Clinical outcomes and complications following primary total elbow arthroplasty using the Latitude prosthesis

Abstract

Background

The Latitude total elbow arthroplasty (TEA) is an implant with limited published data on its performance and outcomes. The aim of this study was to report the short-term outcomes of the Latitude TEA as well as to describe the radiographic outcomes and complications.

Methods

The Latitude was implanted in 20 patients (23 elbows) in a linked configuration. Patients were recalled to clinic for the assessment of their range-of-motion and compared to preoperative values. Administration of functional outcome measures was also performed.

Results

Mean follow-up was 4.7 years (range, 1 to 7.5 years) with four elbows requiring revision. The flexion–extension arc improved from 86.6 to 101.3 (range, 76 to 126) postoperatively (p = 0.04). The average Disabilities of the Arm, Shoulder, and Hand score was 28.1 (range, 5.8 to 50.4) and the average Mayo Elbow Performance Score was 89.6 (range, 76 to 100), with 83% of elbows scoring in the good or excellent range. Radiolucencies were detected in 60% of patients and 31% of these lucencies progressed in size at the time of follow-up.

Conclusions

The Latitude prosthesis provides patients with favorable clinical outcomes with improvements in their range-of-motion and a complication rate comparable to other elbow arthroplasty implants.

Introduction

Total elbow arthroplasty (TEA) is a procedure to treat end-stage arthritis of the elbow in low-demand patients with disabling pain and stiffness, with or without instability. End-stage rheumatoid arthritis (RA) has traditionally been the most common indication for TEA.¹ Other indications include primary or post-traumatic arthritis and complex or comminuted distal humeral fractures in elderly patients with poor bone quality^2,3 and distal humeral non-unions. Initial TEA designs were fully constrained and often failed due to high force transmission at the bone–cement interface secondary to the linked design, especially in younger, more active patients.⁴ In response, unlinked TEA devices were developed that more accurately reconstitute the ulnohumeral articulation to reduce bone–cement force transmission and polyethylene wear; however, such implants are contraindicated when significant osseous or ligamentous deficiency exists.⁴

Clinical outcomes, including pain relief and improvement in range-of-motion (ROM), have been favorable following primary TEA in appropriately selected patients; however, complication rates following such procedures remain high. In a systematic review of 73 studies, Welsink et al.⁵ found a complication rate of 11% to 38% with aseptic loosening being the most common complication (7%). Several different total elbow implant designs exist with few studies that directly compare these implant designs. A study by Little et al.⁶ compared three contemporary TEA implant designs and found similar improvements in ROM and survivorship between the different implants. The five-year survival rates when performed in patients with end-stage RA, ranged from 81% to 93%.

The Latitude TEA (Tornier®, Stafford, Texas) is a total elbow design that permits conversion to either a linked or an unlinked configuration without requiring the removal of well-fixed components. There have been few published clinical reports that focus on the Latitude prosthesis in primary TEA since it has been available for clinical use in 2001. Burkhart et al.⁷ reported favorable clinical outcomes using the Latitude prosthesis in a cohort of 15 patients with primarily post-traumatic arthritis. Additionally, Wagener et al.⁸ reported promising results after using the Latitude TEA arthroplasty in a cohort of 63 patients.

The primary purpose of this study was to report the short-term clinical and functional outcomes of patients who underwent primary TEA using the Latitude prosthesis. The secondary objective of this study was to describe the radiographic outcomes and complications.

Methods

This study was approved by the Conjoint Health Research Ethics Board. Consecutive patients who underwent a primary TEA using the Latitude implant from 2006 to 2013 were considered for inclusion in this study. There were 20 patients who received 23 TEAs (Figure 1, Table 1). All procedures were performed by the senior author (KAH). Indications for the procedure included RA (n = 13 patients, 16 elbows), post-traumatic arthritis (n = 4 patients, 4 elbows), elbow fracture-dislocation (n = 1 patient, 1 elbow), distal humerus non-union (n = 1 patient, 1 elbow), and fibrous dysplasia (n = 1 patient, 1 elbow).

Figure 1.

Summary of the number of patients and elbows contributing to the various outcome measures.

Table 1.

Summary of patients.

Patient	Age	Gender	Year	Side	Indication	Implant Type	Radial head procedure	Complication(s)	X-ray > 2 yrs	DASH/ MAYO	Previous Operation(s)
1	60	M	2006	L	PTA	Latitude	Resected	Aseptic loosening ulnar component (2011)	Y	N	1) ORIF radial head, LUCL repair for terrible triad (2004) 2) Radial head replacement, LUCL reconstruction for recurrent instability (2005)
2	30	F	2007	L	RA	Latitude	Replaced	Humeral spool set screw back-out (2015)	Y	Y	Elbow arthroscopy and synovectomy (2005)
3	74	F	2007	R	RA	Latitude	Replaced		Y	N	Open synovectomy (2006)
	76		2009	L	RA	Latitude	Replaced		Y	N
4	70	F	2008	L	DHN	Latitude	Replaced	Grade I open periprosthetic humeral fracture (Mayo B2) (2009)	Y	N	1) ORIF and revision L proximal ulna x 4 (2002-2003) 2)  ORIF L distal humerus (2006) and revision (2007)
5	69	F	2008	R	PTA	Latitude	Replaced		Y	Y	ORIF distal humerus (2007)
6	51	M	2009	L	PTA	Latitude	Replaced / Revised	1) Dissociated radial head revised (2012) 2) Aseptic loosening ulnar component, loosening of humeral spool, radial head dissociation (2014)	Y	Y	1) ORIF L elbow (1975) 2) Elbow debridement (2006 & 2007) 3) Ulnar nerve transposition (2008)
7	63	F	2009	L	RA	Latitude	Replaced	1) Sterile bursa resection (2013) 2) I&D for PJI, repeat bursectomy x2 (2013) 3) Excision of implants for PJI (2013)	Y	N
8	47	F	2009	L	FD	Latitude	Resected		Y	Y
9	68	F	2009	R	EFD	Latitude	Resected		Y	Y	Radial head replacement with LCL repair for terrible triad (2009)
10	75	M	2009	L	RA	Latitude	Replaced		Y	N
11	66	F	2009	R	RA	Latitude	Replaced		N	N
	67		2010	L	RA	Latitude	Replaced		N	N
12	66	F	2010	R	PTA	Latitude	Resected		Y	Y	1) ORIF distal humerus (2008) 2) hardware removal and ulnar nerve transposition (2009)
13	43	F	2010	R	RA	Latitude	Replaced / Resected	1) Dissociated radial head (2010) 2) Periprosthetic ulna fracture (Mayo B2)(2013), ORIF (2014)	Y	Y	Elbow arthroscopy and synovectomy (2006)
14	31	F	2011	L	RA	Latitude	Replaced		Y	Y	Elbow arthroscopy and synovectomy (2008)
15	70	M	2011	R	RA	Latitude	Replaced		Y	Y	I&D of spontaneous septic elbow arthritis x2 (2001)
16	71	F	2011	R	RA	Latitude	Replaced		Y	N
17	74	F	2011	R	RA	Latitude	Replaced		Y	Y
18	48	F	2012	R	RA	Latitude	Resected		Y	Y	Elbow arthroscopy and synovectomy (2011)
19	56	M	2012	R	RA	Latitude	Resected		Y	Y
	57		2012	L	RA	Latitude EV	Resected	I&D for superficial wound infection (2012)	Y	Y
20	78	F	2013	R	RA	Latitude EV	Intact		Y	Y

Note: A total of 20 patients had a total of 23 total elbow arthroplasties (TEA). Indications included rheumatoid arthritis (RA), post-traumatic arthritis (PTA), distal humerus non-union (DHN), fibrous dyplasia (FD), and elbow fracture-dislocation (EFD). A total of 13 patients were able to return to clinic for the administration of DASH and MEPS questionnaires (L = left, R = right, Y = yes, N = no, ORIF = open reduction and internal fixation, PJI = prosthetic joint infection, I&D = irrigation and debridement).

A retrospective chart review on all 20 patients (23 elbows) was performed (Figure 1). Four patients (4 elbows) had been revised by the time of the chart review. Component revision of the ulnar and/or humeral components defined “revision” and was considered an endpoint to follow-up. The remaining 16 patients (19 elbows) had an average follow-up of 4.8 years (range, 2.6 to 7.5 years) and were invited for clinical and radiographic evaluation. One patient (1 elbow) was lost to follow-up. One patient (1 elbow) with fibrous dysplasia was excluded from the radiographic analysis due to the distortion of the bone from the disease making interpretation difficult. One deceased patient (2 elbows) and two patients (3 elbows) not responding to participate in the study had ROM measures greater than two years after the index procedure that were included in the evaluation, while the deceased patient (2 elbows) and one patient (1 elbow) not responding to the invitation also had radiographs included. In summary, 15 patients (18 elbows) were included in the ROM data, 13 patients (15 elbows) were included in the radiographic data, and 12 patients (13 elbows) completed clinical outcome measures.

Active ROM measurements including elbow flexion, extension, supination, and pronation were obtained preoperatively, two years postoperatively, and at the clinic visit for those who responded to the invitation to participate in the study. Flexion–extension arc of motion and pronation–supination arc of motion were calculated from this data. ROM was measured with a hand-held goniometer with 25 cm arms. Using this method, it has been shown that meaningful changes in flexion or extension are ≥ 5 degrees.^9,10 Outcome data were derived from patient-reported self-assessment instruments, including the Disabilities of the Arm, Shoulder and Hand (DASH) score¹¹ and Mayo Elbow Performance Score (MEPS)¹² at the last clinic visit only.

Radiographic analysis

Lateral radiographs obtained during the immediate postoperative period and at the last follow-up visit were assessed by two surgeons fellowship-trained in upper-extremity surgery (KAH and AJB) to determine the presence of and/or progression of radiolucencies around the humeral and ulnar components in a method previously described (Table 2).^8,13–15 The ulnar and humeral stems were divided into three equally sized zones on both surfaces of the component (dorsal and volar for ulnar component, anterior and posterior for humeral component); an additional zone at the tip of each stem was also included as part of the assessment. Each surgeon independently reviewed the radiographs and recorded the presence or absence of a lucency, and if a lucency was present, if it increased in size (i.e. progressive lucency) or remained stable (i.e. non-progressive lucency). When discrepancy between the surgeons existed, the radiographs were reviewed to reach consensus. The radial head was noted to be either congruent or not congruent with the capitellar component. The stem of the radial head component was not analyzed for signs of radiographic loosening.

Table 2.

Radiographic analysis of 13 patients (15 total elbow arthroplasties).

Patient Number	Side	Component	Region	Progression
1	L	–	–	–
2	L
3	R L	Ulna Humerus Ulna Humerus	Dorsal Tip, Posterior Volar Anterior, Tip, Posterior	Progressive Non-progressive Progressive Anterior = Non-progressive Posterior = progressive
4	L	–	–	–
5	R
6	L	–	–	–
7	L	–	–	–
8	L	Unable to interpret XR
9	R
10	L	LTF
11	R	No XR
	L	No XR
12	R
13	R	Ulna	Tip	Non-progressive
14	L
15	R	Humerus	Anterior, Posterior	Non-progressive
16	R	Humerus	Anterior, Posterior	Non-progressive
17	R	Humerus	Anterior, Posterior	Non-progressive
18	R L	Humerus Ulna Ulna	Anterior, Posterior Volar, Dorsal Volar, Tip, Dorsal	Progressive Progressive Progressive
19	R	Ulna Humerus	Volar, Tip, Dorsal Anterior, Posterior	Progressive Progressive
20	R

Note: Bold indicates patients who were excluded from the radiographic analysis. Dashes indicate the four patients who underwent a revision procedure. Patient 8 had an underlying diagnosis of fibrous dysplasia and the radiographs were uninterpretable. Patients 10 and 11 were lost to follow-up, and unable to obtain radiographs, respectively. Empty cells indicate that no lucency was detected (L = left elbow, R = right elbow, LTF = Lost to follow-up, XR = radiograph).

Surgical technique

Patients were positioned supine and a straight 15-cm incision was made over the posterior elbow. Full thickness flaps were developed, while preserving the triceps attachment on the olecranon. The components were introduced in a method similar to Szekeres and King.¹⁶ For cases where the radial head was replaced, a trial radial head was used to assess articular congruency with the distal humerus. In certain cases, an attempt to replace the radial head resulted in incongruence of the ulnohumeral articulation (patient 18) and the radial head was therefore left resected. In other cases, patients who underwent previous radial head replacement or excision due to trauma or RA, the decision was made intra-operatively to leave the radial head resected (patients 1, 9, 12). The ulnar and humeral components were cemented using cement restrictors and pressurization. All TEAs were linked. Medial and lateral collateral ligaments were not repaired. The common flexor and extensor origins were repaired back to the medial epicondyle and lateral triceps, respectively. Prior to wound closure, a subcutaneous ulnar nerve transposition was performed. A drain was placed routinely for 24 to 48 h postoperatively. Active elbow flexion/extension and forearm rotation were then started gradually after 24 h.

Statistical analysis

Primary outcomes assessed included ROM (difference between pre- and postoperative measures) and patient-reported outcome instruments (DASH and MEPS). A Wilcoxon signed-rank test was used to compare preoperative to postoperative flexion and extension as well as changes in the arc of motion preoperatively and at least two years postoperatively. The pronation–supination arc of motion was also analyzed in this manner. A p-value of 0.05 was considered statistically significant. Statistical analysis was performed using SPSS Statistics (IBM® SPSS Statistics for Windows, Version 23, Armonk, NY).

Results

The average age at primary surgery was 59 years (range, 30 to 74 years) with 10 left-sided TEAs performed and 13 right-sided TEAs in 5 men and 15 women (Table 1). The mean follow-up was 4.7 years (range, 1 to 7.5 years). Twelve of the 20 patients (60%) had undergone a previous operation of their elbow that was related to a fracture and/or dislocation, or synovectomy for inflammatory arthritis before the index TEA (Table 1). Of the patients who had previous elbow surgery, five had multiple procedures (42%). One patient (patient 4) underwent six previous surgeries including an open reduction and internal fixation (ORIF) and three subsequent revision ORIF procedures of her proximal ulna as well as an ORIF and revision ORIF of a subsequent distal humerus fracture before undergoing a TEA (Table 1). The Latitude prosthesis was used in 21 elbows, and the Latitude EV prosthesis in 2 elbows. The mean time to revision was 3.8 years (range, 1 to 5.2 years). Patients 1 and 3 were deceased at the time of the study.

The radial head was replaced in 15 elbows, resected in 7, and remained intact in 1 elbow. Two of the seven resections (patients 1 and 9, Table 1) occurred in patients who underwent a previous radial replacement for radial head fracture; after the radial head arthroplasty was removed, the Latitude radial head component could not be implanted because the radial neck was resected beyond the resection level required for the Latitude TEA. Radial head replacement was attempted in both elbows of patient 19, but was not performed in either elbow because of incongruity of the trial radial head replacements, likely secondary to variations in this patient’s specific anatomy. Patient 8 required radial head resection due to abnormalities within the proximal radius secondary to fibrous dysplasia (Table 1). Patient 18 had very small anatomy and despite placement of the smallest radial head trial (5 mm), the lateral side of the joint was over-stuffed (Table 1); the decision was made to leave the radial head resected to prevent altering ulnohumeral joint mechanics. Patient 12 had a radial head excision at the time of the initial surgery to treat a comminuted fracture.

Range-of-motion

Fifteen patients (18 elbows, 78% follow-up rate) were analyzed for ROM (Figure 1). The average postoperative ROM follow-up was 5.1 years (range, 2.6 to 7.5 years). Although there were apparent increases in mean terminal flexion (123.7° ± 18.0° to 133.9° ± 10.6°), extension (37.1° ± 25.7° to 27.3° ± 16.2°), pronation (68.7° ± 16.5° to 77.6° ± 10.9°), and supination (61.5° ± 24.2° to 71.4° ± 15.7°), these increases were not statistically significant (p = 0.07, 0.29, 0.34, and 0.07, respectively) (Table 3). Analysis of the improvement in flexion–extension arc of motion demonstrated a statistically significant improvement from 86.6° ± 19.6° to 101.3° ± 25.0°, an improvement of 14.7° (p = 0.04) (Table 3). The pronation–supination arc was 130.2° ± 27.7° preoperatively and 145.2° ± 18.2° postoperatively, but did not reach statistical significance (p = 0.08) (Table 3). There were no statistically significant differences in pronation and supination between patients treated with radial head resection (135.8° ± 14.0°) and those who underwent radial head arthroplasty (146.7° ± 17.9°), although the arthroplasty group demonstrated an almost 10° greater arc of rotation (p = 0.42) (Table 3).

Table 3.

Range-of-motion data for 15 patients (18 Latitude total elbow arthroplasties).

	Preoperative	Postoperative	p
Flexion	123.7 ± 18.0	133.9 ± 10.6	0.07
Extension	37.1 ± 25.7	27.3 ± 16.2	0.29
Flex-ex arc	86.6 ± 19.6	101.3 ± 25.0	0.04
Pronation	68.7 ± 16.5	77.6 ± 10.9	0.34
Supination	61.5 ± 24.2	71.4 ± 15.7	0.07
Pro-sup arc	130.2 ± 27.7	145.2 ± 18.2	0.08
Radial head arthroplasty arc	–	146.7 ± 17.9	0.42
Radial head excision arc	–	135.8 ± 14.0

Note: p-Values calculated using a Wilcoxon signed-rank test with a 0.05 level of significance (Flex-ex arc = flexion-extension arc, Pro-sup arc = pronation-supination arc).

DASH score and MEPS

Twelve patients (13 elbows, 68% follow-up rate) were able to return to clinic for DASH and MEPS administration (Figure 1). The mean postoperative follow-up was 4.5 years (range, 2.6 to 7.5 years). The mean total MEPS score was 89.6 ± 13.9 with 8 of 12 elbows (67%) scoring in the “excellent” range (MEPS score > 90), 2 of 12 elbows (17%) scoring in the “good” range, and 2 of 12 elbows (17%) scoring in the “fair” range. The mean DASH score was 28.1 ± 22.3 (Table 4).

Table 4.

Mayo Elbow Performance Score (MEPS) and Disabilities of the Arm, Shoulder, and Hand (DASH) scores for 12 patients (13 Latitude total elbow arthroplasties).

	MEPS	DASH
Mean	89.6	28.1
Excellent	8 (67%)	–
Good	2 (17%)	–
Fair	2 (17%)	–
Poor	–	–

Radiographic analysis

Radiographs of non-revision TEAs (13 patients, 15 elbows, 79% follow-up rate) were analyzed (Figure 1). The average postoperative radiographic follow-up was 4.2 years (range, 2 to 7.5 years). Of the 15 TEAs analyzed, lucencies were identified in 9 TEAs (60%) (Table 2). There were a total of 26 lucencies detected in various regions around the implants. Eight of the 26 lucencies (31%) had progressed in size at the time of final radiographic follow-up. Therefore, the 4.2-year prevalence of progressive lucency formation was 18.6%.

Complications

Complications were evaluated on all 19 patients (22 elbows); one patient was not included in the complication analysis due to loss of follow-up (patient 10). A total of 7 patients (7 elbows) had 11 complications (Tables 1 and 5). The complications included: periprosthetic fractures (n = 2, 1 each humerus and ulna); humeral spool loosening (n = 1); humeral spool set screw back-out (n = 1); aseptic loosening of the ulnar component (n = 2); radial head dissociation (n = 2); recurrent olecranon bursitis (n = 1); superficial wound infection (n = 1); periprosthetic joint infection (PJI) (n = 1).

Table 5.

Complications in 19 patients (22 Latitude total elbow arthroplasties).

Complication	Number (Rate)	Treatment	Reported rates
Radial head dissociation	2 (9.1%)	Radial head component removal	–
Humeral spool set screw back out	1 (4.5%)	Observation	–
Humeral spool loosening	1 (4.5%)	Humeral spool replacement	–
Aseptic loosening	2 (9.1%)	Component revision	van der Lugt et al. 0.8–13% Little et al. 0–18% Hastings et al. 3–30%
Prosthetic joint infection	1 (4.5%)	I&D, Abx, implant removal	van der Lugt et al. 0.7–5.1% Hastings et al. 2–13%
Superficial wound infection	1 (4.5%)	Wound I&D	–
Recurrent olecranon bursitis	1 (4.5%)	Bursal resection	–
Humeral component periprosthetic fracture	1 (4.5%)	Humeral shortening	–
Ulnar component periprosthetic fracture	1 (4.5%)	ORIF	–
New ulnar nerve palsy	0		Hastings et al. 3–20%
Triceps insufficiency	0		Hastings et al. 1–20%

Note: Reported rates are values from different comprehensive systematic reviews and case series of different TEA implants. Radial head dissociation was calculated out of the total number of radial head arthroplasties performed (2 dissociations out of 15 arthroplasties).

The superficial wound infection was treated with an irrigation and debridement and healed successfully. The two cases of periprosthetic fractures were treated with open reduction internal fixation in one case and with humeral shortening and component revision in the other case. The patient who developed a PJI underwent an aseptic olecranon bursa excision four years after the index procedure at a different institution. Two weeks later, the patient presented with an acute PJI that was treated with a two-stage revision arthroplasty. The revised arthroplasty also became infected and was treated with excisional arthroplasty and did not undergo reimplantation.

Two patients developed radial head component dissociation. Patient 6 had a radial head component dissociation 3.8 years after the index procedure and underwent revision of the radial head component. The radial head dissociated a second time. At the time of the second revision surgery, it was noted intraoperatively that the ulnar component had undergone aseptic loosening and that the humeral spool was dissociated from the stem. The loose ulnar component was revised, the humeral spool replaced, and the decision was made to not re-implant the radial head. The second radial head component dissociation (patient 13) occurred within one month of the index procedure and the patient underwent a second operation to remove the dissociated component. The radial head was left resected (Figure 2).

Figure 2.

Radial head component dislocation (white arrow) at one-month postoperative in a patient who underwent a Latitude total elbow arthroplasty. Anteroposterior (a) and lateral (b) radiographs of the elbow. There is also evidence of cement extrusion (dashed white arrow) from a breach in the ulnar cortex with residual cement surrounding the forearm.

There were no cases of new neuropathy/nerve injury or instances of triceps insufficiency at the time of final clinical follow-up. Patient 2 was noted on follow-up radiographs to have backing out of the humeral spool screw (Figure 3). The elbow was stable on examination and the patient reported no issues with the elbow and was therefore managed non-operatively.

Figure 3.

Humeral spool set screw back-out (white arrow) identified incidentally on routine follow-up radiographs. The patient was asymptomatic and was observed.

Discussion

Total elbow arthroplasty is an effective procedure to relieve pain from a multitude of different etiologies. Despite improvements in TEA design, complication rates remain relatively high when compared to other forms of upper extremity arthroplasty.^5,17 The Latitude TEA is a semi-constrained prosthesis with the option for implanting the components in a linked or unlinked fashion. There are only two previously reported studies using the Latitude in the primary setting.^7,8 We have demonstrated that the Latitude TEA, when used as a primary prosthesis to manage end-stage rheumatoid and post-traumatic arthritis, improves flexion–extension arc and produces good to excellent postoperative outcome scores (DASH and MEPS).

Our results indicate that the majority of patients who underwent TEA using the Latitude demonstrated good to excellent clinical outcome scores, as reflected in the MEPS. We reported a mean MEPS of 89.6 (range, 76 to 100), with the majority (83%) of our patients demonstrating a “good” to “excellent” result at an average follow-up of four years. There were only two patients who scored in the “fair” range. One patient had extensive fibrous dysplasia (patient 8, Table 1) throughout both the humerus and ulna, and the second patient developed significant radiolucencies at the tip of her ulnar component (patient 19, Table 1) (Figure 4(b)). Comparing our outcomes with Burkhart et al.,⁷ the MEPS scores were nearly identical with an average MEPS score of 89.2 (76 to 100) in the former study. This is also comparable to the mean MEPS found in a systematic review by Welsink et al.⁵ of 85.3 in patients with RA.

Figure 4.

Non-progressive and progressive radiolucencies of the humeral and ulnar components of the Latitude total elbow arthroplasty at four years postoperative. Evidence of a non-progressive lucency (solid white arrow) posterior to the humeral component at the two week postoperative point (a) with no evidence of progression at four years postoperative (b). The dashed white arrow indicates development of a progressive lucency at the tip of the ulnar component (b) which was not present immediately postoperative (a).

The evaluation of our 30- to 74-year-old patient population using the DASH revealed an average score of 28.1, which is comparable to a 70- to 79-year-old Norwegian cohort (DASH score of 22) with no medical or physical comorbidities.¹⁸ The average DASH score in the study performed by Burkhart et al.⁷ was 8.4 (range, 0 to 28); the discrepancy is possibly due to the observed differences in patient demographics between the two studies. The majority of our patients underwent the procedure for RA (65%) while only one of the 15 (7%) patients in Burkhart et al.’s⁷ study had RA as the primary indication for surgery. As RA commonly affects all joints of the upper extremity, it is likely that our patients had uniformly higher DASH scores due to disability in joints proximal and distal to the elbow. Despite the higher DASH scores, our results indicate that the majority of patients treated with the Latitude TEA demonstrated similar functionality compared to an age-matched cohort of unaffected individuals.¹⁸ Unfortunately, the only other study that has reported outcomes of the Latitude TEA used different outcome instruments; for these reasons, comparison cannot be made between this study and our results.⁸

The statistically significant improvement in flexion–extension arc by almost 15° is also clinically significant as it has been demonstrated that a 100° arc of motion is necessary to independently perform most activities of daily living.¹⁹ The primarily RA patient cohort in this study had a preoperative flexion–extension arc that was well below the functional arc of motion. The Latitude TEA provided an improvement in arc of motion by 15°, well within the average increase in arc of motion previously found in a systematic review of other TEA designs used to treat elbows with RA.⁴ These results are consistent with the findings of Wagener et al.⁸ using the Latitude, as well as other TEA designs, which have demonstrated improvements from 10° to 20° in elbow flexion.⁶ The preoperative pronation–supination arc of 130° indicates that the disease process in this patient cohort did not affect rotation about the elbow as much as flexion and extension (i.e. sagittal plane motion).

In determining complication rates, established TEA-specific complications were noted including infection, aseptic loosening, and periprosthetic fracture. We also reported on complications unique to the Latitude prosthesis, such as radial head dissociation, humeral spool loosening, and humeral spool set screw back-out. In our series, seven elbows in seven patients experienced at least one complication, for a rate of 32% of elbow replacements, or 37% of patients. This complication rate is comparable to previously reported rates, with Voloshin et al.²⁰ publishing a complication rate from 64 studies of 24.3% ± 5.8%. More recently, Hastings et al.^19,21 reviewed the complication rates of four different total elbow prostheses, which ranged from 14% to 72%.

The rate of clinically significant aseptic loosening requiring component revision in our series was 2 out of 22 elbows (9.1%). This is comparable to the rate reported by Voloshin et al.²⁰ in a systematic review of linked TEA designs (5.2% ± 4.5%). As described by Little et al.,⁶ aseptic loosening of TEA prostheses is of concern, with five-year survival rates of 82% for the Kudo TEA, 86% for the Coonrad-Morrey TEA, and 81% for the Souter-Strathclyde TEA.

In an attempt to correlate clinical loosening and signs of loosening on radiographs, the number of progressive lucencies was analyzed and found to be 33%. Despite this relatively high proportion of progressive lucencies, only two elbows in two patients were symptomatic to the point of requiring revision surgery. In the TEA literature, the formation of progressive lucencies has been thought to indicate implant loosening. However, the significance of progressive lucency formation in the Latitude implant, as well as other total elbow implants in general, is unknown as the majority of the patients in this series were asymptomatic. An improved method for detecting clinical loosening in the context of progressive lucencies on plain radiographs is needed to help understand and address this important issue. Due to the senior authors concerns regarding ligament integrity, especially when the predominant indication for TEA was RA, every prosthesis was implanted in a linked fashion. There is no convincing clinical data that unlinked prostheses have lower rates of loosening compared to linked prostheses that have 5 to 7 degrees of varus/valgus laxity in the ulnohumeral articulation.

We have identified a previously reported²² pattern of radiographic lucency around the tip of the ulnar component that developed in three of the fifteen (20%) TEAs reviewed in the current series (Figure 4). Although two of the three TEAs with this loosening pattern were functioning well at the two-year follow-up, one patient scored “fair” with an MEPS of 55. Cheung and O’Driscoll have previously reported ulnar tip lucencies in 10 patients who presented with painful ulnar components using the Coonrad-Morrey prosthesis.²² The proposed mechanism is anterior humeral flange impingement and subsequent pistoning of the ulnar component.²² The first generation Latitude implant included a smooth, cylindrical ulnar component that may have permitted longitudinal translation of the implant within the cement mantle. The potential problem of ulnar component pistoning may be addressed with the new Latitude EV ulnar stem, which is more trapezoidal in shape and includes a proximal plasma sprayed coating to permit bony on-growth/in-growth, thereby increasing implant stability. Biomechanical testing has demonstrated that stem surface treatment with coatings can increase stem stability in an in vitro model of different stem designs.²³ Unfortunately, only 2 of 23 elbows in this series were implanted with the Latitude EV components (when the implant became commercially available) and the numbers are insufficient to perform a direct clinical or radiographic comparison between the two groups.

In the current study, there were two infections out of 22 elbows (9.1%) that required operative intervention: one superficial wound infection and one PJI. The superficial wound infection rate in our series (4.5%) is comparable to other reported soft tissue complication rates after TEA that range from 4%⁸ to 5.5%.²⁴ The deep infection case in our study occurred as a postoperative complication of a bursal resection (patient 7, Table 1). The deep infection rate of one of 22 elbows (4.5%) is similar to previously reported rates in the literature.²⁰ Voloshin et al.²⁰ found a deep infection rate of 3.3% ± 2.9%, with rates in some series as high as 9%. Contributing factors include the predominantly rheumatoid population that underwent the procedure, who are often on immunosuppressive medications.²⁰

A theoretical advantage of the Latitude system over other TEA implants is the ability to replace the radial head component which can improve stability of the elbow and treat concurrent radiocapitellar arthrosis. Management of the radial head remains a challenge for several reasons, including previous resection, tracking issues detected intra-operatively, and dissociation of the radial head component following the arthroplasty procedure. The complication of radial head component dissociation using the Latitude TEA has been previously reported by Wagener et al.⁸ They replaced the radial head component in 42 of their TEAs (61%) and reported 13 disengaged radial head components.⁸ Wagener et al.⁸ concluded that the disengagements did not significantly affect outcome. They removed the disengaged component in only one of these patients due to pain. Regarding the two patients in our series that experienced radial head component dissociation, each patient required a second surgery to remove the component. The radial head component dissociated one month postoperatively in patient 13 and was removed due to tracking issues (Figure 2); after removal of the radial head component, the patient’s elbow was stable and exhibited a pain-free ROM. The second patient (patient 6, Table 1) sustained a radial head component dissociation at 3.8 years after the primary TEA, which was revised. The radial head component dissociated a second time in association with ulnar component and humeral spool loosening. The issues with multiple components in this patient suggest that the radial head dissociation was not due to an implant-related design issue, but overall poor functioning of the entire joint. Radial head replacement can create problems related to tracking and overstuffing the joint. Radial head component dissociation or disengagement may be an issue unique to the Latitude implant. Based on the above results, the senior author (KAH) does not resect the radial head or replace the radial head if it was previously resected.

The final considerations for complications include screw back-out, triceps insufficiency, and the ulnar nerve. There were no cases of ulnar cap screw back-out and one case of humeral spool screw back-out identified (Figure 3). The humeral screw complication occurred sometime between the first and seventh year from the arthroplasty for reasons unknown; the patient did not experience instability nor was instability of the elbow detected on clinical examination. Operative intervention was deemed unnecessary. There were no cases of triceps insufficiency postoperatively, likely related to the triceps-sparing approach utilized in the surgical technique. Multiple surgical approaches to the proximal ulna and distal humerus have been described for TEA including triceps-reflecting, triceps-splitting, and triceps-sparing.¹⁹ A review of 100 TEAs performed using the Bryan-Morrey (triceps elevating or reflecting) approach reported seven cases (7%) of triceps weakness postoperatively, with the clinical significance of triceps weakness unclear.²⁵ The results from this study suggests that performing a TEA through a triceps-sparing approach results in consistent postoperative triceps function with no identified cases of insufficiency. Ulnar nerve dysfunction is a rare complication, reported at 2.9% in a review of 6091 TEAs.⁵ There were no cases of ulnar nerve dysfunction in this series.

Limitations of this study include an inability to assess the magnitude of improvement in functional outcome, due to the lack of preoperative DASH and MEPS outcome scores. Furthermore, there was not a control group that could be used for comparison (i.e. a retrospective review). An additional limitation was the different numbers of patients represented in the ROM and outcome score (DASH and MEPS) data groups; as we were unable to recall all patients for DASH and MEPS scores but included these patients in the ROM data, the populations were not identical (i.e. not all of the patients whose data are represented in the ROM data are represented in the DASH and MEPS score data). Lastly, improvements in radiographic analysis of implants for aseptic loosening are warranted in future studies.

Conclusion

In conclusion, the Latitude TEA, as performed by the authors, significantly improves patient function and results in good to excellent clinical outcomes in the majority of patients who undergo the procedure. The complication rate in the current series was comparable to, if not lower, than other implant designs with the unique complication of radial head component dissociation. Due to these issues with radial head component dissociation, the senior author does not resect the native radial head and does not use the Latitude radial head component if the patient has had a previous radial head arthroplasty or a radial head resection. All components in this series were implanted in a linked fashion. The Latitude TEA is a reliable option for improving end-stage elbow arthritis with consistent improvements in flexion–extension arc and good to excellent clinical outcomes.

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.

Ethical Review and Patient Consent

Conjoint Health Research Ethics Board REB13-0808, University of Calgary.