Posterior Humeral Head Allograft and Open Bankart Repair Through an Anterior Approach

Abstract

Background:

Shoulder instability events can often result in humeral head and glenoid bone defects. Lesion size, patient age, bone quality, and cause of instability affect management. Surgical options are numerous, depending on severity and complexity. In addressing posterior humeral head lesions, remplissage and humeral head allograft have been reliably described, but the approach to addressing these often significant lesions has been variably illustrated. As recently described by Yazdi et al in a systematic review in 2022, osteochondral allografts for Hill-Sachs or reverse Hill-Sachs lesions showed good patient-reported outcomes. This is in agreement with other studies in the literature, including another systematic review by Saltzman et al in 2015 that reported good outcomes after humeral head allografts for humeral head defects, as well as another study by Gerber et al that reported similar promising outcomes.

Indications:

Humeral head allograft should be considered in the setting of instability refractory to nonoperative measures in younger patients with large Hill-Sachs and reverse Hill-Sachs lesions, particularly in those that are engaging with the glenoid through range of motion and are over 30% of the depth of the humeral head.

Technique Description:

Following an examination under anesthesia and diagnostic arthroscopy, a deltopectoral incision was made from the coracoid to the deltoid insertion. The subscapularis tendon and anterior capsule were both carefully released from their humeral insertion and tagged. Following external rotation of ~180°, the Hill-Sachs defect was visualized, debrided, and molded with bone wax. After an osteochondral humeral head allograft was sized and sculpted on the back table, it was positioned and fixated with provisional Kirscher wires followed by 4-0 cannulated, headless compression screws. Finally, an open Bankart repair was completed, followed by a capsular closure and subscapularis repair.

Results:

Humeral head allografts have demonstrated short-term improvements in motion and patient-reported outcome measures and can be used for posterior Hill-Sachs lesions, fully accessible through an anterior approach when anterior instability procedures are also warranted.

Discussion/Conclusion:

Management of large Hill-Sachs and reverse Hill-Sachs lesions with a humeral head allograft using an anterior open approach is a viable option for patients with refractory instability.

Patient Consent Disclosure Statement:

The author(s) attests that consent has been obtained from any patient(s) appearing in this publication. If the individual may be identifiable, the author(s) has included a statement of release or other written form of approval from the patient(s) with this submission for publication.

Graphical Abstract.

This is a visual representation of the abstract.

Download video file ^{(553.8MB, mp4)}

DOI: 10.1177/26350254251320956.M1

Background

We present a unique case of allograft bone grafting of a Hill-Sachs lesion with an open Bankart repair through an open anterior deltopectoral approach. The authors have no conflicts of interest or disclosures to report. Furthermore, no commercial funding was used for the creation of this video.

Indications

Our patient is a 22-year-old right hand–dominant man who had chronic left shoulder instability with multiple dislocations and seizure disorder. His past medical history is also significant for autism spectrum disorder. At the time of surgery, his seizures had been controlled with levetiracetam for almost 6 months.

His physical examination was notable for full range of motion (ROM) and strength but with positive apprehension and relocation testing.

On the patient's preoperative radiographs, you can begin to appreciate the significant Hill-Sachs lesion. Various axial cuts from the computed tomography (CT) scan further identify this Hill-Sachs defect in the patient's posterior humeral head. On the axial CT scan, you can visualize the humeral head defect and note that its width and depth reach at least 50% of the width of the humeral head. Images show a 3-dimensional reconstruction of this defect using his CT scan.

We also obtained magnetic resonance imaging (MRI) to assess the soft tissue changes in addition to the CT shown previously that was obtained to better assess glenohumeral bone loss. On the axial MRI, you again see the Hill-Sachs lesion, in addition to the soft tissue changes about his labrum and glenoid. We can see this anterior inferior glenoid lesion, also known as a Bankart lesion. On the sagittal MRI, we are able to visualize the glenoid face and labral injury, as well as the anterior and inferior glenoid.

We then used cuts of the sagittal and axial CT to calculate the glenoid track and glenoid bone loss. The glenoid track has been defined as 83% of the width of the glenoid, minus width of the glenoid lesion, and this was calculated to be about 18.95. We then used the axial CT to calculate the Hill-Sachs interval, which is measured as the distance from the medial aspect of the rotator cuff insertion to the medial portion of the humeral head defect. It has been reported that if the Hill-Sachs interval is greater than the value of the glenoid track, it is an engaging or off-track lesion, which, in this case, it was. We also used our sagittal CT to calculate the glenoid bone loss based on surface area of the defect over surface area of the glenoid, and in this case, the percent bone loss was 8%. There have been recent reports reidentifying subcritical bone loss as 10% to 13% in addition to critical bone loss as anywhere from 13.5% to 25% bone loss.^7,8 As this case fell in the range of subcritical bone loss from 10% to 13%, it was deemed that a bony augment procedure to the glenoid may not be necessary given that this was the patient's primary index procedure and the patient has recurrent seizures, putting any procedure at risk for failure. In the case of significant glenoid bone loss in the setting of seizure disorder, it is the senior author's (J.F.D.) preferred treatment to use a free bone block fixed with noncanulated screws. Recurrent seizures are not uncommon, and fracture or the loss of fixation in the setting of a primary Latarjet presents a particularly challenging revision scenario due to the altered anatomy, proximity to the axillary artery and nerve, and the unprotected brachial plexus.^1,2,4,5 In this case with mild glenoid bone loss and a large Hill-Sachs lesion, an open Bankart repair was chosen in addition to augmenting the posterior humeral head lesion with an osteochondral allograft (OCA) humeral head.^1,2

Our patient's examination under anesthesia preoperatively is shown. As you can see, he has a 3+ load and shift test anteriorly and 1+ posterior load and shift, in that his humeral head is translatable over the glenoid rim, completely dislocating and remaining in the dislocated position.

Technique Description

We first performed an arthroscopy of the patient's glenohumeral joint to better visualize the labrum, biceps anchor, and any bony or cartilage changes. As noted in these select photos, in this portion of the humeral head and glenoid, the cartilage is intact, with mild changes along the anterior inferior aspect of the glenoid rim, with an anterior labral periosteal sleeve avulsion (ALPSA) lesion of the anterior inferior labrum anteriorly and inferiorly.

We then took the shoulder through external and internal rotation (IR) to visualize the humeral head lesion and axillary pouch, as seen here. The video of the glenohumeral joint notes the complete loss of the labrum anteriorly and inferiorly. This was an ALPSA with no associated loose body or bone fragment along the anterior-inferior glenoid neck. We then visualized the Hill-Sachs lesion as we came superiorly and posteriorly, and we noted the posterior and superior labrum were intact.

Here we are taking the shoulder into abduction and external rotation (ER), showing the Hill-Sachs lesion is engaging with the anterior glenoid.

Our attention then turned to the open portion of the case. Our patient was positioned in the beach-chair position with a folded towel under the scapula to allow for scapular stabilization. Our incision was for a deltopectoral approach rather than the standard axillary incision for an open Bankart repair to allow for greater access to the humeral head. Our incision extended from the coracoid, distally toward the deltoid insertion. We then proceeded to use the deltopectoral interval to reach the glenohumeral joint. Humeral head exposure requires a subscapularis tenotomy, which should be meticulously performed. It is imperative to maintain sufficient tendon for repair on the subscapularis tenotomy. The biceps can be preserved if desired, as was done in this case.

Here is a clip showing the interval between the subscapularis tendon and the anterior joint capsule being developed, and here we show the excellent mobility achieved between the subscapularis and capsule. We then show access to the glenohumeral joint achieved but using traction stiches in both the subscapularis and capsule.

A battery-powered or pneumatic arm holder is preferred for maintaining ER during humeral head exposure. We were able to fully externally rotate the arm to allow the humeral head to come through the anterior interval and rotate for complete visualization of the posterior humeral defect. For the Hill-Sachs lesion, we used a Kobel medially and laterally with the shoulder dislocated, in maximum ER, and a small pointed Homan retractor was used along the posterior humeral head to protect the rotator cuff.

We show full access to the posterior lesion, achievable through this approach.

We then cleared the defect of soft tissue and debrided the bone to a bleeding bed and placed bone wax to use as a mold for sizing the OCA.

The OCA was measured, cut, and sculpted on the back table. We used a humeral head allograft, but the use of other grafts has been reported in the literature, including iliac crest, femoral head, and talar bone grafts.^3,6

We also further prepared the bony bed before placement of the OCA. We then placed the OCA wedge in the defect and secured the graft into place with two 4.0 cannulated headless compression screws over guidewires with excellent compression achieved. Headless compression screws were used to avoid any potential pitfalls with hardware.

Our attention then turned to the open Bankart repair portion of the case. With humeral head retraction with a Fakuda retractor and Link, Homan, and Bankart retractors anteriorly and inferiorly, we were able to access the anterior glenoid with excellent exposure. We then placed our 3 FiberTak (Arthrex) anchors at the 5:30, 4:30, and 3:30 positions on the glenoid face for our Bankart repair and an additional 2 anchors in the humeral head for our future repair of the subscapularis. It should be noted that if glenoid bone augmentation is performed, use of a free bone block and hard body (noncanulated) screws is recommended. ⁴

After passing our suture tapes from the loaded anchors through the anterior labral and capsular tissues, we tied down our repair outside the capsule with a knot pusher, which achieved the desired glenohumeral stability.

We then closed the capsule and repaired the subscapularis tenotomy using the sutures from the previously placed anchors into the humeral head. We ensured there was no overtensioning of the tendon with the arm returned to its appropriate position. The shoulder was well centered and reduced within the glenohumeral joint once the reconstruction was complete, and there was no inferior translation. The wound was then closed in a standard layered fashion.

Results

Postoperative radiographs, notable for the well-positioned OCA secured with the 2 headless compression screws, are shown. Also, the patient's radiographs at his first postoperative visit again show a well-located humeral head with restored sphericity with the allograft and maintained hardware position.

No additional advanced imaging was performed postoperatively, as there was no concern for failure or new trauma, and the patient has difficulty with longer examination times due to his underlying disability.

Discussion/Conclusion

The patient did well immediately postoperatively, was placed in a sling for 8 weeks, and had no pain, further incidents, or seizures despite having to stop his seizure medication, levetiracetam, due to agitation. He is now 7 months postoperative and doing well, with full ROM, intact strength, continuation of physical therapy, and no other issues. He was released on August 1, 2024, to follow-up as needed.

Here are our references, and our full postoperative protocol is delineated in the transcript of this video.

Postoperative Protocol

Phase I (0-6 weeks: no active ROM (AROM) or passive ROM (PROM; pulleys).

At the 1- to 2-week mark, we begin with active-assisted ROM and activity of elbow, wrist, and hand. Pendulums begin within a sling and progress to full in a graduated fashion. Cardiovascular activity is limited to a stationary bike with the sling in place. Submaximal shoulder flexion, extension, and abduction isometrics can be initiated. Scapular retraction exercises are done without resistance.

At the 2- to 4-week mark, IR and ER submaximal isometrics can be started as well as resisted exercises from the elbow distal. Scapular stabilization can continue with further ball-on-table exercises. Cardiovascular training protocol is unchanged.

At the 4- to 6-week mark, cross-arm stretching and sleeper stretches begin for aggressive posterior capsular stretching, with progression to full AROM against gravity only. Additionally, glenohumeral joint mobilizations (posterior/inferior) can begin as well as rhythmic stabilization. Cardiovascular training progresses to a walking progression program (treadmill, bike, elliptical, Stairmaster), and lower extremity strength training can begin.

Scapular retraction can begin prone without resistance, and stabilization transitions from ball-on-table to ball-on-wall.

Phase II (6-12 weeks postoperatively; no pushups, heavy lifting, sports, repetitive overhead use of shoulder; overhead ER starts at week 8, start to wean from sling/immobilizer)

At the 6- to 9-week mark, continue AROM and AAROM with the addition of PROM (goal of full PROM in all planes by 9 weeks). Strength training includes ER and IR with the arm at the side, forward flexion (FF) and scaption to 60° to 90°, and prone rows; transition from sitting to standing. From a scapular stabilization perspective, retraction begins against resistance. Pushups plus triceps extensions, lat pull-downs, and bicep curls begin. Can transition to beginning-level pool program with nothing overhead or upper extremity resistive exercises.

At the 9- to 12-week mark, continue to progress PROM. Low-level upper extremity dynamic stretching program begins. Shoulder strengthening with IR/ER at 30° scaption and stabilization exercises progress. Gentle overhead strengthening begins. Sport-specific training begins at week 10.

Phase III (4-6 months postoperatively; no contact/collision sports/military schools until 9 months postoperatively; same for heavy overhead lifting and 90/90 position)

At the 4- to 6-month mark, strengthening continues with ER/IR at 45° to 90° elevation and FF/scaption to 90° to 120°. Progressive sports training begins at month 5 at 25% to 50% intensity.

At the 6+-month mark, high-level upper extremity dynamic stretching progresses along with strengthening and sport-specific drills. Once patients are able to sustain a tall plank position without increased pain before completing test, they can complete closed kinetic chain upper extremity testing.