Radial Head Resection versus Arthroplasty in Unrepairable Comminuted Fractures Mason Type III and Type IV: A Systematic Review

Unrepairable comminuted fractures of the radial head Mason type III or type IV have poor outcomes when treated by open reduction and internal fixation. Radial head resection has been proposed as good option for surgical treatment, while in the last decades, the development of technology and design in radial head prosthesis has increased efficacy in prosthetic replacement. The present review was conducted to determine the best surgical treatment for comminuted radial head when ORIF is not possible. Better outcomes are reported for radial head arthroplasty in terms of elbow stability, range of motion, pain, and fewer complications compared to radial head excision. Nevertheless, radial head resection still can be considered an option of treatment in isolated radial head fractures with no associated ligament injuries lesion of ligaments or in case of older patients with low demanding function.


Introduction
Surgical treatment for comminuted and unrepairable fractures of the radial head may be challenging. These types of fractures are often associated with multiple ligamentous injuries amounting to elbow instability. Radial head resection has been proposed as good option for surgical treatment, while in the last decades, the development of technology and design in radial head prosthesis has increased efficacy in prosthetic replacement.
The radial head is a secondary valgus stabilizer of the joint and it is involved in transmission of axial force load through the elbow during flexion [1]. It is also a varus and external rotatory constrainer [2]. Comminuted radial head fractures Mason type III and type IV are commonly associated with other injures of the elbow as capitellum and coronoid fractures and/or ligaments disruption, both medial and lateral ligaments and interosseus membrane [3][4][5][6]. Primary goal in surgical treatment is to restore elbow stability in order to preserve the complex physiologic elbow kinematics. In this respect, medial collateral ligament is the primary constrainer in valgus stress. Radial head contributes secondarily to valgus stability [1,7] and its preservation is mandatory in case of fractures that involve soft tissue and ligaments to avoid chronic instability. Many authors have described serious complications in case of resection of the radial head such as proximal migration of radius and longitudinal instability, humeroulnar osteoarthritis [2,[7][8][9], decrease in grip strength, cubitus valgus, and ulnar neuropathy [10,11]. Therefore, radial head arthroplasty has obtained a large consensus in orthopaedic surgeons as primary option of treatment in fractures Mason types III and IV. It allows an anatomical reconstruction and it maintains stability and physiologic kinematics of the elbow if associated with ligament reconstruction. However, oversizing or overstuffing of radial head prosthesis, malpositioning, and aseptic mobilization may lead to a high rate of complications and failure of this surgical procedure. Recent reviews of literature [10,12]  radial head fractures Mason type III or type IV to assess results and indications for radial head replacement or resection.

Materials and Methods
We searched in PubMed electronic database the words (radial head fractures) AND ((artrhoplasty) OR (prosthesis)) AND ((resection) OR (excision)). The guidelines for preferred reporting items for systematic reviews and meta-analysis (PRISMA) were used ( Figure 1). We selected articles of the last 20 years, from 1998 to December 2017. We created an Excel database for collecting data extracted from articles in English language, selecting papers with series of 10 or more patients. Exclusion criteria were articles written in other languages, case reports or reviews, cadaveric or instrumentals studies, clinical studies with no standard questionnaires or scores, and studies in which posttraumatic outcomes were not separated from primary reconstruction of the radial head. We extracted relevant data from the selected articles: type of study, number of patients, age, follow-up, type of surgery performed, clinical results (ROM, DASH score, MEPS score, and VAS), and radiographic results.

Results
The database search identified 152 potentially relevant articles. Abstracts have been analyzed following inclusion and exclusion criteria and a total of 29 papers were selected for the present review. Most of retrospective studies on metal radial head prosthesis have been published in the last ten years in comparison to a lack of studies for radial head excision in the last two decades. Moreover, few articles on comparison of the two surgical techniques have been found. Because of heterogeneity in level of evidence, surgical technique, type of implants, and rehabilitation protocol, we did not perform statistical data analysis. Articles selected are reported in Table 1.

Discussion
From our review of literature clinical results for radial head replacement are reported in Table 2. Most of retrospective studies involve modular monopolar or bipolar prosthesis implanted for irreparable Mason type III or type IV fractures. For most of authors, mid term follow-up has shown satisfactory results in range of motion recovery (average   [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32]. A certain loss of grip strength compared to contralateral side is often described (average loss of strength: 10% respect to the contralateral side). Authors highlight the importance of ligament reconstruction in case of associated injuries. Intraoperative assessment of stability and acute repair of torn ligaments is mandatory for a successful procedure.
Most common radiological modifications include osteoarthritic changes of ulnohumeral joint, capitellum wear for oversizing of radial head prosthesis, periarticular heterotopic ossifications, and radiolucency lines around the stem. Some modifications in radiographic appearance seem to not correlate directly with clinical symptoms: bone resorption around the prosthesis does not correlate with loosening of the prosthesis and does not affect clinical scores. Marsh [21] reports favorable clinical outcomes from short to long follow-up despite a high evidence of radiolucency around the stem and arthritis in his series. Gauci [20] has found no association between neck bone resorption and postoperative symptoms.
Complications (Table 3) described in radial head replacement are in common in almost all the papers: aseptic mobilization of the stem, overstuffing, erosion of the capitellum, osteoarthritis, and heterotopic ossification clinically arising with lateral elbow pain or loss of motion, and posterior subluxation for undersizing. Rare temporary ulnar and radial nerve sensory neuropathies are reported. Though, few papers seem to discourage radial head arthroplasty. Moro [31] reports mild to moderate impairment of ROM and pain for both elbow and wrist in patients treated with a metal radial head implant. Laumonerie [16] describes unsatisfactory result from bipolar radial head prosthesis because of malposition in varus and valgus and oversizing leading to a high rate of reintervention during the three first months after implantation. Flinkkila [23] reports poor results from press    fit radial head prosthesis due to a high rate of loosening. Difficulties on technique of implantation are described by Ashwood [30] for mono-block prosthesis.
Retrospective studies on radial head resection have a longer follow-up and clinical and radiological results are reported in Table 4 [33][34][35][36][37][38][39][40][41][42]. Clinical and radiological complications at long-term follow-up are reported (Table 5). Clinical results show good outcomes in Mayo Elbow Performance Scores (MEPS, from 79 to 100) and Disabilities for Arm Shoulder and Harm scores (DASH, from 4 to 15), a satisfactory recovery of elbow range of motion (average flexion-extension arc of motion: 120 ∘ ) and low scores in VAS scale (from 0 to 4.6). However common complications of this surgical procedure involve ulnohumeral joint due to an higher load compression force that leads to degenerative changes and progressive worsening of cubitus valgus associated to ulnar nerve neuropathy and UCL elongation leading to chronic elbow instability [3,4]. Moreover, proximal migration of radius is often assessed (80% of papers), complications that involve DRUJ impairment leading to wrist pain hand strength reduction and distal radio-ulnar arthritis. Preoperative or intraoperative setting of elbow stability and correction of ligaments injuries is mandatory to avoid early complications. Despite of complications, many authors approve the surgical technique due to good outcomes in mid to large term. Yalcinkaya [36] found no significant correlation between radiological degenerative modifications in elbow and outcomes of clinical scores in patients treated by radial head resection. Antuna [38] reports good clinical results in a large series of patients less than forty years old treated by radial head excision after a mean follow-up of 25 years. Herbertsson [39] reports worst outcomes in excision for Mason type IV fractures although delayed radial arthroplasty is suggested for pain relief and preservation of range of motion in case of failure of other treatments. Finally, few papers compare radial head resection and radial head arthroplasty [34,35] where authors recommend resection as primary option of treatment because of a lack of statistical clinical differences between the two surgical procedures, in case of isolated radial head fractures not associated to ligaments injuries. Nestorson [33] did not found better outcomes by using a press fit radial head prosthesis in Mason type IV fractures and he reports similar functional results after radial head resection despite more osteoarthritic changes. Lopiz [34] suggests resection as a good option of

Ikeda M 11
Mean ulnar variance +1.6mm Mean increase in carrying angle 8 ∘ Mild to severe degenerative arthritis in all patients Jansen RP 18 ROM limitations 11 Degenerative changes 7 increase of cubitus valgus, 7 periarticular ossification, 7 osteoporosis of capitellum, 12 proximal radius migration (from 1 to 5 mm) treatment when ORIF is not possible, reporting a higher rate of complications in the group of patient treated by radial head arthroplasty.

Conclusion
From our review of literature almost all the retrospective studies on radial head arthroplasty report convincing results in terms of elbow stability, range of motion, and pain. Nevertheless, papers on radial head resection report good clinical outcomes in isolated radial head resection with no associated ligament injuries. Few papers compare the two techniques with no substantial differences in terms of clinical outcomes at medium and long follow-up.