Implant Survivorship, Radiographic Evaluation, and Clinical Outcomes of Radial Head Arthroplasty Using The Depuy Synthes Radial Head Replacement System

Abstract

Background:

Radial head arthroplasty (RHA) implants are designed either to promote osseo-integration with an anatomic stem or function as a loose-fitting spacer. Anatomic implants have shown painful loosening leading to revision or removal, making the loose-fitting design an appealing alternative. This study evaluates short-term survivorship and clinical, radiographic, and patient reported outcomes following RHA using the DePuy Synthes monoblock loose fit radial head replacement system.

Methods:

Nineteen patients (mean age 54 SD 17 years), who underwent RHA were enrolled, with a mean follow-up of 14.9 SD 6.8 months. Fractures were classified using Mason and Regan-Morrey systems. Fifteen patients returned for in-person evaluation with x-rays, physician and physical therapist examination and patient reported outcome (PRO) questionnaires. Four remote participants completed PROs only.

Results:

The cohort consisted of 8 females and 11 males, with 10 patients sustaining associated elbow dislocations and 13 with concurrent coronoid fractures. No patient had the implant removed or revised. PROs indicated excellent functional outcomes: Mayo Elbow Performance Score of 90 SD 12, QuickDash of 13 SD 9 (minor disability) and EQ-5D of 0.77 SD 0.18 (mild disability). Grip strength in the operative limb was 90% SD 14% of the non-operative and range of motion was greater than 90% of the non-operative side. Heterotopic bone formation was noted in 7 patients but did not limit motion. Mild periprosthetic radiolucency around the stem was noted in 14 patients. Moderate capitellar erosion was noted in one patient with an Essex-Lopresti injury.

Conclusion:

The DePuy Synthes radial head replacement system demonstrated excellent shortterm outcomes in terms of survivorship, functional recovery and patient satisfaction. Although mild radiographic lucencies and heterotopic ossification were observed, they did not impact functional outcome. These findings support the safety and efficacy of this implant, though longer-term follow-up is necessary to assess its durability and performance.

Level of Evidence: IV

Introduction

Fractures of the radial head and/or neck comprise a third of all elbow fractures, with radial head fractures occurring in 20% or more of all traumatic elbow injuries.¹'³ The main goals for treatment of these injuries include restoration of joint stability, preservation of motion, and maintenance of the relative length of the radius.³ When operative treatment is indicated, either open reduction and internal fixation (ORIF) or radial head arthroplasty (RHA) can be employed to manage radial head fractures. RHA is the preferred treatment option for radial head fractures with three or more fragments, especially in the setting of complex elbow dislocations.^3–5

Radial head implants are generally designed either to promote osseointegration press-fit stem or to function as a loose-fitting spacer.⁴ Press-fit implants aim to replicate the native function of the radial head, featuring a stem designed to encourage osseointegration. However, these implants have been associated with painful loosening, leading to revision or removal rates ranging from 0% to 30%.^6–8 High failure rates secondary to loosening has resulted in several press-fit implants being withdrawn from the market.⁶

Loose fitting designs are intended to act as spacers, settling into an anatomic position during elbow motion without causing pain.^4,5 Laumonerie et al. (2021) found no cases of revision surgery for painful loosening in a systematic analysis of 146 cases of radial head arthroplasty using a modular loose-fitting prosthesis.⁹

Most patients experience much of their recovery in the first 12 months after surgery.^10,11 A minimum 3 year follow up was suggested to provide reasonable reproducibility of outcome results.¹² However, early follow up is crucial for assessing the safety of new implants and recognizing potential failures. Somerson and Matsen¹³ emphasized the importance of timely recognition of adverse results using the FDA Maude database. Several of the problems with radial head implants were evident early, but publication of the failures and removal from the market was delayed.

In 2019, DePuy Synthes introduced its monoblock loose fit implant, the Radial Head Replacement System, to the United States’ market. This study aims to evaluate the safety and effectiveness of this system in the short term by assessing implant survivorship, radiographic outcomes, clinical outcomes, and patient-reported outcomes in a cohort observed between six and 36 months postoperatively.

Methods:

Study Design and Population

This observational retrospective case series evaluated 26 patients who underwent RHA with the DePuy Synthes Radial Head Replacement System (West Chester, PA) by a single surgeon (DCF) between 2020 and 2023. Patients were included if they were 22 years or older and were between six- and 36-months post-surgery. The study team attempted to contact each patient up to four times by phone. Ultimately, 19 patients provided informed consent and were enrolled, 15 completed a one time in-person assessment and four opted for remote data collection as they lived out of the area. Seven eligible patients did not participate due to lack of response (6) or declining participation (1).

Operative Technique

All patients were evaluated with anterior and lateral radiographs and computed tomography (CT) scans when indicated. A lateral Kocher approach was utilized in all cases. The radial head fracture was inspected and indicated for replacement if there were more than 3 fragments present. The fragments were removed, and a saw was used to cut the radial neck at the appropriate length. Broches were used to determine stem size. The head size was measured by reconstructing the resected fragments. If the native head was between two sizes, the smaller size was chosen to minimize the risk of overstuffing the joint. A trial stem was inserted, and the length was evaluated using direct visualization and radiographic landmarks. Coronoid fractures, when present, were addressed based on the size of the fragment. Small tip fractures were either excised or managed non-surgically. For fractures involving the anterior capsule that were too small for screw fixation, capsular sutures were placed and secured through bone tunnels to the dorsal cortex of the ulna. Larger coronoid fractures were stabilized using posterior-to-anterior screw fixation. Once the final implant was placed, the lateral ligament complex was repaired to its anatomic origin using a suture anchor. Stability was tested and a medial repair was performed if needed. An ulnar nerve decompression was added if a medial repair was required or if the patient had pre-operative ulnar nerve symptoms. If there was residual instability after a medial and lateral repair was performed, an Internal Joint Stabilizer (IJS, Skeletal Dynamics, Miami, FL) was placed on the lateral side.

Implants

The Synthes Radial Head System is a smooth stem cobalt chrome monoblock implant. The implant offers three head diameters (19, 22, and 25 mm), three head heights (0, 3, and 6 mm), and four stem diameters (5.5, 6.5, 7.5, and 8.5 mm), resulting in a total of 30 different size combinations. The set includes radiolucent trials to allow visualization of fit and prevent overstuffing.

Post-Operative Protocol

The operative elbow was placed in a posterior splint for 7-10 days to allow the wounds to heal. The patients were then placed in a hinged brace and allowed full active motion. Physical therapy was routinely ordered. Patents with a lateral ligament reconstruction were instructed to perform flexion/extension exercises with the forearm in pronation and pronation/supination exercises with the elbow at 90 degrees of flexion for 6 weeks to protect the lateral ligament repair.

Data Collection

The radial head fracture was classified according to the Mason classification^14,15 and associated coronoid fractures were classified using the Regan Morrey clas-sification¹⁶ using injury radiographs and CT scans.

In-person participants were consented and administered outcome questionnaires, received two-view (anterior/posterior and lateral) x-rays of the affected limb, and were examined by a physician (KPB) and a licensed physical therapist (JME). Remote participants did not undergo x-rays or physical exams due to travel constraints, which limited their assessment to patient reported outcome questionnaires.

Implant Survivorship and Radiographic Outcomes

Survivorship was defined as the implant remaining in place for the entire follow-up period. Heterotopic ossification was identified according to its location and severity. Capitellar erosions were graded as none, mild, moderate or severe. Periprosthetic lucency was rated according to the system of Grewal as none, mild, moderate or severe.¹¹ Degenerative changes were rated using the Broberg and Morrey classification.¹⁷ Radial length was evaluated to determine the presence of gross over lengthening by evaluating the congruity of the medial ulnohumeral joint line.¹⁸

Clinical and Patient-Reported Outcomes

Demographic and injury data were abstracted from medical records. The physician (KPB) evaluated each patient for deformity, sensory and motor nerve function, and tenderness, while the physical therapist (JME) measured grip strength and range of motion using a hand-grip dynamometer and goniometer, respectively. Patient reported outcome measures included the Mayo Elbow Performance Score (MEPS), QuickDASH, and EuroQol-5D (EQ-5D) questionnaires.

Results:

Patient Demographics

The mean age of the 19 patients at the time of injury was 54 (SD = 17) years with a mean follow up of 14.8 (SD = 6.8) months. Patient demographics are documented in Table 1. Over half (53%) of the patients had radiographically documented dislocations which were reduced in the emergency department. Radial head and coronoid fracture classification, as well as associated ligament injuries are documented in Table 2. One patient was noted to have an Essex-Lopresti injury to the forearm. External fixators were not used in any cases.

Table 1.

Number of Patients Evaluated and Characteristics of Those Patients

Characteristic	n
Total	19
In-person	15
Remote	4
Gender (assigned at birth)
Male	11
Female	8
Smoking status (at follow-up)
Never smoker	17
Former smoker	1
Current smoker	1
Insurance type
Medicaid	5
OTJ	2
Medicare	3
Commercial	7
VA	2
Characteristic (n=19)	Mean (SD)	Range
Time to follow-up (months)	14.9 (6.8)	6.9-31.5
Age at time of surgery (yrs)	54 (17)	27-77
BMI at time of surgery	31 (6)	23-46

(n=19).

Table 2.

Injury and Surgical Information for Remote and In-Person Participants

Injury Information	n	Injury/Surgery Information	n
Surgical elbow		Broberg-Morrey OA grade
Left	9	Grade 0	16
Right	10	Grade I	2
Isolated		Grade II	1
No	14	Grade III	0
Yes	5	Lateral ligament injury
Associated dislocation		No	3
No	9	Yes	16
Yes	10	Repair	16
Mason classification		No repair	0
I	0	Medial ligament injury
II	2	No	13
III	7	Yes	6
IV	10	Repair	4
Coronoid fracture		No repair	2
No	6	Internal joint stabilizer
Yes	13	No	18
ORIF repair	3	Yes	1
Excision repair	6	Ulnar nerve decompression
Capsular repair	1	No	17
No repair	3	Yes	2
Regan Morrey coronoid fracture classification (n = 13)
I	6
II	5
III	2

(n=19).

Complications and Revision Procedures

There were no revision procedures for instability or implant failure. One patient was advised to have the IJS removed but elected to leave it in place. All implants remained in place at final follow-up.

Functional Outcomes

Patient reported outcome scores were obtained for all 19 patients (Table 3). The MEPS mean score was 90 (SD = 12, scale range 0-100, >90 indicating excellent performance and <60 indicating poor performance), with 17 patients reporting good or excellent performance and 2 reporting fair performance. No patient had poor performance. The mean QuickDash score was 13 (SD = 9, scale range 0-100, 0 indicating no disability) indicating minor disability and symptoms. The mean EQ-5D score was 0.77 (SD = 0.18, scale range 0-1, 1 indicating no disability) indicating mild disability.

Table 3.

Patient Reported Outcomes for all In-Person and Remote Participants

Outcome	Mean (SD)	Range
EQ-5D	0.77 (0.18)	0.39-0.94
QuickDASH	13 (9)	0-27
MEPS	90 (12)	65-100
MEPS Category	n
Excellent	7
Good	10
Fair	2
Poor	0

(n=19).

Physical examination outcomes were available for the 15 patients that returned for a clinical exam (Table 4). Grip strength in the operative extremity was 90% of that in the non-operative extremity. Flexion/extension range of motion was 14 - 142 degrees in the operative elbow and -4 to 145 degrees in the non-operative elbow. Pronation/supination range of motion was 70 - 66 degrees in the operative elbow and 72 - 72 degrees in the non-operative elbow. One patient reported mild pain over the radial forearm, one patient reported mild tenderness over the radial head and/or forearm, and three patients had residual ulnar nerve sensory changes.

Table 4.

Clinical Physical Examination Outcomes for In-Person Participants

Outcome	n
Tenderness
No	11
Yes	4
Olecranon	1
Ulnar nerve	1
Forearm	1
Radial head & Forearm	1
Deformity
No	14
Yes	1
Valgus	1
Varus	0
Numbness
No	9
Yes	6
Ulnar nerve	3
Median nerve	1
Lateral forearm	2
	Mean (SD)	Range
Grip strength (lbs)
Surgical	71 (22)	39-113
Non-surgical	80 (25)	45-128
% of non-surgical	90% (14)	66-122%
Maximum active elbow flexion (deg)
Surgical	142 (6)	130-151
Non-surgical	145 (5)	135-155
% of non-surgical	98% (4)	93-107%
Maximum active elbow extension (deg)
Surgical	14 (9)	-1-32
Non-surgical	-4 (7)	-18-3
% of non-surgical	90% (5)	82-96%
Maximum forearm supination (deg)
Surgical	66 (11)	32-80
Non-surgical	72 (7)	60-85
% of non-surgical	92% (20)	46-118%
Maximum forearm pronation (deg)
Surgical	70 (14)	32-89
Non-surgical	72 (5)	63-83
% of non-surgical	98% (23)	43-131%

(n = 15).

Radiographic Outcomes

Radiographic outcomes were evaluated for all 15 patients that returned for an examination and are described in Table 5. No hardware fractures, dislodgement, or peri-prosthetic fractures were found. Heterotopic bone was noted in 7 patients. The location is described in Table 5. None of the heterotopic bone was thought to limit range of motion of the elbow which was on average greater than 90% as compared to the non-surgical side. Moderate capitellar erosions were noted in one patient with an Essex-Lopresti injury of the forearm. Radiolucency around the stem was observed in 14 of the 15 patients and was mild in 9 cases. Radial length measurements showed no patients to have evidence of significant overstuffing of the elbow as judged by asymmetric widening of the medial radiocapitellar joint. Early evidence of osteoarthritis was noted in 3 patients (Table 2).

Table 5.

Radiographic Outcomes for In-Person Participants

Outcome	n
Heterotopic ossification
No	8
Yes	7
Radial neck	2
Radial neck and lateral ligament	1
Anterior capsule	2
Anterior and lateral neck	1
Lateral condyle	1
Capitella erosions
No	14
Yes	1
Mild	0
Moderate	1
Severe	0
Radiolucency
No	1
Yes	14
Mild	9
Moderate	4
Severe	1

(n = 15).

Discussion

This early term follow-up of the Synthes Radial Head Replacement System for unrepairable radial head fractures showed the implant to be safe and effective in restoring elbow joint stability and function. No revision surgeries were required for painful loosening or implant failure. Radiographic stem lucency and heterotopic bone were common, but nearly all patients had good or excellent objective and subjective outcome scores.

The patient reported outcome scores in this study are comparable with those in other studies for similar modular, smooth stem radial head replacements. In a systematic review of 5 previous studies reporting early and midterm outcomes following modular smooth stemmed radial head implants, Laumonerie et al. (2021) found a mean MEPS score of 87.7 (range from 83-91), similar to our reported score of 90.⁹ Two patients had a fair result, both scores were lowered by greater pain intensity. Chen et al. (2018) found a QuickDash score of 11.7 using the same modular implant, similar to our reported score of 13.¹⁹

Smooth stem radial head implants are meant to rotate within the radial shaft, thus lucencies are anticipated. Previous research often shows a high incidence of painless lucencies surrounding the stem. Five separate studies examining the short to medium-term results of a modular smooth-stemmed prosthesis found that the rate of radiographic lucencies varied between 7% and 80%.^10,11’^20,22 Longer term follow-up studies show minimal progression of the lucencies and no deterioration in outcome parameters.^10,23 Chen et al. (2018) noted a high percentage of mild residual forearm pain at midterm follow up after smooth stemmed implants.¹⁹ They reported the pain did not affect patient outcomes and was not correlated with the degree of lysis. Despite our excellent patient reported outcomes, we also noted mild forearm pain in a small percentage of our patients.

Overstuffing of the elbow occurs when an excessively long or overly wide prosthesis is implanted, leading to capitellar erosions, reduced motion range, and elbow joint degeneration.²⁴ The Synthes Radial Head System features a monoblock design with 30 different size combinations. In contrast, comparable smooth stem modular systems offer many more options,¹⁸ raising concerns that limited sizing choices could result in mis-sizing the implant. Accurate measurements of radial length using postoperative radiographs are challenging.²⁴ We opted to assess the congruity of the medial ulnohumeral joint to detect any significant overlengthening of the radius. We found no instances of elbow overstuffing, suggesting that the limited number of available implants did not pose a sizing issue.

The affected limb exhibited a slightly reduced range of motion compared to the unaffected limb; however, these differences were minimal and unlikely to impact daily activities.^25,26 Prior research indicates that range of motion stabilizes within 12 months and does not typically change thereafter.^10,11 Strength assessments revealed minor impairments relative to the contralateral limb, consistent with previous findings.^10,11,20

Despite the positive findings, our study has several limitations. The relatively small sample size and the retrospective nature of the study may introduce biases. Additionally, the follow-up period is relatively short for assessing long-term outcomes and potential late complications. Laumonerie et al. (2017), in a study that included both smooth and bone ingrowth stems, suggested that the minimum follow up for radial head implants should be at least 3 years to adequately capture a majority of the reoperations.¹² However, midterm follow-up studies using only smooth stemmed implants showed no increased failures after 1 year,^10,20 suggesting shorter follow-ups may be adequate. It is, however, vitally important to perform early follow-up studies on new implants to determine the safety and efficacy of the implant. This is especially true for radial head implants given the number of devices that are removed from the market for failures.¹³

In conclusion, the Synthes Radial Head Replacement System demonstrates excellent short-term outcomes in terms of implant survivorship, functional recovery, and patient satisfaction. Our results are congruent with several other studies using similar smooth stemmed implants. While the results are promising, longer-term studies are necessary to fully understand the durability and effectiveness of this implant.