Abstract
Purpose:
To report the midterm outcome of a novel reconstructive technique using a contoured iliac crest bone graft for partial radial head replacement in the treatment of complex elbow dislocation.
Material and methods:
Between January 2008 and December 2013, 10 patients (5 women, 5 men; mean age, 43.8 years; mean follow-up duration, 65.9 months) with complex elbow dislocation who underwent the partial radial head replacement with the contoured bone graft were included in the study. The irreparable radial head defects averaged 49% of the articular surface (range, 36–60%). The fracture involved the entire head in four patients and partial head in six patients.
Results:
At the final follow-up, the mean elbow extension was 8°, flexion 143°, supination 76.5°, and pronation 73°. The mean Mayo elbow performance index was 93.2 points and the Broberg–Morrey functional rating score was 94.1 points, with seven excellent cases, two good cases, and one fair case. Radiographic union was achieved in all but one, at an average time of 6.89 weeks (range, 6–10 weeks). The final radiographs demonstrated no evidence of degenerative change in eight patients, mild arthritis in one patient, and moderate arthritis in one patient.
Conclusion:
This technique is a viable option in the treatment of the large radial head defect in complex elbow dislocation when more than 40% of the original head is still available for incorporation.
Keywords
Introduction
Reconstruction of the radial head and restoration of the radiocapitellar contact is usually recommended in complex elbow dislocation to prevent re-dislocation, persistent instability, and post-traumatic arthritis (PTA). 1,2,3 However, the management of radial head fractures with large irreparable defects remains a challenging problem. 3,4
In the clinical setting of complex elbow dislocation, discarding the unrepairable fragments, termed partial radial head resection, showed poor results. 5 According to in vitro biomechanical studies, a progressive decrease in radiocapitellar joint stability is observed, with increasing size of the radial head defect. 6,7 A wedge defect greater than one-third of the articular surface is considered significant as the radial head is therefore unable to contribute to joint stability. In this situation, replacement of the entire radial head with a metal prosthesis is preferred. 3,8 However, in some countries, prostheses are not available or covered by the countries’ health insurance. 9 Alternatively, the replacement of the whole or partial radial head with a fresh-frozen allograft has been described. 10 –12 While replacement of the entire radial head allograft demonstrated variable results, others reported a favorable outcome with partial radial head replacement. 12 However, the use of allografts requires a bone bank that can supply radial head allografts of various sizes, and many countries do not have ready access to such facilities.
The purpose of the present study is to report on the midterm results of a reconstructive technique using partial replacement of the radial head with contoured iliac crest bone grafts in 10 patients with complex posterior elbow dislocations. The rationale for using this technique is that the radial head can be preserved by converting a complicated fracture into a simpler configuration that is amendable for open reduction and internal fixation (ORIF). The surgeon is able to discard the irreparable bone fragments and restore the original radial head profile using the remaining radial head as a template.
Materials and methods
Between January 2008 and December 2013, a total of 10 patients with a complex elbow dislocation underwent iliac crest reconstructive bone grafts. The medical record was reviewed for demographic data and for preoperative, intraoperative, and postoperative information. All patients were contacted for clinical and radiographic assessments. All patients were also informed that data regarding their clinical and radiological examinations would be submitted for publication. The protocol of this report was approved by the institutional review board.
Patients
The study group consisted of five women and five men with a mean age of 43.8 years at the time of surgery (range, 22–67 years). The average duration of postoperative follow-up was 65.9 months (range, 36–110 months).
The mechanism of injury included a fall on the outstretched hand from a standing height in three patients, a fall from a greater than standing height in three patients, a sport-related injury in two patients, a bicycle accident in one patient, and a motor vehicle accident in one patient. The mean period from injury to operation was 3.5 weeks (range, 0–7 weeks). Five patients were referred to our institution from other centers 4 weeks after the injury. One patient had already undergone partial radial head excision and an external fixator was applied because of instability issues. The concomitant elbow lesion injuries included coronoid fracture (all cases), olecranon fracture (patients 1 and 5), avulsion of lateral collateral ligament (LCL) complex (all cases), and avulsion of medial collateral ligament (MCL) (patients 3 and 8). In view of the patterns of fracture dislocation, 13 the posterolateral (PL) rotatory injury (terrible triad) was revealed in eight patients and the posterior olecranon fracture dislocation (Monteggia lesion) was seen in two patients. In both patterns, the instability still persisted after attempted restoration of the ulnohumeral articulation.
Four patients sustained complete radial head fracture: irreparable fragments (patients 5 and 9) and missing fragments (patients 3 and 4). Partial radial head fracture was seen in six patients: prior partial head excision (patient 1), irreparable fragments (patients 7 and 10), and impaction with deformation of the fracture fragment (patients 2, 6, and 8) (Figure 1(a) to (d)). Size of the defect was defined as the area involved at the joint surface articulating with the capitellum. The intraoperative digital photographs were taken and imported into the image processing program (ImageJ, National Institutes of Health, Bethesda, MD, USA). Size of the defect was quantified by the following equation: Percentage of bone defect = (defect surface area/whole articular surface area) × 100%. Due to the marginal loss on the defect, the best-fit ellipse was drawn along the visible margin of the intact portion to assume the whole articular surface boundary. The average size was 49% of the total articular surface (range, 36–60%). When the radial head was divided into four quadrants using the radial tuberosity referenced as the straight posterior line according to the study of Capo et al,
14
the bone defect was located in the anterolateral (AL) and anteromedial (AM) quadrants in five patients, AM and posteromedial (PM) quadrants in three patients, and AM quadrant alone in two patients. The coronoid fractures were categorized based on O’ Driscoll’s classification system
15
: tip subtype 2 in four patients, AM subtype 2 in four patients, and basal subtype 2 in two patients. All olecranon fractures were classified as type IIIB in Mayo classification because of the comminution and accompanying elbow instability.
16
Partial replacements of the radial heads with contoured iliac crest bone grafts were performed in injuries resulting in posterior elbow instability, with large defects of the radial head. (a) Patient 1: prior fragment excision. (b) Patient 3: missing fragment. (c) Patient 10: severe comminuted fragments. (d) Patient 8: deformation of the fracture fragment.
Operative technique
The radial head was exposed through the extensor digitorum communis splitting approach. The situation of the radial head fracture was assessed, all fracture fragments were identified, and the articular defect was determined. Prior to the radial head reconstruction, the associated coronoid fracture was repaired. A medial approach was required if the fracture was inaccessible. The olecranon was reduced and fixed through the posterior approach, if required.
In partial head fractures, the irreparable fracture fragments were removed. The articular margin of the undamaged portion was shaped into a smooth straight line using a microsagittal saw cutting perpendicular to the joint surface. When the fracture involved the AM and PM quadrants, and the defect was not well visualized when the forearm was fully supinated, additional removal of the bone in the AL quadrant was performed in order to improve access to the defect. The second bone cut was made along the head–neck junction to meet the first cut so that the contact area increases and creates a stable platform for the bone graft. The autogenous iliac bone graft was harvested and contoured with the remaining undamaged portion to reform a full radius of the radial head, and two 1.2-mm Kirschner wires were inserted for temporary fixation. A motorized burr was used to smooth the graft edge, recreate the curvature of the articular disk, and flush the radiocapitellar articulating surface of the graft with the native articular surface. The graft was secured to the original radial head with one or two 3.0-mm cannulated headless screws (Cannulated Screw Systems, OsteoMed, Christchurch, Dorset, UK). The forearm pronation and supination was checked. Residual bony prominences of the graft causing the impingement of the proximal radioulnar joint during forearm rotation were removed (Figure 2(a)–(d)).

Reconstruction technique in partial radial head fracture. (a) The irreparable fracture fragments were removed. (b) The articular margin of the undamaged portion was shaped to create the stable platform for the graft. (c) The bone graft was contoured with the corresponding remaining portion to reform a full radius of the radial head. (d) The graft was secured to the original radial head with the cannulated headless screws.
For fractures involving the entire radial head, one or two of the largest fragments comprising at least 40% of the articular surface were preserved to be used as a core fragment. The remaining attached periosteum was left intact in order to maintain the blood supply. In cases where the periosteum of the preserved bone has already been detached completely, or the exposure was inadequate, an on-table reconstruction technique was performed as described by Businger et al (Figure 3(a) and (b)).
17
Care was taken to realign the fracture line of the preserved fragment with the corresponding line at the radial neck to prevent malrotation of the reconstructive head. Any irreparable small fragments were removed. The articular margin of the preserved radial head and head–neck junction was fashioned in the same way as performed in the partial head fracture. The iliac bone graft was harvested and contoured to match the defect. The graft and the preserved bone fragment were realigned and secured with one or two headless screws. The reconstructed radial head was then secured to the radial neck with an anatomical proximal radius locking plate (LCP Proximal Radius Plates 2.4; Synthes, Paoli, PA, USA). The plate was placed on the safe zone as described by Smith et al.
18
On-table reconstruction technique in entire radial head fracture. (a) Patient 4: the largest radial head fragment comprising 40% of the articular surface was preserved. (b) Patient 3: the two largest fragments were preserved.
The avulsed LCL was reattached with a transosseous suture or a bone anchor (FASTak II Suture Anchor system; Arthrex, Naples, FL, USA). The annular ligament was repaired using absorbable sutures. Concentric reduction in radiocapitellar and ulnohumeral joints through the entire range of motion was assessed under clinical examination and fluoroscopy.
Postoperative management
Postoperatively, the elbow was immobilized in 90° of flexion. Active and active-assisted range-of-motion exercises were initiated within 1 week after the surgery. We did not use any medications or irradiation as prophylaxis against heterotopic ossification. Weight bearing was not allowed until the radiologic consolidation was documented.
Evaluation
The evaluation consisted of an interview, a physical examination, radiographic examination, and outcome assessment. Outcome measurement was performed at the final follow-up visit with the Mayo elbow performance index (MEPI), the Broberg–Morrey functional rating index, and the patient-based Disabilities of the Arm, Shoulder and Hand (DASH). The morbidity including residual pain, sensory disturbances, and functional limitations associated with harvesting bone from anterior iliac crest was documented.
Anteroposterior and lateral radiographs of the elbow were reviewed for the congruity of the radial head, osseous union, joint incongruity, and PTA. At the time of final follow-up, radiographic signs of PTA were assessed according to the criteria of Broberg and Morrey as grade 0 (normal joint), grade 1 (slight joint-space narrowing and minimum osteophyte formation), grade 2 (moderate joint-space narrowing and moderate osteophyte formation), and grade 3 (severe degenerative changes and destruction of the joint). 19
Results
The results are summarized in Table 1. Intraoperatively, the concentric reduction of the elbows was achieved in all cases without the need for MCL repair or immobilization with the hinged external fixator. Malreduction with the 25° angulation of the radial head was observed in one case of the entire head fractures (patient 4). In this series, there was no incidence of infection, secondary fracture, or implant dislocation. No patient had morbidity associated with the bone graft donor site, had stiffness of the fingers, or had developed a complex regional pain syndrome. In three of four patients who had the plate fixation, the plate was removed postoperatively due to irritations with the annular ligament. All but one fracture achieved radiographic union, with the evidence of bridging trabeculae and obliteration of the bone interface at an average time of 6.89 weeks (range, 6–10 weeks). In patient 3, the follow-up radiographs showed that one piece of the original radial head fragment interface was not united. This fragment was removed at the time of plate removal.
Details of the study group and results.
DASH: Disabilities of the Arm, Shoulder and Hand; MEPI: Mayo elbow performance index; f, female; m, male; R, right; L, left; AL, anterolateral; AM, anteromedial; PM, posteromedial; PL, posterolateral; OF, olecranon fracture; BS 2 CF, basal subtype 2 coronoid fracture; AMS 2 CF, anteromedial subtype 2 coronoid fracture; TS 2 CF, tip subtype 2 coronoid fracture; DRF, distal radius fracture; MCL, medial collateral ligament; LCL, lateral collateral ligament; ROM, range of motion; Fle, flexion; Ext, extension; Pro, pronation; Sup, supination; PTA: post-traumatic arthritis.
aMayo elbow performance index (MEPI), score: 90–100 points, excellent; 75–89 points, good; 60–74 points, fair; <60 points, poor.
bBroberg–Morrey functional rating index, score: 95–100 points, excellent; 80–94 points, good; 60–79 points, fair, <60 points, poor.
cDisabilities of the Arm, Shoulder and Hand (DASH) Outcome Measure.
dRadiographic rating of PTA according to the criteria of Broberg and Morrey.
The final arc of elbow movement averaged 135° (range, 125°–145°), with a mean flexion of 143° (range, 140°–150°) and a loss of extension of 8° (range, 0°–25°). The final arc of forearm rotation averaged 149.5° (range, 125°–170°), with a mean pronation of 73° (range, 55°–80°) and supination of 76.5° (range, 55°–90°).
At the time of final follow-up, the functional results according to categorical ratings of MEPI and Broberg–Morrey functional rating index were excellent in seven cases (Figure 4(a) to (d)), good in two cases, and fair in one case. The average MEPI score was 93.5 points (range, 70–100 points). The average Broberg–Morrey functional rating index was 94.1 points (range, 65–100 points). The average score on the DASH was 6.97 points (range, 0–24.2 points). Hand grip power of the injured side averaged 95.4% of the normal side, with the mean being 27.0 kg for injured arm and 28.3 for normal arm. There was no evidence of instability of the elbow clinically and radiographically in any patient.

A 67-year-old female suffered a complex elbow dislocation. (a) Preoperative images show a large defect of the radial head. (b) Fifty percent of the radial head was replaced with the iliac crest bone graft. (c) The radiographs at 9 years postoperatively show well-maintained radiocapitellar and ulnotrochear joint spaces, with no sign of arthritis. (d) Postoperative images demonstrated the range of motion.
In eight patients, the radiographs taken at the last follow-up demonstrated a well-maintained radiocapitellar and ulnotrochear joint spaces with no sign of arthritis. Grade 1 arthritis was observed in patient 4 who had the malreduction, and grade 2 in patient 3 who had a nonunion.
Discussion
The radial head is the important structure providing elbow stability. 2 The present study evaluated the reconstructive technique to restore the radial head in the complex elbow dislocation using the contoured iliac crest bone grafts.
In general, it is preferable to preserve the radial head with the ORIF when possible. 20 However, factors such as a large irreparable portion from significant comminution, the small size of fragments with limited available subchondral bone, impaction or deformation of the fragments, and bone loss can contribute to tenuous fixation and early failure. 2,21 The use of contoured iliac crest bone graft can convert multiple irreparable fragments to a large single fragment sufficient to accept the screw. Consequently, the total number of articular fragments is reduced to two or three, making the reduction and fixation simpler and more feasible. Ring et al. described a high incidence of complications in ORIF if more than three articular fragments were present. 22 The iliac bone graft also provides structural support and overcomes the bone loss. The graft, which contains osteogenic, osteoinductive, and osteoconductive properties, is likely to enhance the bone healing and possibly reduce the risk of nonunion. 23 In the present series, 9 out of 10 fractures healed completely. One patient had a nonunion in a piece of the original fracture fragment, but the bone graft had healed completely within the radial neck and remaining original fragment.
The contoured bone grafting technique assists the surgeon to restore the original radial head profile using both the irreparable portion and the remaining portion of the radial head as the templates. Defect in the AM quadrant of the radial head was found in all patients. This finding corresponds to a recent summary report of imaging in radial head fractures associated with the elbow instability. 14 We found that supinating the forearm can provide greater visibility of the defect; however, when the bone loss is located in the AM and PM quadrants, the exposure still might be inadequate. Further removal of the bone in the AL quadrant may be necessary to increase the working area in order to obtain the appropriate reduction and fixation.
Following the reconstruction with contoured bone graft, there was no evidence of recurrent instability in any patients. During a 3- to 9-year-follow-up period, the mean range of movement, MEPI, and DASH score were comparable with the results in the prior reports on the treatment of comminuted fractures with radial head arthroplasty. 24,25 Patients who had entire radial head fracture with malalignment occurring following the reconstruction had the worst outcome. The high incidence of implant removal was seen in the patients receiving plate fixation. The same evidence was observed in several other studies of plate fixation in the treatment of the radial head fracture. 18,26,27 The various incidents of the graft-site morbidity following the anterior iliac crest bone graft harvesting had been reported in the literature. In the present series, no donor site morbidity was revealed at the latest follow-up evaluation. 28,29
Interestingly, despite the lack of cartilage surface of the iliac crest bone graft, development of symptomatic osteoarthritis with pronounced functional impairment is not apparent in the present series. Radiographically, the radiocapitellar and ulnotrochear joint spaces were well maintained with no sign of arthritis in 8 of 10 patients at the midterm follow-up. Radiographic evidence of arthritic changes was observed in two patients. Similar findings were also documented when the autogenous iliac crest bone grafts have been used in the reconstruction of the articular bone defect of the glenoid and coronoid. 30 –32
Limitations of this study include small sample size and no comparison with other operative treatments. The dissimilar patterns of injury, consisting of the PL rotatory injury and posterior olecranon fracture dislocation, were combined. The different extent of osseous and ligamentous injury in each type may have significant effects on the outcome. Larger comparative studies with longer follow-up in the specific injury pattern are needed to confirm the value of this procedure.
Conclusion
Based on the findings in the present study, the use of the contoured iliac crest bone graft to repair the large noncontained defect of the radial head due to irreparable, deformed, or missing fracture fragments in complex elbow dislocation provided satisfactory results. This technique may be considered as an alternative option when more than 40% of the original radial head is still available for incorporation.
Footnotes
Acknowledgment
The authors thank Michele D. Spring, MD, of the Armed Forces Research Institute of Medical Sciences for editorial assistance.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplementary material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
