Radial head replacement – A comprehensive review

Abstract

Background

Radial head fractures represent 1–4% of all adult fractures and 1/3rd of elbow fractures (Kaas et al., 2010). Radial head is an important secondary stabilier of the elbow. The aim of the treatment of radial head fractures is to achieve stability and good functional outcome. Radial head replacement (RHR) is indicated when robust reconstruction and fixation of the fracture fragments is not possible.

Methods

This article explores history and evolution, anatomical considerations, biomechanics, implant designs, indications, surgical outcomes and controversies in radial head replacement based on current evidence.

Results

There is a wide range of RHR designs available without conclusive evidence to support the superiority of one design over the other. Ranges of motion, functional outcomes and complication were comparable among different designs with a high incidence of complications reported in many studies.

Conclusion

RHR remains a good option in unreconstructible radial head fractures, with potential to regain excellent function. It is imperative to continue with the quest to innovate and improve on current designs, to reduce complications in the long term.

1. History & evolution

Radial head fractures represent 1–4% of all adult fractures and 1/3rd of elbow fractures.¹ Speed (1941) used a ferrule cap for RHR which was designed with the cast of a resected radial head.² Carr (1951) surgically treated radial head fractures using Kellog-Speed metallic cap with good outcomes.³ Cherry (1953) published a case series of acrylic RHR describing its use in preventing proximal displacement of the radius in certain fracture patterns. They were unable to draw a definitive conclusion regarding its role in all situations.⁴

In 1969, Silicon RHR was introduced by Swanson for comminuted fractures and for failed radial head resections. Until the 80's, the most popular radial head implant was the silicon implant.⁵ In 1981 Morrey reported a case series using silastic radial head implants with satisfactory results in an acute setting. He also recognised the limited role of silastic implants. Further reports of early implant failure and reactive synovitis with silastic implants began to emerge and their use gradually declined.^6,7

Harrington et al. preferred metal radial head replacements as it was radiographically visualised with ease and felt it was more resistant to mechanical failure. Their early results (1981) and long-term outcomes reported good outcomes with titanium RHR.⁸

Vitallium radial head prosthesis (Howmedica, UK) emerged in the 80s and outcomes were published in 1993. Cadaver studies were conducted by compression testing at 90° elbow flexion and axial loading of radius. Compared to vitallium, silicon implants deformed easily, allowing proximal radius migration under load. In a case series (36 patients, mean follow-up 4.5 years) satisfactory outcomes were reported following vitallium RHR.⁹

A floating design RHR with integrated articulation allowing the radial head to change position with elbow movement was introduced in the 80s. Judet et all reported early results in 1996 confirming good performance with improved functional scores.¹⁰

Between 1988 and 1995, frozen-allograft RHR was used for the treatment of radius which had proximally migrated. The authors remained cautiously optimistic about the indications and outcomes. No further reports are available to date.¹¹

Since 2000, several anatomical studies emerged looking at the scientific rationale behind various RHR, taking into account variations in radial head anatomy. The radial neck and diaphysis angle were studied in detail. Anatomical RHR designs began to emerge. However, even with anatomically designed prostheses, accurate placement of the implant is crucial to achieving anatomical alignment.12, 13, 14, 15

Radial head replacement has evolved considerably. A range of designs are available which could be monopolar, bipolar, modular, monoblock, loose-fit stems or fixed stems. Fixed stems can be press-fit or cemented. The head design could be anatomic or non-anatomic. The bearing surfaces of RHR can be metallic alloys, pyrocarbon or high-molecular-weight polyethylene.

2. Anatomical considerations & biomechanics

The elbow is a complex joint with three articulations (radio-capitellar, proximal radioulnar, ulnohumeral). Stabilisers are static and dynamic. Primary static elbow stabilisers are the anterior bundle of the medial collateral ligament, lateral ulna collateral ligament (LUCL) and ulnohumeral articulation. Secondary static stabilisers include radio-capitellar articulation, anterior capsule, common extensor and common flexor tendons. Dynamic stabilisers are the musculotendinous units crossing the elbow joint.

The radial collateral ligament, LUCL and annular ligament constitute the lateral ligament complex. LUCL is the main stabilizer of the elbow to varus loads. The medial ligament complex consists of anterior, posterior and transverse bundles. The anterior bundle resists valgus loads.

The head of the radius is elliptical in shape with a concave articular surface with dimensions varying between individuals. The head articulates with the convex shaped capitellum.

In cadaver studies, the radial head diameter ranged between 22.6 mm and 24.3 mm. The articular surface depth measured 2.4 mm and the mean offset of the head from the neck was 4.2 mm. The average proximal diaphysis-neck angle measured 17°. A cam effect was created due to the variable offset of the head from the neck.^13,16, 17, 18, 19

The radial head is vital in providing stability in axial load, valgus and external rotation.²⁰ Isometric elbow flexion under resistance can generate forces as high as 4 times the body weight across the elbow. The radio capitellar joint takes up to 60% of these loads.²¹

An invitro study looking at the transmission of forces through elbow in extension demonstrated that 60% of the forces were transmitted through the radiocapitellar and only 40% through the ulnohumeral joint.²² During valgus stress, 30% of the resistance is transmitted to the radial head²³

Radial head replacement designs should take all of the above anatomical and biomechanical factors into account.

3. Implant design

There is a wide range of RHR designs available in the market.

3.1. Monopolar RHR

This commonly used implant is available in mono block or modular designs. The stem designs are fixed stem or intentionally loose fit. The fixed stem can be press fit or cemented. In order to achieve a tight fit in the intramedullary canal, press-fit stems are either plasma sprayed or grit blasted.23, 24, 25 The smooth stem option can either be used as size for size fit with the canal or over-reaming the canal to intentionally produce a loose fit.

The radial head designs can be anatomical or non-anatomical.

The understanding of radio capitellar biomechanics improved significantly following several biomechanical studies. It was hypothesised that monoblock designs did not replicate native radiocapitellar biomechanics. An anatomical design was introduced with the objective to distribute a low contact pressures more evenly than non-anatomic implants.

3.2. Bipolar RHR

Taking into account the complex anatomy and biomechanics of the radiocapitellar articulation (shape of the radial head with offset of 10°–15° at the radial neck), bipolar RHR was introduced to improve joint congruency during range of movement, as it was thought mono block designs did not achieve this.

The advocates of the bipolar designs and loose fitting monopolar smooth stems designs believe that the RHR aligns itself within the radiocapitellar joint and the PRUJ and compensates for the lack of anatomical shape.

Overall, the rationale for the various designs are as follows.

• 1.Anatomical designs restore normal anatomy and replicate the native radial head.
• 2.Bipolar designs allow a degree of rotation between the radial head and neck, with a 10–15° movement in all planes, adapting to the anatomy and biomechanical requirements.
• 3.
  Loose fitting stems self-centre, thereby improving congruency in the

  • radiocapitellar joint.

4. Surgical indications

There is strong evidence to suggest that radial head fractures with three or fewer fragments can be managed successfully with ORIF.²⁶ Reconstructing the radial head will restore normal elbow kinematics.

Indications for RHR include.

• 1.Radial head fracture with more than three fragments
• 2.Terrible triad injuries, Monteggia variant fracture dislocations, transolecranon fracture dislocations when the radial head is unreconstructable
• 3.Symptomatic nonunion or malunion. Pain from joint incongruity requiring RHR is not unusual.
• 4.Radiocapitellar arthritis –Treating with RHR only is an option. However, RHR along with resurfacing the capitellum (Radio Capitellar Replacement) has been a recent development.²⁷ Indications, surgical algorithm and outcomes are not well established.

5. Surgical approach

The surgical approach for radial head arthroplasty could be Kaplan²⁸ or Kochers.²⁹ Boyd's³⁰ approach is often used during surgery for Monteggia variant and trans-olecranon fracture dislocation.

In Kaplan approach (Fig. 1) the plane between EDC and ECRB is developed, providing excellent visualisation of proximal radius and coronoid.

Fig. 1.

Radial head replacement - Kaplan approach.

In Kocher's approach, the plane is developed between anconeus and ECU. The Modified Kocher's approach preserves the LUCL by anteriorly elevating the anconeous.

In Boyd's approach, the incision is posterior and the forearm fascia is incised and a sub periosteal dissection of the anconeus is performed to access the radial head. The original approach described by Boyd in 1940 and has been modified over the years.³⁰

While performing RHR, attention to radial head diameter, length, alignment and tracking is crucial.

Radial head Diameter: The resected radial head fragments are used to estimate the head diameter. Incorrect diameter will result in poor tracking.

Radial length: Understuffing or overstuffing of the radio capitellar joint by even 2.5 mm could alter elbow kinematics and radiocapitellar contact pressure.^31,32 The radial head should align with the lateral coronoid, lesser sigmoid notch. Image intensifier is used to establish parallel lines at the medial and lateral joint surface (Fig. 2).

Fig. 2.

Intraoperative radial head length assessment.

Alignment and tracking:

The axis of the forearm lies between the ulna fovea to the centre of the radial. Restoration of this axis is paramount for restoring the normal kinematics.³³ Radial neck cut and the implant is aligned perpendicular to this axis.

5.1. Outcomes and complications

There are numerous studies looking at outcomes and complications following RHR using various types of implants from centres across the world. Overall, the range of movements and functional outcomes are reported to range from good to excellent but with relatively high incidence of complications.

5.2. Anatomical press fit designs

This design was introduced to reproduce the normal architecture of the radial head. J.C. Levy et al. reported the outcomes for 19 patients with anatomical press fit RHR (30 months average follow-up). They reported an excellent range of movements and good functional outcome scores, but a 40% incidence of radiolucency in the range of 2–3 mm. Stress shielding was noted in 67%. Press fit RHR in fractures associated with ligament injury showed higher rates of loosening which warranted further investigation.³⁴

Long-term survival rates of monopolar screw fix prosthesis and a monopolar loose fit prosthesis were reported by Schnetze et al. Although the overall functional scores and range of motion were satisfactory, there was a 37% complication rate and a 28% rate of revision (9.5-year median follow-up). At 18 years, the RHR survival was 75.1%. Most implant failures were noted in the first year.³⁵

Marsh et al. reviewed their 8-year experience with 55 RHR with smooth modular stem and metallic radial head implant for unreconstructible fractures. The Elbow range of movement and strength were significantly reduced when compared with the contralateral elbow. Stem lucencies were noted in 45%, ulnohumeral arthritis in 38%, heterotopic ossification in 36%, and radioulnar synostosis in one patient. There were no cases of revision surgery.³⁶

Chen et al. described their experiences with 32 loose-fitting stem RHR where none underwent revision or implant removal (mean follow-up 8.9 years).³⁷

Mid-term results for 77 modular bipolar RHR of 4 different models were reported by Laumonerie et al. The follow-up period was 74 months. The average MEPI score was 90.2. 40 complications were encountered with 38.9% re-operations. 24.7% of implants were removed and the mean time of removal was 21 months. The RHR survival rate was 60.8% at ten years.³⁸

Popovic et al. reported 55 patients who underwent RHR with a cemented bipolar prosthesis. There were no revisions at 8.4 years.³⁹

5.3. Controversies

Accurate sizing of radial head is crucial for a good clinical outcome. A number of parameters can be useful intraoperatively to guide decision making-the resected radial head, sigmoid notch, lateral ulnohumeral joint surface, posterior lateral synovial fold.

Cadaver studies looking at the resected radial head versus sigmoid notch suggest that the sigmoid notch is an unreliable reference point for implant diameter sizing.⁴⁰

A CT scan-based study (Doornberg et al.) showed that the radial head was 0.9 mm (average) proximal to the proximal aspect of the lesser sigmoid notch. There were significant differences between patients with some being proximal and others distal.⁴¹

Van Riet et al. examined cadaver specimens. For sizing the radial head, they recommend measuring the length from the proximal radial neck to the proximal lesser sigmoid notch.⁴²

Frank et al. recommend using the lateral ulnohumeral joint surfaces as a guide when implanting RHR. Overstuffing of the radial head implant by greater than 2 mm resulted in obvious gapping at the lateral joint surface.⁴³ However, Rowland et al., in a X ray-based study demonstrated that the lateral joint space is not a reliable marker of over stuffing as it was wider than the medial.⁴⁴

Another proposed landmark is the synovial fold at the radiocapitellar joint. This structure is present in more than 85% of patients and is 0.8–1.5 mm proximal to the radial head. The prosthesis should not be proximal to this fold.⁴⁵

5.4. Stem fit

Debate exists over the optimal fixation of RHR. Various implant designs include smooth stems with either an over-reamed loose fit or a size-for-size fit, cemented stem and press fit ingrowth stem.

The effect of stem fit on the contact biomechanics of the radiocapitellar joint has been studied using finite element models to evaluate whether the loose fit is better than rigidly fixed stems. A study by Szmit et al. suggested that 1–2 mm over-reamed smooth stem provided ideal contact biomechanics and probably functioned like a bipolar implant.⁴⁶

Studies comparing the effect of implant design on the concavity compression stability of the radiocapitellar joint concluded that anatomical designs were more effective than circular RHR in providing stability.⁴⁷

Similar studies comparing native, monopolar, bipolar heads with or without soft tissue presence found that bipolar implants were highly dependent on soft tissue integrity to achieve radio capitellar stability and anatomical designs achieve this by concavity compression.^48,49

5.5. Designs and survivorship

A study on outcomes of three implants (total 114 RHR) with and without cement fixation reported 85% survivorship of the implant, revision free period of 10 years. The designs used were 1. nonanatomic smooth stem design (n = 60) 2. nonanatomic design with grit-blasted stem, 3. curved stem design (n = 21), 4. anatomic design with straight stem which is grit-blasted (n = 33). There were no survivorship differences between cemented and uncemented stems. Stress shielding, implant tilting and loosening were seen commonly.⁵⁰

Study comparing smooth and porous stem designs (6.3 years average follow up) showed that porous implants were more likely to show osteolysis, loss of elbow flexion and overstuffing. Functional scores were similar in both groups.⁵¹

Sershon et al. reported 10-year follow up of bipolar smooth cementless stem RHR. The long-term survival was 97%. There was evidence of non-progressive radiolucency, but this did not correspond with poor clinical results.⁵²

Berschback et al. compared bipolar smooth stem RHR with monopolar press-fit designs. Short and mid-term outcomes were similar. They warned about osteolysis and proximal radius bone loss in press-fit designs.⁵³

Songy et al. could not identify similar osteolysis and loosening in their study on monopolar press fit RHR.⁵⁰

6. Bearing surface

Commonly used RHR bearing surfaces include metal alloys, pyrocarbon and high molecular weight polyethylene.

Watts et al. reported early results of anatomic press fit short stem RHR with pyrocarbon implant. The survival rate was 94% with good functional ROM in 76% at 17 months. 11% of implants were revised at an average of 38.6 months after implantation.⁵⁴

Medium term results of pyrocarbon RHR in 65 patients over a 7-year period showed cortical resorption around the neck in 92% with no mechanical failure.⁵⁵

Long term results (9 years) of modular design with a titanium stem and pyrocarbon head showed satisfactory outcomes. However, there was radial neck osteolysis without progression to stem loosening or failure.⁵⁶

In a systematic review of monopolar and bipolar RHR compiling the results from 9 studies and 591 patients. There were 365 monopolar, 226 bipolar designs and ranges of motion and complication were comparable in both groups. Revision and implant removal rates were comparable between the implants.⁵⁷

Chen et al. suggested that heterotrophic ossification was more pronounced in bipolar RHR than unipolar RHR.⁵⁸

Agyeman et al. published a systematic review comparing loose fitting smooth stems versus fixed stem (cemented or press-fit). There was no difference in outcomes, but they reported higher risk of complications in the fixed-stem group.⁵⁹

Heijink et al. in a systematic review of 30 publications and 727 patients (1940–2015) concluded that there was no evidence supporting one implant over the other but silicone prostheses were found to be biomechanically insufficient. The implants reviewed in this article include both monopolar (70%) and bipolar (30%) designs. 32% were press fit, 32% intentionally loose, 21% cemented and 15% were fixed implant with expandable stem.⁶⁰

This study looked at implants made of various materials including cobalt-chromium (70%), pyrocarbon 15%, titanium (9%), and Vitallium (6%).^60. The type of fixation, implant design and material used had no significant effect on the functional outcome or the revision rate. Revision rate was 8% in both groups. The complications reported were loosening (23%), overstuffing (20%), subluxation (18%), stiffness (14%), lateral elbow pain (13%), dissociation of the prosthesis(5%), infection(4%), malposition(2%) and for periprosthetic fracture (2%).

Summary.

• ⁃RHR remains a good option in unreconstructible radial head fractures, with the potential to regain excellent function.
• ⁃It is vital for the surgeon to have a good understanding of elbow biomechanics, knowledge of various implant options and the pros and cons of each.
• ⁃Surgical Tips: The radial neck cut should be perpendicular to the axis of the forearm. The resected radial head fragments are assembled to estimate the RHR diameter. When choosing the definitive radial head implant, 2 mm is deducted from this measurement to account for the articular surface thickness. During the trial of the implant, ensure that the radial head aligns with the lateral coronoid, lesser sigmoid notch. Another useful landmark is the synovial fold and the RHR should not be proximal to this fold (Fig. 3). An intraoperative Image intensifier should be used to ensure that medial and lateral joint surfaces are parallel.

Fig. 3.

Synovial fold.

• ⁃It is prudent to counsel the patient regarding outcomes and complications.
• ⁃The radiological complications and revision rate following RHR is worrying, with a high incidence of osteolysis around the RHR stem in most studies for all implant types, with a higher incidence in press fit RHR.
• ⁃The hypothesis that monopolar RHR is more superior in restoring stability compared to bipolar prosthesis has not been conclusively proven in comparative clinical studies.
• ⁃There is a lack of conclusive evidence in the literature to support the use of one RHR implant type over the other

6.1. Future directions

Further evaluation of RHR in high quality randomised multicentre studies could help answer some of the current concerns.

To continue with research in this area to further refine implant designs, fixation methods to achieve a highly functioning RHR with low complications and revision rates;

Funding/sponsorship

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Informed consent (patient/guardian), mandatory only for case reports/clinical images

Consent obtained for clinical images.

Institutional Ethical Committee Approval (for all human studies)

This is a review article, and no patients were involved. Ethical Committee Approval was not required.

Authors contribution

David S Thyagarajan: Methodology, Literature search, investigation, Conceptualization, Writing- Original draft preparation, Visualisation, Formal analysis, Writing- Reviewing and Editing.

Declaration of competing interest

None.