaments, stiffness is lower. Walz et al. demonstrated the efficacy of the dual tunnel TightRope technique with buttons in reproducing and exceeding the vertical and horizontal forces of the native CC ligaments.

Reconstructions that incorporated the ACJ restored the biomechanical characteristics of the native ACJ specifically by reducing undesirable AP translations that could not be overcome by isolated CC techniques. Nüctern et al. also identified the significant role in joint stiffness contributed by the AC ligaments. Using a constant technique of CC ligament reconstruction while comparing four different AC reconstruction, Beitzel et al. showed the benefit of the additional ACJ reconstruction and the superiority of the stability of the direct wrapping and suturing technique. The utility and efficacy of reconstructing the ACJ with an intramedullary free semitendinosus graft in vitro have been shown. Garg et al. demonstrated the superiority of intramedullary over various extramedullary ACJ reconstructions. Shu et al. documented the effectiveness of a novel reverse CAL transfer when compared with isolated CC ligament reconstruction and comparable to intramedullary allograft tendon. When AC and CC reconstructions are simultaneously performed, significantly greater horizontal stability has been shown. Kowalsky et al. showed that the AC ligaments rather than the articulation of the distal clavicle may have a protective role for a CC reconstruction, hence a supportive rationale for DCR, when necessary. On the other hand, biomechanical testing of the reconstruction may mirror the native state even when the AC ligaments are not included.

The CAL transposition as used in the Weaver-Dunn reconstruction is, experimentally, one-fifth the ultimate load of the intact CC ligament. Deshmukh et al. demonstrated that the Weaver-Dunn method was significantly more lax than the native ligaments. In their model, CC augmentation was less lax than the Weaver-Dunn reconstruction but more lax than the native ligament. Harris et al. showed that CAL transfer in the manner of Weaver-Dunn was the weakest and least stiff experimental reconstruction when compared with the intact CC ligament, CC screw fixation, woven polyester vascular graft slings, and suture anchor technique with No. 5 braided polyester. The Weaver-Dunn procedure, even with modifications, fails to reproduce the load-to-failure durability of the intact AC and CC ligament complex. In a cadaver model that assessed the motion between two points fixed respectively on the distal clavicle and acromion process, an unaugmented Weaver-Dunn reconstruction demonstrated significant multiplanar laxity when compared with an intact joint. In the same study, a reconstruction augmented with CC suture with anchor into the coracoid was neither more nor less mobile than the intact joint. In contrast, LaPrade et al. recognized that when the Weaver-Dunn reconstruction was performed with coracoid transclavicular cerclage, motion at the ACJ was restored to near-normal values. In their comparative study, Lee et al. drew similar conclusions, but also documented the superiority of the transclavicular coracoid cerclage that used semitendinosus. From their investigation, Luis et al. concluded that for sufficient resistance to undesirable ACJ motion, the Weaver-Dunn reconstruction should always be accompanied by supplemental stabilization methods. A similar conclusion was reached by Wellman et al., who used only the medial one-half of the CAL together with polyester cord secured with flip buttons. Transfer of the CAL as per Weaver-Dunn at the time of CC hamstring tendon grafts did not yield superior stability or biomechanical strength. Several biomechanical studies have been completed that illustrate that ACCR more closely approximates the stiffness of the CC ligament complex and produces less anterior-to-posterior translation at the ACJ as compared with the Weaver-Dunn procedure. Other studies have identified the inferior biomechanical properties of the Weaver-Dunn reconstruction when compared with alternate, more “anatomic” or alternative reconstructions. The stabilizing advantages of additional coracoid reconstruction at the time of the Weaver-Dunn method have been shown.

Portions of the conjoined tendon have been used for experimental CC reconstruction. Sloan et al., in a cadaver model of single load to failure, demonstrated that the strength of the lateral half of the conjoined tendon (265 N) was at least as strong as that of the CAL (246 N) in a simulated reconstruction, although it lacked the strength of the intact CC ligaments (621 N). They cited potential advantages of this technique, including retention of the CAL integrity and significantly longer autologous graft material.

The tendon of the pectoralis minor tendon morphology and tensile strength has been investigated in cadavers and compared with that in the AC ligament. It can offer structural properties that render it as suitable as, or even more suitable than, the CAL. Although it has yet to be evaluated clinically, its use for this purpose circumvents the potential complications that may arise from the sacrifice of a normal and functionally important part of the coracoacromial arch.

The rationale for the use of biologic CC ligament substitutions has been investigated. The requirement for a biologic graft may not depend so much on the strength of the reconstruction, which can be comparably achieved with nonbiologic materials, but rather the necessity of the presence of a biologic substrate. In load-to-failure studies, anatomic allograft reconstructions have been shown to be superior to modified Weaver-Dunn, nonanatomic allograft, anatomic suture, and graft-rope techniques. An experimental anatomic reconstruction with semitendinosus tendon in a cadaver model failed to show significant elongation with cyclic loading in a manner similar to the intact CC ligament complex. In addition, the stiffness and ultimate load of the anatomic reconstruction decreased significantly by 40% and 25%, respectively, compared with the intact CC ligament complex. Comparing reconstructions using semitendinosus supplemented with either hook plate or PDS suture braid in paired cadavers, Dierckman et al. demonstrated superior biomechanical properties with the hook plate, although no difference to maximum load to failure. CC ligament reconstructions with tendon grafts demonstrate greater stability and less translation in vitro when compared with the Weaver-Dunn procedure in its original or modified forms. Accurate graft positioning and tensioning are required to reduce the incidence of failure. An in vitro investigation by Stübig et al. demonstrated the superiority of 3D C-arm flat detector navigation compared with free-hand arthroscopic technique aided by 2D fluoroscopy for the creation of transosseous tunnels based on anatomic landmarks. In a similar in vitro study, the advantages of electromagnetic navigation compared with standard minimally invasive reconstruction methods in terms of tunnel accuracy and time of preparation were observed. Constraint on the graft has been shown to be reduced experimentally if fixation takes place in the sitting position with the arm in 90 to 120 degrees of abduction. An in vitro CC reconstruction with semitendinosus looped beneath the coracoid and through anatomic clavicular drill holes comparing knot versus interference screw fixation showed no differences in the biomechanical properties except in the mode of failure: graft elongation versus slippage at the screw-tendon interface, respectively. A theoretical advantage is avoidance of additional stress risers in the coracoid process. It has been shown that a CC sling compared with a transcoracoid tunnel results in more pure anterior translation and AP translation of the clavicle under load testing. Comparing graft fixation techniques, Tashjian et al. found that square knotting had superior ultimate strength compared with interference screw fixation but the greatest elongation with cyclic loading versus side-to-side suture fixation and interference screw fixation. The side-to-side construct was the stiffest of the three methods of fixation. The superiority of biointerference screw fixation has been shown, and can be further enhanced with augmentation with nonabsorbable sutures. In cadaver specimens, Breslow et al. identified comparable fixation and stability achieved by a CC fixation method with sutures compared with one with suture anchors.

The concept of surgical recreation of separate CC ligaments has support, both theoretically and from experimental investigations. Different tendons have been used, and have shown no differences in the mechanical properties of in vitro reconstruction. Tunnel placement in the area of the clavicle corresponding to the attachment of the CC ligaments has the best bone density, and correlates with higher loads to failure experimentally. In a virtual shoulder simulated from normal shoulder MRIs, the kinematics of normal CC ligaments could not be restored by three common methods of CC reconstruction, leading the authors to conclude that it may be prudent to select graft fixation techniques most compatible with the type of instability being treated. In addition, the importance of simultaneous AC ligament reconstruction cannot be overlooked.

The direction being taken by investigators with interest in ACJ stabilization is toward the use of stronger materials or methods. There is growing evidence, mostly experimental, that supports a strategy of repair or reconstruction that includes both the AC and CC ligaments. Experimentally, a comparison of retention or excision of the distal clavicle in conjunction with reconstruction of the CC and AC ligaments revealed no significant difference in motion versus the intact state. It is not entirely clear at this time whether rigid or nonrigid fixation is better. The greatest attention has been focused on direct methods to achieve CC stability and, for the most part, indirect methods to achieve AC stability. Therein lies the challenge. It is intuitive that the preservation of the clavicle in its entirety would afford the best opportunity for preserving strength, motion, and function. However, even with maintenance of a congruent ACJ reduction, the necessity for excising the lateral clavicle might not be obviated. Biologically compatible materials as well as grafts are receiving more attention, as are methods considered less invasive.

Results

When compared with nonoperative treatment, operative treatment can result in less pain and better endur­ance, especially during work. Operative treatment resulted in improvement with regard to pain and patient satisfaction both in the short term and long term in several studies. However, from the meta-analysis of Smith et al., no differences were observed in strength, pain, throwing ability, or OA. Better cosmesis and greater duration of sick leave occurred with operative treatment. From their systematic review, Korsten et al. recognized that subjective and objective outcomes were better with operative treatment, although radiographic abnormalities were more common. While subjective and radiographic outcomes were better with operative treatment, objective parameters (UCLA, C-M, ASES, Acromioclavicular Joint Instability [ACJI] scores) were not significantly different. In high-performance overhead athletes, complete pain relief and return to normal was achieved more with operative treatment (92% vs. 80%). Return to professional football required nearly 60 days. A trend toward earlier return to Australian rules football with a more satisfactory outcome with operative treatment was reported by Cardone et al.

Direct AC and CC ligament repair via transosseous sutures that were protected with a temporary 4.5-mm CC screw for 8 weeks resulted in excellent clinical and radiographic outcomes at a mean follow-up of 72 months. Methods that temporarily transfix the ACJ with pins can produce some excellent and good results. A recent study by Verdano et al. reported a mean C-M score of 92.7, DASH score of 3.2, and SST score of 11.4. Reduction was maintained or there was minimal loss (less than 3 mm) in 85.7% of the patients. Increasingly stable if not improving long-term benefit up to 11 years has been shown. Fair and poor results have been reported to be 50% to 100% as common as excellent and good results. Intermediate outcomes, with 60% excellent and good outcomes, were observed. From what might be the report with the longest mean follow-up (24.2 years), by Lizaur et al., the clinical outcomes (Imatani, DASH, UCLA, SST, and VAS satisfaction) were both satisfactory and pain free in 92.1%. Radiographically, reduction was maintained in 86.8%, and the incidence of moderate to severe OA of the ACJ was 16%.

The PDS triple cerclage technique described by Sandmann et al. resulted in 94% excellent and good results, a mean C-M score of 94.3, ASES score of 94.6, and DASH score of 3.46. With respect to the CC distance, they found 85% less than 5 mm, 12% 5 to 10 mm, and one patient with more than 10 mm, none of which affected the clinical outcome. The AC and CC weave with nylon tape resulted in a mean Constant score of 84.9, ASES score of 81.7, and satisfaction in 88.2%, with 94% radiologic reduction. Occasional concomitant direct CC ligament repair did not impact the outcome.

Hook plates have been used extensively in Europe, resulting in reports from various centers as far back as 33 years. * In the earliest reports, the follow-up period was relatively short, with the results mixed and inconsistently reported. † Since then, the clinical and radiographic reliability of the method has been verified in multiple studies, and its use has become more widespread ( Fig. 9-68 ). Consistently high mean C-M scores (>92) have been reported. In 225 patients with a mean follow-up of 36 months, the mean Taft score in a report by Kienast et al. was very good and good in 84%. Additionally, in their study, both AC and CC intervals were appropriately reduced. Tavakoli et al. reported no redislocations. Minor joint asymmetry may exist at the completion of treatment. A poor correlation between clinical and radiographic results is not uncommon. On the contrary, Di Francesco et al. noted a clear correlation between radiographic findings and clinical results. They utilized postreconstruction MRI to assess ligamentous healing and the presence of scar tissue. Healing was present in 37 (88%) (stable) and not present in 5 (12%) (unstable). This correlated with greater than 50% joint subluxation and only fair clinical results. The clinical scores are lower in the infrequent presence of persistent anterior-posterior instability. Return to work and sport, often early, is the norm.

Treatment with hook plate. A, Preoperative anteroposterior (AP) image. B, Preoperative axillary lateral image. C, First postoperative image (overcorrection). D, Preoperative AP removal. Note osteolysis of the acromion at hook. The overreduction was reconciled. E, Final image. Note persistent osteolysis, favorable reduction, and remodeling of the screw holes.

* References .

† References .

Gstettner et al. performed a comparative study of hook plate versus conservative treatment, and achieved better clinical (Constant 90.4 vs. 80.7, SSV 89.2 vs. 77.6, excellent/good 87.5% vs. 58.8%) and radiographic (CC interval 12.1 mm vs. 15.9 mm, reduction in sports activity 16.4% vs. 23.5%) results in hook plate cases. Ejam et al. observed no significant differences in the outcomes of acute or late operations. Compared with the Phemister method, the Acroplate favored a more complete and rapid recovery without mobility complications.

More recent reports out of Asia have highlighted the ability of the hook plate to produce excellent clinical scores and radiographic parameters. Chen et al. reported no loss of reduction and improved clinical outcomes after plate removal. The mean C-M score was 89. Dou and Ren reported 100% excellent and good clinical outcomes.

With the MAAP, Liu et al. reported a mean Constant score of 94, excellent and good results in 94%, and anatomic reductions in 88%.

The outcome of the use of CC screws is gener­ally excellent and good in more than 77% of patients. This is the case even when ossification forms between the clavicle and coracoid or is encouraged to form. A relatively recent report by Esenyel et al. highlighted a mean Constant score of 98, excellent results in 86.7%, and normal alignment in 96.7%. Percutaneous screw insertion is more difficult, and results in higher rates of complications.

Bone-encircling and bone-engaging fixation with a multitude of flexible materials produce generally favorable clinical outcomes. If desired, return to preinjury activity levels can be expeted. Martetschläger et al. identified an 83% survivorship rate at 2 years using these techniques. Implants specifically designed for initial mechanical and supplemental biologic stabilization demonstrate a favorable immunologic response as well as a histologic response that suggest a functioning implant. In one of the longer term (mean, 70 months) studies using PDS cerclage, Greiner et al. reported favorable clinical outcomes (C-M 91.7, DASH 5, and SSV 92). CC interval increase of more than 5 mm was observed in 20% of the patients; OA was observed in 36%.

Favorable outcomes in the C-M, DASH, and Taft scores can be expected with transclavicular nonabsorbable sutures secured to the coracoid with suture anchors. A similar outcome was reported with the addition of supplemental suture anchor fixation across the ACJ. Somers and Van der Linden reported favorable clinical and radiologic outcomes with high satisfaction rates using a similar technique performed arthroscopically.

The clinical results of reconstructions using synthetic ligament (Ligastic) indicate a mean C-M score of 82 versus 90 on the healthy side. Despite this score, satisfaction for the postoperative result was high (96%), although some patients complained of residual pain (33%), scar concerns (7%), weakness (1%), work issues (1%), and sport issues (1%). A greater CC distance on the operative side did not adversely affect the outcome. Kany et al., using double braided polyester, reported excellent and good outcomes in 88% of their patients. The efficacy and safety of the LARS device has been shown. Motta et al. identified complete and incomplete ossifications in 96% of patients who recovered normal ACJ motion as demonstrated by 90 degrees of abduction and 0 degrees of external rotation against resistance and cross arm test. They showed full clinical recovery, and no correlations were recognized with radiographic or clinical outcomes.

Ye et al. reported a mean C-M score of 95.3 and markedly improved VAS score with transclavicular and subcoracoid passage of titanium cerclage multifilament cables with 82% anatomic reduction. They recommended removal 6 months after implantation to avert fatigue breakage.

Transosseous sutures spanning the CC interval may be secured through holes in metal buttons held intimately to the bone cortex by knotting the passed suture over the button. With such an open technique, an early (mean, 12 to 18 months) clinical success rate of nearly 90% together with radiographic stability restoration was achieved. At a mean follow-up of 24 months, De Carli et al. noted very high outcome scores (mean C-M, 98.2; ASES, 100; ACJI, 87.9), with the only significant difference from nonoperative treatment being for the ACJI score. Subjective and radiographic parameters favor operative treatment. Yi and Kim recognized that when the surgeon controls variables of clavicle and coracoid tunnel alignment, oriented as perpendicularly as possible, both clinical and radiographic improvement are favored. Shin and Kim reported a mean C-M score of 97.5 at 25.6 months. Loss of reduction did not influence the C-M score significantly. With a slightly longer follow-up (mean, 39 months), Rosslenbroich et al. showed a mean C-M of 94.7 and a mean Taft score of 10.8. An influence of age on outcomes was observed, with younger patients achieving higher outcome values.

Grassbaugh et al. showed an advantage to suture button techniques compared with other techniques, some historical, for the rate of revision and radiographic outcomes. However, in their review of their experience with eight different methods of treatment, they could show no statistically significant differences in clinical function or pain regardless of the treatment method. They observed poorer results for the treatment of chronic injuries and in the presence of DCR.

Some series of arthroscopic suture button techniques show a higher rate of successful clinical versus radiographic outcomes at early-to-intermediate term follow-up. Original buttons with a diameter of 6.5 mm migrated more often than a wider button (10 mm). No significant CC interval differences were observed by Cohen et al. who also obtained a C-M score of 91 despite piano key deformity and superior-inferior mobility in 40%. Pan et al. reported excellent functional recovery, no or minimal pain, and no radiographic failures at a mean of 2 years. The influence of tunnel orientation on the results is inconclusive. Glanzmann et al. emphasized the ability to eliminate pathologic posterior translation radiographically. Cook et al. obtained as many excellent and good results as fair and poor results. If failures were excluded, the mean C-M score reported by Spoliti et al. improved from 89.7 to 92.4.

Open techniques deemed “anatomic” reveal definite improvement in both pain and function, with an ASES score of 92 and Constant score of 94.7. Three failures yielded a success rate of 82%. Similar outcomes with respect to ASES scores as well as favorable SANE scores (89.1), QuickDASH (5.6), and excellent satisfaction were reported by Millett et al. Radiographic reduction was maintained in 82% of the patients. Arthroscopic-assisted techniques with interference screw autograft fixation in bone tunnels in both the clavicle and coracoid yielded a C-M score of 96.6, minimal (<2 mm) to mild (2 to 4 mm) or less displacement, and excellent functional and subjective outcomes.

Studies that have compared the results of AC and CC fixation report conflicting outcomes. Favorable results were reported by Weaver and Dunn and subsequently reiterated by other authors who performed the operation as originally described. Modified Weaver-Dunn operations also produced favorable outcomes. With a modification of the Cadenet method, Cerciello et al. reported mean C-M score of 94.3, VAS pain score of 0.91, 92% satisfaction 88%, complete reduction, and no degenerative changes. Sood et al., from their systematic review of the published literature available at the time, concluded that there was a low level of evidence to support the use of CAL transfer. In addition, supplemental fixation did not improve the results secondary to complications of its usage. A high rate of recurrent deformity was expected. Acute and chronic cases fared equally. The outcomes with retention or excision of the distal clavicle did not differ unless the retained clavicle manifested OA, in which case the results were inferior.

Comparing the Vukov and Phemister methods, Radovanovic et al. found similar outcomes but a shorter recovery time with the technique of Vukov.

Mixed results have been reported for the treatment of type III injuries with transposition of the tip of the coracoid process, which includes the short head of the biceps and the coracobrachialis to the clavicle. Skjeldal et al. advised that the procedure not be performed in patients with acute type III injuries. The proximally based conjoined tendon with CC sutures method of Jiang et al. resulted in superior clinical scores (ASES 91.4, C-M 90.6, SST 10.9), with overall excellent or good results in 89% of the patients. The majority returned to the same work and sports.

The early results of using autologous tendon indicate both efficacy and safety when grafts such as gracilis are utilized. A comparison study analyzing the results of reconstruction with Ethibond sutures through tunnels and semitendinosus autograft demonstrated no clinical or radiographic differences. Material costs related to certain newer implants may not completely offset any realized savings gain accompanying their usage.

Several studies report the lack of correlation between clinical outcome and abnormal radiologic findings. However, in the occasional instances of AP instability, inferior outcome scores are to be expected.

Complications

The systematic review of Korsten et al. determined that complications were more prevalent with operative treatment. Isolated ACJ fixation has a higher complication rate than CC fixation.

The ability of the stabilization method to maintain the operative reduction is considered the hallmark of the worthiness of the procedure. The incidence of recurrent deformity has been reported to be as high as 80%. The definition of loss of reduction with “anatomic” reconstruction varies from author to author. Anything other than anatomic reduction (referencing the contralateral side) varies from 6% to 100% ( Fig. 9-69 ). It is generally believed that after treatment, the extent of congruity or incongruity as well as the presence of OA and the development of heterotopic ossification appear to have no correlation with the outcome, regardless of the method.

Early loss of reduction of acute type III dislocation treated with synthetic sutures tied over cortical buttons.

A high failure rate is expected for soft tissue reconstructions if either AC or CC fixation techniques are not incorporated as additional protection from deforming forces. However, a very concerning complication of ACJ repair or reconstruction is hardware failure. Pins and wires used to stabilize the ACJ can break or loosen and subsequently migrate, as reported by Lyons and Rockwood and others ( Fig. 9-70 ). Serious injury and death can result from pin migration. The materials used for CC loop techniques and foreign body reactions to these materials can result in failures, the majority of which result from the erosion or fracture of either the clavicle or the coracoid process ( Fig. 9-71 ). Screws, wires, cables, or nonmetal synthetics can suffer fatigue breakage, leading to recurrence of the deformity ( Fig. 9-72 ). Screws can lose their purchase in the coracoid and pull through ( Fig. 9-73 ).

Steinmann pin fixation was chosen for an acromioclavicular joint repair. After breaking, it migrated to the right lung.

(From Rockwood CA, Green DP,eds. Fractures . 2nd ed. Philadelphia: JB Lippincott; 1984.)

Coracoclavicular loop fixation with vascular graft. A, In situ vascular graft that has eroded into the clavicle. B, Bulky vascular graft removed because of painful soft tissue irritation.

(Illustration courtesy Steven Lippitt, MD.)

Failure of nonabsorbable sutures used for a type III injury, with recurrence of deformity and reactive changes within the clavicle.

Loss of fixation of the coracoclavicular screw used for a type III injury, with recurrence of deformity.

The propensity for complications related to the engaging hook was realized in the static cadaver investigation by ElMaraghy et al. Hook positioning was unpredictable, especially because of the widely variable acromial and sometimes distal clavicular morphology. In their study, the incidence of subacromial bursa violation was 87%, supraspinatus muscle belly contact was 60%, and the focal contact of the tip of the hook with the undersurface of the acromion was 60%. These phenomena were validated and others recognized including subacromial impingement and RC lesions (partial thickness) in the study by Lin et al., who performed dynamic sonography of the subacromial space and environs after hook plate procedures. The engaging hook of the hook plate can lead to acromial osteolysis or erosions with varying frequency and extent, sometimes progressing to a fracture ( Fig. 9-74 ). No acromial osteolysis secondary to the hook was reported by Metzlaff et al., who removed plates between 10 and 13 weeks from the time of implantation. While the duration of implant retention may yield excellent stability, the price paid in terms of acromial osteolysis and ACJ degeneration has been recognized. It can also lose its purchase beneath the acromion process ( Fig. 9-75 ). Loss of reduction can spontaneously occur after removal of the implant, but is infrequent. Reports of hardware failure are not unique to the hook plate used by Kienast et al. Other complications related to the hook plate include medial clavicle fracture and screw failure ( Fig. 9-76 ). Hypertrophic incisional scars were observed in as many as 63% of patients. Additional complications reported include progressive degeneration of the ACJ, frozen shoulder, hematoma, and wound-healing problems.

A and B, Two examples reveal a large zone of osteolysis around the hook of the plate indicative of substantial motion unwisely performed in the presence of the hook plate. C, Fracture of the acromion process as a result of osteolysis.

Hook plate no longer secured beneath the acromion process.

Fracture of clavicle propagating from the medial screw.

Loss of reduction by suture anchor malpositioning was reported by Choi et al. Other notable factors contributing to loss of reduction include obesity, heavy, muscular limbs, and noncompliance with postoperative protective measures.

Contemporary techniques that promote “anatomic” reconstruction, including bone tunneling performed at the time of transosseous CC reconstruction, are solely responsible for fractures through the clavicle and coracoid, although coracoid looping techniques are not without risk. With the recent explosion of these techniques, the frequency of iatrogenic fracture is increasing ( Fig. 9-77 ). In a series of 27 patients, complications were more common in coracoid tunneling (80%) versus coracoid looping (35%) methods. To avoid iatrogenic clavicle fractures from tunnels, several authors adhere to techniques whereby biologic graft encircles both the clavicle and the coracoid. A coracoid fracture with subcoracoid passage of tendon graft was reported by Tomlinson et al. Millett et al. reported a revision and complication in 8 of 31 cases (25.8%).

A to C, Iatrogenic fracture of the coracoid secondary to malposition of the tunnel created to place the cortical button.

Many of these techniques employ flipping buttons, sometimes placed arthroscopically, that may migrate with loss of surgical correction in as many as 92% of cases. * Most often, this phenomenon is related to tunnel malpositioning, expansion of the drilled tunnel, and sometimes fracture through the respective bone ( Fig. 9-78 ). As an exception, Motta et al. identified #5 FiberWire failure in four patients with ligamentous hyperlaxity. Cook et al. showed that medialized tunnels were a significant predictor of early loss of reduction. Carofino and Mazzocca advised minimization of tunnel diameter spaced no closer than 20 to 25 mm with the most lateral tunnel 10 to 15 mm from the lateral edge of the clavicle. This phenomenon was diminished if the conoid tunnel was placed at 25% of the clavicle length from the lateral edge of the clavicle. The incidence of posterior instability, static or dynamic, may be as high as 53%. Instances when clavicular and coracoid buttons were not aligned perpendicular to each other suffered greater loss of reduction. In this study, the authors also observed that younger patients achieved superior clinical and radiographic outcomes. Patients with complications did not fare any different than those without complications, but they were less satisfied with their operation and did not experience good or excellent results. Rosslenbroich et al. identified a 22% loss of reduction (subluxation more than 50%). Shin and Kim reported a complication rate of 44%, including clavicle fracture propagating from a drill hole, fixation failures, clavicular erosion, and ACJ arthrosis. Despite a variable degree of loss of reduction (more than 50% difference in 33%) and complications in 44%, the mean C-M scores in patients with complications and reduction loss were 95.2 and 95.6, respectively. Woodmass et al. performed a systematic review of complications of arthroscopic treatment of AC dislocations. Soft tissue irritability over the clavicular hardware was the most common complication in acute cases.

Tunnel malposition leading to early loss of reduction of an acute type III dislocation treated with synthetic sutures tied over cortical buttons. A and B, Preoperative images. C and D, Early postoperative images. E and F, Images taken 5 weeks later.

* References .

In a long-term follow-up using PDS cerclage, complete redislocation was observed in 2 of 50 patients (5%). The rate was 9% with the triple cerclage technique in a shorter study by Sandmann et al. One redislocation was observed with nylon suture weave.

Synthetic ligament complication may be unique to the device. Use of Ligastic-type reconstruction was abandoned as a result of progressively increasing osteolysis appearing at the site of ligament passage through the clavicle in 22% of the patients. A 14% incidence of OA or osteolysis was noted as well. Kany et al. observed a 21% loss or reduction and a 12% incidence of night pain and capsulitis. One sleeve migration occurred. One instance of clavicular tunnel enlargement was observed in each of two reports.

Infection, usually superficial, at the site of ACJ reconstruction has been reported. * Staged reconstruction may be possible after the removal of all foreign material, rendering the operative site sterile. An aseptic response to braided polyester synthetic ligament, Nottingham Surgilig (Surgicraft), resulting in osteolysis of the distal clavicle was reported by Dearden et al.

* References .

Technical overcorrection of the CC interval, while enticing to preempt anticipated superior subluxation after repair, may result in compression of the brachial plexus. Entrapment of the brachial plexus with devastating consequences resulting from subcoracoid looping techniques has been documented.

Inadvertent injury to subclavian neurovascular structures during clavicular drilling associated with placement of a hook plate is a potential complication that can be averted with a keen knowledge of recognized safe drilling angle and depths.

Posttraumatic arthritis has been shown to be more common after operative than after nonoperative treatment. Posttraumatic arthritis can result and is more likely in ACJs that remain incongruent at the completion of their treatment.

Calvo et al. have shown that the incidence of CC ossification is higher after operative treatment. Millbourn reported that the incidence and severity were not different with operative treatment. In the report by Greiner et al., one instance of revision operation was necessary secondary to prolific CC calcification. At 10 years of follow-up, posttraumatic arthritis was reported to be more common when AC fixation, as opposed to CC fixation, was used.

Scapular dyskinesis and scapular malposition, inferior medial border prominence, coracoid pain and malposition, and dyskinesis of scapular movement (SICK) syndrome have been reported in 11.7% and 2.9% of patients with type III injuries.

Salvage of Failed Coracoclavicular Stabilization.

Depending on the index operation, any number of salvage options may be available as reported by Geaney et al. Confounding factors include loss of the integrity of the clavicle or coracoid due to erosion or fracture, clavicle foreshortening, and tissue exhaustion (CAL, conjoined tendon, deltotrapezius fascia). The optimal conditions for revision exist when the clavicle and the coracoid remain intact, and specific points for materials anchorage are preserved. CC looping techniques or their variations, with supplemental but temporary rigid fixation, would appear to offer the best options. For these extreme cases, there may be advantages to using biologic as opposed to synthetic reconstruction techniques. The first report of the use of autologous tendon between the clavicle and the coracoid was used for this indication. The continued use of autologous semitendinosus tendon is supported by its strength properties and results in a few cases. Deveci et al. incorporated the use of autogenous anterior tibialis tendon through transosseous tunnels with a hook plate to offload the bone tendon interfaces during the healing period and optimize integration of the tendon in the tunnel. They reported lower C-M scores (mean, 77.7) than for index procedures but no loss of reduction. The use of a braided polyester ligament in this setting has been described.

Salvaging iatrogenic fracture of the coracoid process may prove difficult. It usually thwarts any opportunity to achieve stabilization for intraosseous fixation between the clavicle and scapula unless substantial, high-quality bone remains available at the intersection of the base of the coracoid process with the body of the scapula and the glenoid neck. The best option may be the use of a hook plate and separate internal fixation of the clavicle fracture if feasible. Transposition of some or all of the origin of the conjoint tendon with or without bone is another consideration.

Recurrent deformity resulting from iatrogenic fracture of the clavicle may be reconciled as simply open reduction internal fixation of the fracture. When reconstruction is not deemed possible as a result of deficits of bone and soft tissue, total claviculectomy, reported for other indications, could be considered as the terminal, definitive reconstruction.

Author’s Preferred Method

I originally used Dacron vascular CC loops for type III injuries. When the technique fell out of favor due to erosion, nonabsorbable sutures ranging in size and number were passed in various combinations around and through the clavicle. When this method resulted in frequent failures, a more rigid form of fixation was found with the Rockwood screw. Even with the best visualization for optimal screw placement into the coracoid process, failures continued. A combination of Rockwood screw and modern nonabsorbable suture (FiberWire) did not solve the problem of failures with recurrent deformity. Since 2004, I have been using a hook plate in conjunction with nonabsorbable CC sutures ( Fig. 9-79 ).

The hook plate used by the author (Synthes). A, Early version of the plate. B, Current version of the plate.

With the patient in the beach chair position, a skin incision is made in Langer’s lines over the clavicle in a line just lateral to the tip of the coracoid process ( Figs. 9-80 and 9-81 ). Generous subcutaneous flaps are raised and the deltotrapezius fascia is incised over the lateral clavicle, extending over its end, where the supporting structures are disrupted and just onto the acromion process ( Figs. 9-82 and 9-83 ). The ACJ is inspected and debrided as needed, preserving the entire length of the clavicle ( Fig. 9-84 ). Sufficient subperiosteal exposure of the clavicle is performed to permit application of the shortest plate. The surgical approach at this location does not jeopardize the integrity of the major neurovascular structures, which are more than 4.5 cm from the clavicle.

Acromioclavicular joint deformity apparent at the time of the operation.

Skin incision in the relaxed skin tension lines.

Superficial dissection to expose the deltoid, trapezius, clavicle, and acromioclavicular joint.

Deep dissection between the deltoid and the trapezius, with entry into the injured acromioclavicular joint.

Exposure of the lateral clavicle. Note the disruption of the intra-articular disk.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Tumors and Related Conditions Private: Shoulder conditions in children Fractures of the proximal humerus Anterior instability: Arthroscopic surgery and outcomes Multidirectional instability—open inferior capsular shift Evaluation of rotator cuff pathology: History, examination, and imaging

Stay updated, free articles. Join our Telegram channel

Join