Midterm Results of Osteochondral Lesions of the Talar Shoulder Treated with Fresh Osteochondral Allograft Transplantation

Adams, Samuel B. Jr. MD; Viens, Nicholas A. MD; Easley, Mark E. MD; Stinnett, Sandra S. DrPH; Nunley, James A. II MD

Journal of Bone & Joint Surgery - American Volume:
doi: 10.2106/JBJS.J.00141
Scientific Articles

This article was updated on May 4, 2011, because of a previous error. On page 648, in the Conclusions section of the Abstract, the text had previously read “These midterm results in a small group of patients indicate that structural fresh-frozen allograft transplantation can be a successful surgical option in the treatment of large osteochondral defects of the talar shoulder.” The text now reads “These midterm results in a small group of patients indicate that structural fresh allograft transplantation can be a successful surgical option in the treatment of large osteochondral defects of the talar shoulder.”

An erratum has been published: J Bone Joint Surg Am 2011;93(12):1158.

Background: With osteochondral lesions of the talar shoulder, their size, the articular cartilage geometry, and the loss of the medial or lateral articular buttress often preclude treatment with traditional osteochondral autograft techniques. We hypothesized that fresh, large osteochondral allograft transplantation is a viable treatment option for patients with such lesions.

Methods: A retrospective review was conducted of patients who underwent fresh talar shoulder allograft transplantation between 2000 and 2007. All patients failed initial conservative management. Preoperatively, a visual analog pain scale of 0 to 10 (with 0 denoting no pain and 10 denoting the worst pain imaginable) and the Lower Extremity Functional Scale were administered. At the time of most recent follow-up, the visual analog pain scale, the Lower Extremity Functional Scale, the American Orthopaedic Foot & Ankle Society ankle-hindfoot scale, and the Short Musculoskeletal Function Assessment questionnaire were administered. Radiographs were assessed for allograft incorporation and joint deterioration.

Results: Eight patients with a mean age of thirty-one years and a mean follow-up of forty-eight months were included. There was a significant decrease (p < 0.05) in pain, from a mean of 6 points (range, 5 to 8 points) preoperatively to a mean of 1 point (range, 0 to 2 points) postoperatively. The mean Lower Extremity Functional Scale score improved from 37 points initially (range, 24 to 52 points) to 65 points (range, 31 to 75 points; p < 0.05) at the time of final follow-up. The mean postoperative American Orthopaedic Foot & Ankle Society ankle-hindfoot score was 84 points. The mean Short Musculoskeletal Function Assessment dysfunction index score was 13.3 points and the mean bother index score was 14.3 points. Radiographic lucencies at the graft-host interface were seen in five patients. Four patients required an additional surgical procedure. No patients needed to undergo subsequent arthrodesis or arthroplasty.

Conclusions: These midterm results in a small group of patients indicate that structural fresh allograft transplantation can be a successful surgical option in the treatment of large osteochondral defects of the talar shoulder.

Level of Evidence: Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.

Author Information

1Division of Orthopaedic Surgery (S.B.A. Jr., N.A.V., M.E.E., and J.A.N. II) and Department of Ophthalmology (S.S.S.), Duke University Medical Center, Durham, NC 27710. E-mail address for S.B. Adams Jr.: samueladamsmd@gmail.com. E-mail address for J.A. Nunley II: nunle001@mc.duke.edu

Article Outline

Osteochondral lesions of the talus are defects of the articular cartilage of the talus and the underlying subchondral bone1-3. They can pose a formidable treatment challenge to the orthopaedic surgeon due to the poor intrinsic ability of cartilage to heal as well as the tenuous vascular supply to the talus. Traditional osteochondral grafting techniques such as the osteochondral autograft transfer system (OATS) and mosaicplasty have been described as viable treatment options4,5.

However, a subset of these lesions involves the talar shoulder. These are typically large lesions involving a substantial amount of the weight-bearing portion of the talar dome and either the medial or lateral articular surface (Fig. 1). The size of the lesion, the two-dimensional articular cartilage geometry, the three-dimensional osseous geometry, and the loss of the medial or lateral articular buttress may preclude treatment with current cylindrical osteochondral autograft techniques. However, for young patients without diffuse joint involvement, viable treatment options short of arthrodesis or arthroplasty have rarely been reported. In 2001, Gross et al.6 reported encouraging results with use of fresh allograft transplantation for talar shoulder lesions, and Raikin7 recently reported successful results with use of fresh osteochondral allografts for large cystic lesions of the talus.

Our hypothesis was that large fresh geometrically contoured talar allografts would provide clinically and radiographically successful outcomes in patients with large osteochondral lesions of the talar shoulder.

Back to Top | Article Outline

Materials and Methods

Study Design and Preoperative Evaluation

A retrospective review was conducted of all patients who underwent fresh osteochondral allograft transplantation for osteochondral lesions of the talar shoulder at our institution between December 2000 and July 2007. Patients with cystic and noncystic lesions were included. Previous operative or nonoperative treatment had failed for all patients. A minimum follow-up of two years was required. Institutional review board approval was obtained prior to initiation of this study.

Preoperative assessment included physical examination, weight-bearing radiographs, magnetic resonance imaging (MRI) and/or computed tomography (CT), and a visual analog pain scale (ranging from 0 to 10, with 0 denoting no pain and 10 denoting the worst pain imaginable), and the Lower Extremity Functional Scale (LEFS)8. The LEFS is a validated lower-extremity-specific outcome measure with twenty items. It is scored from 0 to 80, with 80 points representing the best physical function. This scale was based on the World Health Organization's model of disability and handicap and is applicable to all ages. Questions mostly pertain to activities of daily living, such as job performance, housework, donning shoes and socks, walking, picking up groceries, sitting, and standing8.

The required size of the donor talus was calculated from radiographs, MRI, and CT of the contralateral talus. The patients were placed on a waiting list and treated with supportive modalities until the fresh graft became available.

All osteochondral allografts were obtained from nationally recognized suppliers who comply with the procurement and processing guidelines of the U.S. Food and Drug Administration, the American Association of Tissue Banks9, and Clinical Laboratory Improvement Amendments. The general time line of the graft preparation was as follows: Once a donor was identified, the talus was harvested under sterile conditions within twenty-four hours after death. Next, the allograft underwent disease-testing and sterility-culturing for approximately fourteen days. During this time, the donor's medical chart was reviewed for factors that might lead to an unacceptable graft (e.g., high-risk behavior). If the graft passed these screens, proprietary processing techniques were performed and the graft was packaged for shipping (at approximately day twenty-one after harvest). The graft was maintained at 2°C to 4°C throughout this process. The graft typically arrived at the hospital in one day. Prior to shipping, the surgeon was notified of graft approval and contacted the patient. Patient and surgeon availability then ultimately affected the time of allograft implantation, but, typically, the delay was less than one week. Every effort was made to perform the procedure as soon as possible. During this entire time, the allograft was kept refrigerated. No age, sex, tissue, or blood-type matching was performed. Serum anti-human leukocyte antigen (serum anti-HLA) antibodies were not measured.

Back to Top | Article Outline
Surgical Technique

The ankle joint was approached through either a medial or lateral malleolar osteotomy, corresponding to the location of the shoulder lesion. The talar shoulder lesion was identified and debrided to stable articular cartilage margins and bleeding cancellous bone. The edges were cut square with a fine saw to allow for correct geometric fitting of the allograft. The dimensions of the lesion were marked onto the corresponding location of the fresh allograft talus. The graft was then cut with an oscillating saw, and it was stabilized with one or two countersunk titanium mini-fragment (2.0-mm) headed cancellous screws. The medial or lateral osteotomy was repaired with standard internal fixation. Intraoperative fluoroscopy was performed to confirm the placement and fit of the graft and the alignment of the malleolar osteotomy. One patient with a large cystic lesion underwent cancellous bone-grafting from the tibia at the time of allograft transplantation.

Postoperatively, the ankle was immobilized for eight to ten weeks in a non-weight-bearing cast. The patient was then transitioned into a controlled-ankle-motion walking boot, which was worn until the osteotomy was fully healed.

Back to Top | Article Outline
Postoperative Evaluation

At the most recent follow-up, all patients underwent a physical examination and weight-bearing radiographs were made. Additionally, all patients were administered the same visual analog pain scale that they had completed before the operation, the LEFS, the American Orthopaedic Foot & Ankle Society (AOFAS) ankle-hindfoot scale, and the Short Musculoskeletal Function Assessment (SMFA) questionnaire10,11. The SMFA questionnaire consists of the dysfunction index, which has thirty-four questions for the assessment of patient function, and the bother index, which has twelve questions for the assessment of how much patients are bothered by functional problems. Both of these scales are scored from 0 to 100 points, with 0 representing no dysfunction or bother and higher scores indicating poorer function.

Two patients were out of state and were followed by a local orthopaedic surgeon. These patients completed the outcomes questionnaires via mail and their clinic notes and radiographs were mailed to our institution for review. All postoperative imaging, performed at the most recent follow-up, was reviewed for graft collapse, graft nonunion, osteotomy nonunion, or loss of ankle-joint space. If graft failure was a concern, a CT scan was obtained in addition to weight-bearing radiographs.

Back to Top | Article Outline
Statistical Analysis

Statistical analysis was performed on the preoperative and postoperative point scales with use of a paired t test. The preoperative and postoperative LEFS scores were compared with use of the Wilcoxon signed-rank test. Patient demographics were compared with the outcomes measures with use of the Spearman correlation. Statistical significance was set at p < 0.05.

Back to Top | Article Outline
Source of Funding

There was no external funding for this study.

Back to Top | Article Outline


Patient and Allograft Demographics

Eight patients with a mean age of thirty-one years (range, seventeen to forty-four years) were included in this study. There were three men and five women with a mean follow-up of forty-eight months (range, twenty-five to ninety-one months). The mean time from the clinic visit when the decision was made to treat the patient with osteochondral allograft transplantation to the time of surgery was 104 days (range, forty to 174 days). The cause and location of the osteochondral lesion as well as duration of symptoms, the presence of cysts in the lesion, lesion volume, and allograft volume are summarized in Table I. The mean lesion volume, as measured preoperatively on MRI or CT, was slightly greater than the actual allograft volume. This greater volume likely represented edema that was seen on the MRI scans and that interfered with our ability to identify the true edge of the lesion.

Only three patients could recall a specific traumatic event. None of the grafts approached the talar sulcus or crossed the midline of the talus (i.e., there were no hemi-talar replacements in this series).

All allografts used in this study were fresh (never frozen). The mean age of the donors was twenty-six years (range, seventeen to forty years). The mean number of days from host death to allograft implantation was twenty-six (range, twenty-four to twenty-seven days). The allografts were obtained from five males and three females.

Back to Top | Article Outline
Outcomes Measures

A summary of the outcomes-measures data is presented in Table II. There was a significant decrease in pain at the time of final follow-up (p < 0.05), from a mean of 6 points (range, 5 to 8 points) preoperatively to a mean of 1 point (range, 0 to 2 points) postoperatively. Likewise, there was a significant improvement in the mean LEFS score from 37 initially to 65 at the time of final follow-up (p < 0.05, Table II).

The mean postoperative AOFAS ankle-hindfoot score was 84 points (range, 71 to 91 points). The mean pain component of the AOFAS ankle-hindfoot score was 30 points (range, 20 to 40 points), and the mean functional component was 44 points (range, 35 to 50 points). The mean SMFA dysfunction index score was 13.3 points (range, 0.7 to 55.9 points) and the mean bother index score was 14.3 points (range, 2.1 to 66.7 points). These index scores are similar to the reported normative “uninjured” overall population scores10.

Final follow-up radiographs demonstrated no lucencies, graft subsidence, joint-space narrowing, or degenerative changes in three patients. Three patients with medial grafts were found to have partial lucency along the lateral border at the interface of the allograft and the host bone (Fig. 2). These three patients were functioning well with almost complete resolution of preoperative symptoms and, therefore, no further imaging was obtained. One additional patient appeared to have superior graft resorption and lucency along the lateral border of the graft at the interface with the host bone. A CT scan was obtained. No graft resorption was identified. The lucency at the junction of the allograft and host bone was only found at the anterior graft-recipient interface. The graft was believed to be stable, and the patient's symptoms had improved. No further treatment was performed.

One patient (Case 6) had persistent pain and a “clicking” sensation and was found to have lucency surrounding the entire graft. A CT scan was acquired, which suggested possible nonunion of the allograft and prominent hardware. Ankle arthroscopy was performed, revealing partial delamination of the graft at the graft-host interface, prominent fixation, and scar tissue along the medial gutter. The dorsal fixation was removed, and the graft was found to be stable; therefore, it was left in situ. This same patient underwent a second arthroscopic debridement of the medial gutter at approximately four years after graft placement. There was fraying of the cartilage edges of the graft, but no progression of the delamination. This patient consistently had the worst outcomes-measures scores. At the latest follow-up, the patient did not wish to have any further treatment.

Three other patients required additional surgery. One patient underwent removal of painful medial malleolar hardware from a united medial malleolar osteotomy. One patient underwent revision open reduction and internal fixation of an ununited medial malleolar osteotomy. One patient with a medial allograft was found to have varus malalignment of the ankle and underwent supramalleolar and calcaneal osteotomies to protect the graft. Unfortunately, this patient initially presented with radiographs that had been made elsewhere and that offered only limited visualization of the tibia, and the patient refused to have additional radiographs made at our institution. It is likely that the amount of preoperative varus was underappreciated.

Back to Top | Article Outline


The treatment of osteochondral lesions involving the talar shoulder remains a challenge. Circular autologous osteochondral autograft plugs are a successful treatment option for talar dome lesions4,5 but may not be ideal for shoulder lesions due to a lack of structural support. The use of structural allograft transplantation for osteochondral lesions of the talar shoulder has rarely been reported in the literature. The patient population in the present study demonstrated decreased pain and improvement in physical function at a mean follow-up of four years after undergoing this procedure. There were no allograft failures, but one patient did experience partial delamination of the graft at the graft-host interface, with a probable fibrous union.

Patients with osteochondral talar lesions typically complain of pain affecting the ability to perform activities of daily living. The patients in this series reported significant improvement in pain and LEFS scores. The LEFS is a specific functional assessment of the lower extremity. Moreover, the sensitivity to change of the LEFS has been reported to be superior to that of the Short Form-368. This is perhaps an argument for using more regional or musculoskeletal-specific outcomes measurements to assess orthopaedic conditions. Interestingly, although the small size of our patient population limited statistical analysis, the mean scores for both of the SMFA indices for our patients (13.3 for the dysfunction index and 14.3 for the bother index) were similar to the mean scores of a sampling of healthy people from the U.S. population (12.7 for the dysfunction index and 13.8 for the bother index)10.

Previously, Gross et al.6 reported on nine patients with similar talar lesions who underwent fresh osteochondral allograft transplantation. At a mean follow-up of eleven years, six grafts remained in situ. The three failed allografts demonstrated radiographic and intraoperative evidence of fragmentation or resorption, and these patients subsequently underwent ankle arthrodesis at thirty-six, fifty-six, and eighty-three months, respectively, following the allograft surgery. Standardized outcomes measures were not used in that study. The relatively late conversion to arthrodesis indicates the need for long-term monitoring of patients who have been managed with osteochondral allografts.

Raikin7 recently reported on fifteen patients who underwent bulk fresh osteochondral allografting for large-volume cystic lesions of the talus. The mean volume of the cystic lesions was 6059 mm3. At a mean follow-up of 4.5 years, the mean AOFAS ankle-hindfoot score was 83 points. Only two grafts failed and necessitated the performance of ankle arthrodesis. Some form of graft collapse, graft resorption, or joint-space narrowing was seen in all patients. Additionally, Jeng et al.12 reported on total ankle allograft transplantation for the treatment of ankle arthritis in twenty-nine patients, with a mean follow-up of twenty-four months. Of the successful grafts, the mean AOFAS ankle-hindfoot score was 84 points, which was the same as our overall mean score.

Lucency at the interface of the allograft and host bone was found in five of eight of our patients. This is in contrast to the results of Görtz et al.13, who reported that the graft-host interface was not visible in any of twelve patients who underwent fresh osteochondral allografting of the talus at a mean follow-up time of thirty-eight months. The presence of graft-host lucency did not seem to affect the treatment outcome in four of our five patients. One patient who was found on CT to have lucency surrounding the graft underwent two arthroscopic debridements. Despite some cartilage delamination, the graft was stable. It is possible that the lucent areas in these patients are filled with fibrocartilage. As previously mentioned, Gross et al.6 reported radiographic evidence of resorption in three failed allografts. Perhaps the lucencies seen in our patients represent early graft resorption; therefore, continued monitoring and long-term follow-up are needed.

It is important to mention the differences in the radiographically measured volume of the lesion and the true volume of the lesion or the allograft volume. We routinely obtain either an MRI or a CT scan, and sometimes both, to aid in the determination of the size of the lesion. We realize that this is still an estimated volume. The size of the lesion is often overestimated on MRI scans, secondary to surrounding bone-marrow edema14; however, having an estimated size is still very helpful. It provides the treating surgeon with reassurance that the lesion is indeed large enough to proceed with allograft transplantation, and it provides a better estimate as to whether the lesion approaches the talar midline (the sulcus). It is our general practice that if the lesion does approach the talar midline, the allograft becomes more of a hemi-talar replacement. In that situation, we use an anterior approach instead of a medial or lateral malleolar osteotomy.

Although it is generally thought that chondrocytes are immunoprivileged secondary to their surrounding extracellular matrix15, a recent report indicated the presence of an immunologic response to cartilage-specific protein in eight of fourteen osteochondral allografts16. Additionally, marrow elements are not considered immunoprivileged17. Meehan et al.18 demonstrated evidence of serum anti-HLA cytotoxic antibodies in ten of eleven patients who underwent fresh tibiotalar osteochondral allografting. Interestingly, the one patient without cytotoxic antibodies was taking immunosuppressant medications after a kidney transplant. Although a correlation of positive cytotoxic antibodies and graft survival was not made, the authors speculated that the immune response may play an important role in graft survival. However, the role of cytotoxic antibodies or HLA-matching remains unknown. We do not routinely perform these tests; this is in concordance with many other published reports on osteochondral allografting of the talus6,7,12,13. We do copiously irrigate the allograft, prior to implantation, in an attempt to reduce the antigenic load.

Damaged adult cartilage, as in an osteochondral lesion of the talus, lacks the ability to repair itself. Therefore, the basis of allograft transplantation is to provide viable chondrocytes that can maintain the cartilage extracellular matrix and thus provide long-term survival of the joint. Fresh osteochondral allografts have been shown to contain viable chondrocytes up to seventeen years after transplantation19,20. On the contrary, using frozen osteochondral allografts, Enneking et al.21,22 demonstrated absence of viable chondrocytes and evidence of cartilage breakdown as early as one year after transplantation. Therefore, we prefer to use fresh osteochondral allografts rather than cryopreserved grafts. However, it is important to note that chondrocyte viability decreases with time in fresh allografts. Williams et al.23 reported a significant decrease in viable chondrocytes of human fresh osteochondral allografts by day twenty-eight. However, at day twenty-eight, the mean percentage of remaining viable chondrocytes was still 70% of the starting amount. Although the transplants in this study were implanted prior to twenty-eight days, the mean time from donor death to implantation was twenty-six days. We make every effort to perform the transplantation as soon as the allograft is released, but there are many patient, surgeon, and hospital factors that ultimately dictate the timing of the procedure.

Disease transmission is a concern with any transplantation procedure, and disease transmission resulting from osteochondral allografting remains a concern. The current estimated risk of human immunodeficiency virus (HIV) transmission through allograft tissue is one in one million24. There have been three reported cases of HIV transmission, two cases of hepatitis-C-virus transmission, and one case of hepatitis-B-virus transmission through allograft tissue24. We are unaware of any instances of disease transmission in our patient population. We do counsel our patients about the risks and benefits of receiving allograft tissue.

This study is limited by its small sample size, the length of follow-up, the use of an unanchored pain scale, and the retrospective nature of the study. Unfortunately, this procedure is relatively rare and most studies reporting outcomes are limited with respect to sample size. The mean follow-up of forty-eight months is on par with other published series7,13.

Osteochondral lesions of the talar shoulder in young patients without widespread joint disease present a treatment challenge. We present midterm pain, functional, and radiographic data to support the use of structural talar allograft transplantation in this population. We believe that this procedure does not preclude repeat allograft transplantation, conversion to total ankle arthroplasty, or ankle arthrodesis.

Investigation performed at Duke University Medical Center, Durham, North Carolina

Back to Top | Article Outline


1. Berndt AL Harty M. Transchondral fractures (osteochondritis dissecans) of the talus. J Bone Joint Surg Am. 1959;41:988–1020.
2. Giannini S Vannini F. Operative treatment of osteochondral lesions of the talar dome: current concepts review. Foot Ankle Int. 2004;25:168–75.
3. Pappas AM. Osteochondrosis dissecans. Clin Orthop Relat Res. 1981;158:59–69.
4. Al-Shaikh RA Chou LB Mann JA Dreeben SM Prieskorn D. Autologous osteochondral grafting for talar cartilage defects. Foot Ankle Int. 2002;23:381–9.
5. Hangody L Füles P. Autologous osteochondral mosaicplasty for the treatment of full-thickness defects of weight-bearing joints: ten years of experimental and clinical experience. J Bone Joint Surg Am. 2003;85:25–32.
6. Gross AE Agnidis Z Hutchison CR. Osteochondral defects of the talus treated with fresh osteochondral allograft transplantation. Foot Ankle Int. 2001;22:385–91.
7. Raikin SM. Fresh osteochondral allografts for large-volume cystic osteochondral defects of the talus. J Bone Joint Surg Am. 2009;91:2818–26.
8. Binkley JM Stratford PW Lott SA Riddle DL. The Lower Extremity Functional Scale (LEFS): scale development, measurement properties, and clinical application. North American Orthopaedic Rehabilitation Research Network. Phys Ther. 1999;79:371–83.
9. Jacobs N. Establishing a surgical bone bank. In: Fawcett BK, editor. Tissue Banking. Arlington: American Association of Blood Banks; 1987. p 67–96.
10. Barei DP Agel J Swiontkowski MF. Current utilization, interpretation, and recommendations: the musculoskeletal function assessments (MFA/SMFA). J Orthop Trauma. 2007;21:738–42.
11. Swiontkowski MF Engelberg R Martin DP Agel J. Short musculoskeletal function assessment questionnaire: validity, reliability, and responsiveness. J Bone Joint Surg Am. 1999;81:1245–60.
12. Jeng CL Kadakia A White KL Myerson MS. Fresh osteochondral total ankle allograft transplantation for the treatment of ankle arthritis. Foot Ankle Int. 2008;29:554–60.
13. Görtz S De Young AJ Bugbee WD. Fresh osteochondral allografting for osteochondral lesions of the talus. Foot Ankle Int. 2010;31:283–90.
14. Lee KB Bai LB Park JG Yoon TR. A comparison of arthroscopic and MRI findings in staging of osteochondral lesions of the talus. Knee Surg Sports Traumatol Arthrosc. 2008;16:1047–51.
15. Langer F Gross AE. Immunogenicity of allograft articular cartilage. J Bone Joint Surg Am. 1974;56;297–304.
16. Phipatanakul WP VandeVord PJ Teitge RA Wooley PH. Immune response in patients receiving fresh osteochondral allografts. Am J Orthop (Belle Mead NJ). 2004;33:345–8.
17. Langer F Czitrom A Pritzker KP Gross AE. The immunogenicity of fresh and frozen allogeneic bone. J Bone Joint Surg Am. 1975;57;216–20.
18. Meehan R McFarlin S Bugbee W Brage M. Fresh ankle osteochondral allograft transplantation for tibiotalar joint arthritis. Foot Ankle Int. 2005;26:793–802.
19. Convery FR Akeson WH Amiel D Meyers MH Monosov A. Long-term survival of chondrocytes in an osteochondral articular cartilage allograft. A case report. J Bone Joint Surg Am. 1996;78:1082–8.
20. McGoveran BM Pritzker KP Shasha N Price J Gross AE. Long-term chondrocyte viability in a fresh osteochondral allograft. J Knee Surg. 2002;15:97–100.
21. Enneking WF Campanacci DA. Retrieved human allografts: a clinicopathological study. J Bone Joint Surg Am. 2001;83:971–86.
22. Enneking WF Mindell ER. Observations on massive retrieved human allografts. J Bone Joint Surg Am. 1991;73:1123–42.
23. Williams SK Amiel D Ball ST Allen RT Wong VW Chen AC Sah RL Bugbee WD. Prolonged storage effects on the articular cartilage of fresh human osteochondral allografts. J Bone Joint Surg Am. 2003;85:2111–20.
24. Mroz TE Joyce MJ Steinmetz MP Lieberman IH Wang JC. Musculoskeletal allograft risks and recalls in the United States. J Am Acad Orthop Surg. 2008;16:559–65.

Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor a member of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity.

Copyright 2011 by The Journal of Bone and Joint Surgery, Incorporated