➢ The use of fresh osteochondral allografts has become popular in many joint-preserving orthopaedic procedures and shows early promising results within the shoulder.
➢ Distal tibial allograft contains a stout cartilaginous layer that appears to have highly congruent curvature and concavity to the glenoid, which makes for an optimal allograft option for instability.
➢ In the setting of large Hill-Sachs lesions, the use of a humeral-head osteochondral allograft is essential to restore geometry, stability, and mechanics of the native glenohumeral joint.
➢ One must be cautious with the treatment of glenoid chondral lesions with osteoarticular grafting procedures because of the depth of the glenoid compared with the depth of subchondral bone on the graft necessary to achieve a press fit, and advanced imaging is recommended when planning an operative intervention.
➢ Optimizing joint-preservation treatment with osteochondral allografts will rely on the long-term results of these procedures, and careful patient selection, preoperative discussion, and realistic expectations are necessary.
The use of osteochondral allografts to correct articular defects has been well established, with most attention centered on correction of osteochondral defects of the knee1-3. The lack of availability of tissue and the logistics of graft transplantation have limited the use of fresh osteochondral allografts in North America. Under guidelines from the American Association of Tissue Banks, osteochondral allograft tissue is held until microbiologic and serologic testing is complete4. Currently, this allows the patient <2 weeks to receive a size-matched graft to maintain adequate chondrocyte viability. Investigation is ongoing to develop techniques to prolong storage and to preserve chondrocyte viability5. Osteochondral allografts have been used for a variety of shoulder conditions and are most frequently employed for articular defects related to trauma, instability, and degenerative conditions.
In this review, we will discuss the use of osteochondral allografts for a variety of shoulder-specific conditions. We review both negative outcomes and positive clinical results reported in the literature for each treatment. Additionally, we discuss technical considerations and practical applications for frequently challenging pathology.
Glenoid Bone Loss
The stability of the shoulder is dependent on the complex interaction of several static and dynamic mechanisms that help to maintain a functional glenohumeral articulation. With repeated dislocation, deficiencies in the osseous architecture of the glenoid and humeral head may occur. Literature has suggested that the recurrence of shoulder instability is a potential complication after surgical stabilization6. Failure rates of both open and arthroscopic stabilization have been reported between 7% and 19%7, unless all of the injured stabilizers are adequately assessed and treated.
Recognizing glenoid bone loss as a potential cause for failure after an anterior instability surgical procedure has been emphasized in the literature8-13. Anatomically, glenoid bone loss decreases the available articular arc length and the depth of articular conformity to the humerus, causing glenohumeral mismatch and a loss of the joint’s normal concavity-compression9,14. Furthermore, bone loss leads to a smaller area by which the glenoid can resist axial forces, which allows for an increase in relative shear forces imparted to a repaired capsulolabral interface8. Alterations in biomechanical stability are noted when glenoid bone loss approaches 15% to 20%15. A cadaveric study has shown that a 30% glenoid defect increases the anteroinferior glenohumeral contact pressures by 300% to 400%16. Further evidence suggests that bone reconstruction of the glenoid is recommended if there is a deficiency of more than 20% to 25%10,17-19.
Operative treatment is ultimately guided by the extent of osseous injury to the glenoid and humerus, patient-specific factors, and the surgeon’s personal preference and level of experience. Several popular techniques have been published for osseous glenoid reconstruction, including variations of coracoid transfer17,19-23, iliac crest autograft13, glenoid allograft24,25, and distal tibial allograft26. Globally, the Latarjet procedure has gained substantial popularity. Several studies have cited favorable clinical results and few recurrent instability events. However, other studies have shown a short-term complication rate of up to 25%, consisting of infection, instability, and nerve injury, along with the occurrence of coracoid osteolysis that may impact long-term outcomes27,28. Furthermore, the transferred coracoid does not have articular cartilage, and the development of arthritis is a concern and is still being evaluated in biomechanical and retrospective studies17,29,30.
Along with the lack of articular cartilage, poor congruence with the radius of curvature of the glenoid has also been identified with autologous iliac crest bone grafts31. With an autologous iliac crest bone graft, the surgeon exposes the patient to the general risks and morbidity involved with graft harvesting32. Hypothetically, a nonanatomic glenoid reconstruction, poor resulting geometry of the glenoid arc, nonanatomic extra-articular repair of the capsulolabral tissues, and no reconstitution of the chondral surface may lead to earlier degenerative changes26.
With regard to glenoid osteochondral allografts, there are case reports of good short-term outcomes for both open and arthroscopic approaches24,25; however, to our knowledge, the approaches have not been compared in biomechanical studies.
The use of a fresh distal tibial osteochondral allograft for glenoid reconstruction has gained much attention over the past several years since the publication by Provencher et al.26. Widely available in tissue banks, the lateral aspect of the distal part of the tibia contains a stout cartilaginous layer that appears to have highly congruent curvature and concavity to the glenoid. In an initial biomechanical pilot study, randomly selected, non-sized, matched cadaveric specimens from different donors proved that the humeral head and lateral aspect of the distal part of the tibia showed excellent articular arc conformity. The surgeon is able to customize the graft to fit flush along the glenoid, which is crucial in trying to normalize glenohumeral articular contact pressure29. Furthermore, using a weight-bearing corticocancellous bone is thought to allow for better screw fixation and host-graft incorporation. Potential drawbacks include antigenicity, theoretical concerns in achieving osseous union, and poor donor bone quality26.
The open technique for anterior instability has been described utilizing a deltopectoral, subscapularis tendon-splitting approach26. The anterior aspect of the glenoid is identified and the amount of bone loss is confirmed from the preoperative computed tomography (CT) scan. Once bone loss measurements are finalized, the lateral third of the distal tibial allograft can be sized accordingly and can be cut using a sagittal saw (Fig. 1). The surgeon then has the freedom to make alterations to the graft allowing for a customized, anatomic fit to replicate the area of missing glenoid. A lag technique is utilized to securely fix two 3.5-mm fully threaded cortical screws across the graft into the glenoid. If available for repair, the capsule and labrum can also be stitched down to the screw heads before final tightening. Although less common, recurrent posterior glenohumeral instability, either developmental or traumatic, can also be successfully addressed using both open and arthroscopic techniques with distal tibial allograft augmentation33.
Cadaveric studies have compared the distal tibial allograft with other graft alternatives. Dehaan et al.31 evaluated the radius of curvature measurements for various autograft sites and allograft options to determine which would provide the best fit for glenoid augmentation. Analysis demonstrated that 94% (32 of 34 specimens) of the distal tibial allografts were ideally congruent to the glenoid’s radius of curvature. A comparison between coracoid transfer and distal tibial allograft for cadaveric reconstruction of a 30% glenoid defect showed significant improvement in glenohumeral contact areas and lower glenohumeral peak forces with the distal tibial allograft34. Clinically, this mechanical advantage of the distal tibial allograft may play a role in improving postoperative outcomes and long-term preservation of the articular surfaces.
Early results and biomechanical studies are promising, as the distal tibial allograft appears to be a viable option for reconstruction of glenoid defects for shoulder instability. We await follow-up results assessing long-term efficacy along with radiographic and functional outcomes.
Popularized by Hill and Sachs in 194035, the presence of an osseous impaction defect of the posterolateral aspect of the humeral head has been associated with an increased rate of recurrence of anterior shoulder dislocation. Burkhart and De Beer have reported that an “engaging” Hill-Sachs lesion is associated with a clinically important worse prognosis8. The unstable injury pattern will be orthogonal to the long axis of the humerus and will align longitudinally with the anterior glenoid rim when the humerus is in a position of function36,37. The extent of humeral-head defect must be taken into account along with the presence of glenoid bone loss36. Furthermore, the larger the humeral-head defect, the greater the importance and possible need for operative intervention. Flatow and Warner described clinically insignificant lesions as <20% of the articular surface, variably significant lesions as 20% to 40% of the articular surface, and clinically important lesions as >40% articular involvement38. Depending on the acuity and the extent of the abnormality, the surgical procedure can include osseous disimpaction39, soft-tissue repair (relocation of the infraspinatus tendon into the defect)40, humeral rotational osteotomy41, tightening of anterior structures (subscapularis tendon and anterior capsule)42, osseous autograft or allograft42-45, and metal implants. Yamamoto et al. presented the concept of the glenoid track, which is 83% of the glenoid width because of the thickness of the rotator cuff insertion. If a humeral lesion is medial to this measurement, then lesions are considered to be off-track and will engage the anterior aspect of the glenoid during physiologic motion36.
The same osseous allograft bone plug technique that is commonly performed for femoral condyle lesions was initially introduced by Kropf and Sekiya for filling a moderate-sized Hill-Sachs defect46. The treatment of the Hill-Sachs lesion in the shoulder may be more forgiving than osteochondral lesions in the knee because the shoulder joint does not receive the excessive loads seen in the knee. The osteochondral transplants not only serve to recreate the articular surface of the humeral head, but also, more importantly, create stability by preventing engagement with the glenoid. Furthermore, with use of an allograft, one is able to avoid altering osseous or muscular anatomy, a painful implant, and donor-site morbidity present with other techniques.
Biomechanical studies have highlighted the powerful stabilizing concavity-compression mechanism of the glenohumeral joint47,48. The stability of a convex humeral head and concave glenoid and labrum is achieved by compression with shoulder girdle muscles. Sekiya et al.49 highlighted the importance of recreating the concavity-compression stabilizing effect by restoring the humeral-head articular surface with an osteochondral allograft for lesions of >37.5% in their cadaveric study.
The technique for the humeral-head allograft can be performed through either a deltopectoral or a deltoid-splitting approach. Depending on the technique, anterior instability can be addressed before or after humeral-head graft fixation. Once the defect is exposed, either it can be shaped into a chevron pattern (Fig. 2) with a sagittal saw or it can be reamed into a circular core by use of a sizing guide to perfectly fit an osteochondral plug. A size-matched humeral-head allograft is important for recreating the contour of the humeral head in larger defects. In their basic science study, Gebhart et al.50 discovered that patient height, maximum humeral length, epicondylar breadth, and sex were most correlated with humeral-head curvature. These variables may serve as the basis of a future nomogram that can be used to estimate the radius of the humeral head for allograft size-matching. The conformity of the allograft to the original curvature of the humeral head is essential to restore the geometry, stability, and mechanics of the glenohumeral joint. In our own experience, it can take 3 to 6 months to obtain a fresh, size-matched humeral-head allograft. Based on the work of Provencher et al.26 on the distal tibial allograft for the glenoid, there may be indications for using a fresh talar allograft, which is generally more accessible. We are unaware of any particular way to size-match these grafts. Smaller defects (<15 to 20 mm) can be treated with ≥1 osteochondral plugs to fill in the majority of the defect. With smaller plugs, it is not as necessary to have a size-matched humeral head. Osteochondral plugs from a femoral head may be easier to obtain and will aid in filling the lesion. Figures 3-A and 3-B show preoperative CT images of a 17-year-old male football player with repeated shoulder dislocations. Because of the medial location of the Hill-Sachs lesion and the young age of the patient, it was decided that he would benefit from an osteochondral allograft transplant. Access was obtained using a deltoid and infraspinatus-splitting approach directly over the defect. The preoperative measurements of the width of the defect were approximately 15 mm. On the basis of the need for only a smaller graft size, a fresh femoral-head allograft was obtained. After the defect was exposed, the graft size was determined. The central area of the defect was reamed using a 15-mm reamer to an approximately 8-mm depth of the bone. While the femoral-head allograft was undergoing fixation on the back table, another reamer was used to core a 15-mm plug (Fig. 4). The plug obtained from the femoral head was measured to ensure an appropriate fit into the defect (Fig. 5). Two osteochondral plugs from the femoral-head allograft were used to fill the majority of the defect using press-fit fixation (Fig. 6). Fixation of the graft can also be obtained through the use of several different types of implants, including cortical bone screws, headless compression screws, bioabsorbable screws, and plastic screws. In a biomechanical cadaveric experiment, Puskas et al.51 found similar initial fixation and failure strength for both antegrade and retrograde fixation techniques using 3.75-mm cortical screws.
Clinically, Miniaci and Gish37 used massive humeral-head allografts in 18 patients with great success to treat recurrent glenohumeral instability that could not be treated successfully with traditional repair. The authors reported no recurrent episodes of instability with return to near-normal activity at a 2-year follow-up. Two patients required a secondary procedure for the removal of symptomatic screws used in graft fixation. Gerber and Lambert have studied the use of allograft transplantation for treating chronic locked posterior shoulder dislocations44. Four patients with 40% loss of the articular humeral head were managed with reconstruction using an allogeneic segment of the femoral head. Stability was restored and was maintained in each patient at a mean follow-up of 68 months.
In the appropriate population, osteoarticular allograft reconstruction of the humeral head appears to decrease failure rates from instability and has a goal of returning the patient to a high level of function and hopefully slowing the progression of glenohumeral arthritis. Future publications highlighting long-term patient satisfaction scores and radiographic follow-up are much anticipated.
Chondral Defects and Degeneration
In addition to reconstructing moderate to large Hill-Sachs lesions, osteochondral allografts have also been used for replacing focal chondral defects of the humeral head and glenoid. Lesions are often encountered during arthroscopic treatment of other shoulder injuries52-56. It has been reported that 5% to 17% of patients undergoing rotator cuff repair and 6% to 29% of patients treated for impingement symptoms will have grade-III to IV chondral lesions52-55. Many young or active patients with early-stage joint degeneration wish to maintain a high level of activity because of recreational interests or occupational demands. Arthroplasty may not be a practical treatment option secondary to concerns regarding implant durability57. Treatment options include arthroscopic debridement, microfracture, corrective osteotomies, and osteochondral transfers.
Microfracture is an effective reparative treatment modality for focal cartilage lesions58,59. However, with larger defects, humeral-head and glenoid osteochondral allografts may provide a viable alternative. Open biologic total shoulder arthroplasty using a press-fit humeral head was reported with some success in 7 patients with bipolar glenohumeral lesions due to post-arthroscopic glenohumeral chondrolysis60. McCarty and Cole published a case report of a 16-year-old girl who presented with bipolar glenohumeral chondrolysis after thermal capsulorrhaphy61. Because of the devastating scope of cartilage loss, young age, and high functional demand, the authors elected to combine biologic resurfacing of the glenoid using a lateral meniscal allograft with a complete osteoarticular allograft replacement of the humeral head. At a 2-year follow-up, the patient had remarkable improvement in several functional outcome scores and radiographic stability of the construct.
The treatment of glenoid chondral lesions with osteoarticular grafting procedures has been limited in part by concerns about the depth of the glenoid compared with the depth of subchondral bone on the graft necessary to achieve a press fit. Cvetanovich et al.62 performed a cadaveric imaging study and found that, in the majority of shoulders examined, the glenoid can support center-based grafts 16 to 20 mm in diameter at a depth of 4 mm, covering a mean of 51.9% of the glenoid. The authors encouraged the use of three-dimensional CT methodology to evaluate glenoid depth in shoulders with osteoarthritis and chondral lesions when planning operative intervention. At this time, to our knowledge, no important case studies have shown the long-term outcomes of using glenoid osteochondral allograft for focal chondral defect.
Although there have been short-term successes using open approaches for osteochondral allograft resurfacing, Gobezie et al.63 also described an all-arthroscopic biologic total shoulder resurfacing technique. With the use of osteochondral allografts of the humeral head and medial tibial condyle or distal tibial plafond, the technique is performed entirely through the rotator interval, limiting the damage to the subscapularis muscle.
Ultimately, optimizing joint-preservation treatment will rely on the long-term results of these procedures and postoperative documentation of the natural history of disease.
Fresh-frozen grafts are essentially acellular because of the freeze-thaw process. The selection of fresh osteochondral grafts over frozen grafts has the advantage of chondrocyte viability. Saltzman et al.64 conducted a systematic review of several Level-IV and V studies of 36 patients using osteochondral allograft transplants for large humeral head defects. Radiographic studies at the time of final follow-up (mean [and standard deviation], 57.02 ± 34.14 months) showed allograft necrosis in 8.7% of cases, resorption in 36.2% of cases, and glenohumeral arthritic changes in 35.7% of cases. Complication rates were between 20% and 30%, and the reoperation rate was 26.7%. However, only 3 patients received fresh allografts, and there were no reports of graft resorption, necrosis, or arthritic changes at the time of final follow-up.
Allograft chondrocytes have been shown to have viability in the range of 37% to 99% at up to 6 years in one study65. Another case report documented donor chondrocyte viability at 29 years66. However, the limitation of fresh allograft use is that this viability is short-lived in an ex vivo environment. The American Association of Tissue Banks has issued guidelines for the processing of allograft tissues, which result in a 14-day screening period for infectious diseases. Despite heavily regulated harvesting practices, implantation of osteochondral allografts has been reported to be a source of bacterial infection67, and risks must be thoroughly explained to the patient. The chondrocyte viability drops precipitously after 28 days68. This results in a 2-week window in which to schedule the surgical procedure once an appropriately sized graft is located and is cleared for implantation.
Osteochondral allografts have been used to successfully treat a variety of shoulder conditions and to improve clinical results for complex chondral defects. Joint stability may be restored when bone stock has been lost from the glenoid, the humeral head, or both. Furthermore, young, active patients may avoid well-known complications of prosthetic glenoid replacement for osteoarthritis.
As with all allograft procedures, there can be complications related to the allograft and the complex nature of the procedure. Minimizing the time until fresh osteochondral allograft implantation is crucial to maintaining the viability of the cartilage. Optimizing joint-preservation treatment with osteochondral allografts will rely on the long-term results of these procedures, and careful patient selection, substantial preoperative discussion, and realistic expectations are necessary.
Investigation performed at the University of Missouri, Columbia, Missouri
Disclosure: No external funding sources were used in the execution of this research. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated