Recurrent acromioclavicular joint dislocation with an associated coracoid fracture following acromioclavicular joint reconstruction

Abstract

Acromioclavicular joint injuries are common and account for up to 12% of all shoulder girdle injuries and have been reported as being as high as 50% in sports-related shoulder injuries. While the majority of acromioclavicular joint dislocations can be treated non-operatively, there are certain injury configurations, which can include high-grade dislocations in overhead athletes, where surgery may be indicated.

The surgical management of acromioclavicular joint instability has moved towards recreating the action of the coracoclavicular ligaments by resuspending the clavicle on to the coracoid. Multiple techniques using high-strength sutures, synthetic ligaments, tendon allografts or autografts passed either around or through the coracoid process have been described. However, an unusual, but significant, complication associated with these techniques is an iatrogenic fracture of the coracoid process.

We report the case of a patient with an iatrogenic coracoid fracture following two failed acromioclavicular joint resuspensory reconstructions using a synthetic ligament. This injury was successfully treated with an autologous hamstring graft reconstruction, initially protected with a hook plate.

Background

Acromioclavicular joint injuries are common and account for up to 12% of all shoulder girdle injuries and have been reported as being as high as 50% in sports-related shoulder injuries. While the majority of acromioclavicular joint dislocations can be treated non-operatively, there are certain injury configurations, which can include high-grade dislocations in overhead athletes, where surgery may be indicated.

The surgical management of acromioclavicular joint instability has moved towards recreating the action of the coracoclavicular ligaments by resuspending the coracoid from the clavicle. Multiple techniques using high-strength sutures, synthetic ligaments, tendon allografts or autografts passed either around or through the coracoid process have been described.¹ However, an unusual, but significant, complication associated with these techniques is an iatrogenic fracture of the coracoid process.

Case history

A 29-year-old right-handed man, who had previously served in the armed forces, was referred to our unit after two failed previous acromioclavicular joint reconstructions using a Surgilig (Surgicraft Ltd, Redditch, UK) synthetic ligament. He had sustained a type 5 acromioclavicular joint dislocation following a direct impact to his shoulder while snowboarding 15 months earlier. He was treated at his local hospital, where he underwent a suspensory acromioclavicular joint reconstruction using a 12cm Surgilig synthetic graft. The graft appeared to have ‘stretched out’ postoperatively, and the procedure was revised four months later using an 11cm Surgilig graft. Unfortunately, some time after the second procedure, the patient sustained a coracoid fracture with an associated failure of the reconstruction and superior displacement of the lateral end of the clavicle. As a result of the injury, the patient retired from the army and changed profession to an information technology consultant.

He was referred to our tertiary unit, where his presenting complaint was of pain at the acromioclavicular joint with continuing instability, crepitus, a prominent lateral clavicle and restriction of glenohumeral abduction. On examination, his previous sabre incision scar was noted and the lateral end of his clavicle was displaced superiorly. This was tender on palpation and was unstable but could be easily reduced. Glenohumeral abduction was restricted to 130° by pain and he also had pain on cross-body adduction. He had no irritability on palpation around the coracoid and there did not appear to be any pain or functional deficit on loading the biceps or coracobrachialis.

None of the previous imaging was available. Plain radiographs taken in our unit demonstrated a superiorly displaced lateral clavicle with a screw and washer in place. There was an associated coracoid fracture which did not appear to be significantly displaced (Fig 1).

Figure 1.

Anteroposterior (A) and axillary (B) plain radiographs of the shoulder demonstrating a previous acromioclavicular joint reconstruction with a superiorly displaced lateral clavicle with a screw and washer in-situ. There is an associated coracoid fracture (white arrow).

Treatment options were discussed with the patient including non-operative measures and surgery. He had previously undergone an extensive course of conservative therapy, including physiotherapy, and had attempted to return back to swimming and other activities but was unable to do so because of pain and restriction of movement. Taking the patient’s age into account and his desired level of activity, a decision was made to proceed with a revision acromioclavicular joint reconstruction using a hamstring autograft, fixed between the base of the coracoid and the lateral end of the clavicle, which would initially be protected with a supplementary hook plate for six months.

Surgical procedure

Having discussed the surgical procedure with the patient and obtaining informed consent, the procedure was undertaken under general anaesthesia and an interscalene block. The patient was positioned in the beach chair, the shoulder and ipsilateral knee were prepared and draped in the standard fashion and prophylactic antibiotics administered.

Initially, the gracilis tendon was harvested form the ipsilateral knee using a longitudinal incision over the pes anserinus using a standard tendon-stripper and technique. The graft was folded in half. The doubled-over end was whip-stitched using a non-absorbable Number 2 FiberWire suture (Arthrex, Naples, FL) with the suture ends left long. The two free ends of the graft were initially left and would be cut to length and prepared once the required distance from the base of the coracoid to the clavicle was measured.

At the shoulder, the previous sabre incision was reopened. Sharp dissection was used to carefully raise full-thickness deltoid and trapezius flaps from the lateral end of the clavicle. The screw and washer were removed from the lateral end of the clavicle and the old Surgilig implant released. The lateral tip of the clavicle was then cut off square and a clavicular ‘socket’ was created by clearing out the medullary canal using a small Volkmann’s spoon to a depth of 20mm. The cortical bone at the inferior edge of the entrance to the socket was rounded off, to protect the graft when inserted. A 1.5mm drill bit was then used to drill single drill holes through the anterior and posterior cortex of the clavicle, 20mm from the lateral end, into the medial end of the medullary socket.

Dissection was then taken inferiorly following the Surgilig implant down to the base of the coracoid. There appeared to only be fibrous tissue within the loop of the Surgilig, where it had previously wrapped around the coracoid, and the construct was removed. The residual tip of the coracoid was identified and appeared to be firmly secured by fibrous tissue. This was left in place.

The base of the coracoid was then dissected out and cut square to expose the central medullary canal. Under image-intensifier guidance, a 2mm Beath pin was drilled down the centre of the canal aiming slightly medially and inferiorly. The Beath pin was drilled through the posterior cortex of the scapula, pushed through the overlying infraspinatus muscle belly and out through the skin (Fig 2). A 7mm cannulated drill bit was then passed over the Beath pin and the coracoid stump drilled to a depth of 30mm.

Figure 2.

Intraoperative image intensifier; the Beath pin has been inserted into the stump of the coracoid and passes medially and inferiorly out of the back of the scapula.

The whip-stitch suture ends of the graft were threaded through the eyelet of the Beath pin, which was then pulled by its pointed end out of the back of the shoulder. The suture ends were then retrieved as they passed out of the shoulder and pulled tautly. This pulled the looped end of the graft down to the base of the coracoid socket. A thin guidewire was then inserted into the socket, alongside the graft, and a cannulated 7 × 25mm blunt threaded titanium RCI interference screw (Smith & Nephew, Andover, MA) screwed into the socket while the graft was held in tension.

Having secured the graft into the coracoid, the clavicle was pushed down into its reduced position and, using a measuring tape, the distance from the base of the graft into the medial end of the clavicular socket was measured. The two free ends of the graft were then cut to just short of this length and were both whip-stitched, using a Number 2 Fiberwire suture. Using shuttle sutures passed through the anterior and posterior drill holes in the clavicle, the whip-stitch suture end of one of the graft limbs was shuttled into the clavicular socket and out of the anterior drill hole and the other out of the posterior hole. The posterior suture end was brought over the top of the clavicle and clipped to the anterior suture end, but the graft limbs were not pulled into the socket.

The hook of a left-sided 3.5mm seven-hole LCP™ clavicle hook plate (Synthes, West Chester, PA, USA) was then inserted under the acromion and the lateral end of the clavicle reduced as the plate was pushed down into position. The medial end of the plate was secured using three locking screws inserted into slots medial to the end of the clavicle socket (Fig 3).

Figure 3.

Intraoperative image intensifier; axillary view demonstrating the interference screw inserted into the base of the coracoid and the clavicle reduced and held in position by the hook plate.

Having stabilised the clavicle, while carefully directing the graft limbs into the clavicular socket, the anterior and posterior suture ends were pulled taut. Once we were happy with the tension of the graft construct, the suture ends were tied together over the anterior edge of the clavicle (Fig 4). The deltotrapezial flaps were then imbricated using a braided absorbable suture and the wounds closed in layers.

Figure 4.

Diagram of the completed construct; the inferior end of the graft is secured into the coracoid stump with an interference screw, the clavicle has been stabilised with a hook plate and the superior end of the graft has been pulled into the clavicular socket and is secured into position by whip-stitch sutures that have been tied over the front of the clavicle.

After the procedure, the patient was instructed to keep his arm in a sling for four weeks and to undertake gentle pendular exercises only while avoiding any weight-bearing activities. At the two-week review, the patient’s wounds were well healed. At the eight-week review, he had minimal pain and had regained movements with abduction at 120°. Radiographs demonstrated a well-reduced acromioclavicular joint reconstruction (Fig 5). As planned, the hook plate was removed six months later with no complications.

Figure 5.

Postoperative anteroposterior (A) and axillary (B) plain radiographs demonstrating the revision acromioclavicular joint reconstruction. The interference screw is located within the base of the coracoid and the hook plate is holding the lateral clavicle in a reduced position.

At the patient’s most recent follow-up, 12 months after removal of the hook plate, he was pain free, had a full range of motion and had an equal and symmetrical shoulder contour compared with his contralateral shoulder. His plain radiographs demonstrated the lateral end of the clavicle to have remained in a reduced position (Fig 6).

Figure 6.

Anteroposterior (A) and axillary (B) plain radiographs one year after removal of the hook plate. The interference screw can be seen passing down the base of the coracoid and the lateral clavicle is in a reduced position.

Discussion

Multiple techniques have been described to stabilise the acromioclavicular joint, with the most successful being aimed at reconstructing the coracoclavicular ligaments. The Weaver–Dunn procedure, with its numerous modifications, is based on transferring the coracoacromial ligament from the anterior edge of the acromion into the lateral clavicle.^1,2 Newer techniques aim to more accurately recreate the action of the coracoclavicular ligaments by resuspending the coracoid from the clavicle.^3,4 Techniques using synthetic ligaments, suture anchors, suture buttons and autograft and allograft tendons have been described.¹ They all involve either passing the graft or implant around or through the coracoid process.

Multiple complications, including infection, clavicle fracture, coracoid fracture and implant failure following acromioclavicular joint re-suspension procedures have been reported.^5,6 Milewski et al reported a 78% complication rate in their series of 27 patients undergoing a resuspensory reconstruction for high-grade acromioclavicular joint injuries.⁷ They also found a 20% incidence of coracoid fractures in their coracoid tunnel group.⁷

There have been limited other reports of coracoid fractures following resuspensory acromioclavicular joint reconstruction. Bindra et al reported a coracoid fracture with a loss of reduction in a patient eight weeks after they had undergone a TightRope (Arthrex, Naples, FL, USA) acromioclavicular joint reconstruction following an acute high-grade acromioclavicular joint injury.⁸ The patient was treated non-operatively and ultimately became pain free but was unable to return to full sporting activities. Gerhardt et al reported on their case of a coracoid fracture after initial fixation using a transclavicular–transcoracoid technique augmented with a GraftRope (Arthrex, Naples, FL, USA).⁹ At week one postoperatively, the graft construct had pulled through the coracoid causing recurrence of the initial displacement. This was revised by fixing the coracoid fracture with a cannulated screw and looping a semitendinous allograft around the coracoid and clavicle, which was tied to itself. They used a hook plate to protect their repair. The patient achieved an adequate range of motion with minimal pain at final follow-up. However, of note, the hook plate was not removed and they did not report whether the clavicle fracture had healed, so it is not possible to know whether the soft-tissue reconstruction had actually been successful.

The Surgilig is a double-looped synthetic braided polyester ligament, which loops around the coracoid and clavicle and has been used successfully for acromioclavicular joint reconstructions.^10–12 However, Sarda et al have described a case of a patient who developed pain and discomfort over the coracoid process one year after an acromioclavicular joint reconstruction using a Surgilig.¹³ Plain x-ray and computed tomography demonstrated osteolysis around the coracoid and they elected to remove the Surgilig. At the time of surgery, on removal of the implant, the acromioclavicular joint remained unstable and a subsequent Weaver–Dunn procedure was required.² The patient was pain free at follow-up. A coracoid fracture following a Surgilig acromioclavicular joint reconstruction has been reported by Jeon et al in a series of 11 patients.¹⁴ They felt that this was due to excessive loading of the construct by the patient, against their clinical advice, in the first three postoperative weeks. The patient underwent a subsequent Dewar–Barrington procedure, which involved transferring the tip of the coracoid and the conjoint tendon and attaching it to the anterior edge of the clavicle with a screw.¹⁵ The patient had a poor outcome, with elevation of 90° and external rotation of 30° at final follow-up, which the authors suggested might have been due to impingement of the coracoid on the supraspinatus tendon.

In our reported case, where the patient had undergone a revision Surgilig operation within three months of the primary procedure, none of the previous imaging was available. While a fracture of the coracoid, leading to upward displacement of the lateral end of the clavicle, was evident on the presenting x-rays, we do not know when this occurred. It is possible that the coracoid fracture may have actually occurred soon after the index procedure and may have, in fact, been the reason for the failure of this procedure, and so was already present at the time of the revision operation. Whether the fracture had occurred as a result of some form of trauma at the time of either of the surgical procedures, in the postoperative period or as a result of osteolysis associated with the Surgilig implant is not clear.

On initially assessing the patient, it was evident that his continuing symptoms were due to his acromioclavicular joint instability and that his coracoid fracture was asymptomatic. From his imaging we classified his coracoid fracture as a type II injury according to Ogawa et al.¹⁶ In their series of 67 coracoid fractures, they found that type II fractures were generally asymptomatic and recommended that they should be treated non-operatively.¹⁶ Taking this into account, and the patient’s continuing acromioclavicular joint symptoms, we decided to undertake a soft-tissue acromioclavicular joint reconstruction without attempting to fix or disturb the distal tip of the coracoid. We chose to use a hamstring graft attached from the base of the coracoid into the lateral end of the clavicle. As this was a second revision procedure, we also protected our soft-tissue reconstruction with a lateral clavicle hook plate. At our latest review, 18 months after the procedure and 12 months after removal of the hook plate, the patient had maintained a pain-free and fully functional recovery and the lateral clavicle remained in a reduced position on x-ray.

In conclusion, to our knowledge, the current case report is the first to describe a patient with a coracoid fracture following a resuspensory acromioclavicular joint reconstruction successfully treated with a hamstring tendon reconstruction. With an increasing trend towards surgical intervention for acromioclavicular joint injuries, it is important for surgeons to be aware of this potential complication and its management.