Joint replacement surgery for elbow tumours: a systematic review of outcomes

Abstract

Background

Tumour resection followed by joint reconstruction is a surgical option in the appropriate patient. The evidence for such reconstructive surgery of the elbow joint is limited. The aim of this study is to review the literature to evaluate the outcomes of joint replacement surgery in tumours of the elbow.

Methods

A systematic review of PUBMED and EMBASE databases was conducted. Case series and comparative studies reporting results after total elbow arthroplasty, modular endo-prosthetic replacement and custom prosthesis were eligible for inclusion.

Results

Eleven eligible studies were identified (n = 134). At mean follow-up of 44 months, the overall revision rate was 14% and complication rate was 28%. The mean Mayo Elbow Performance Score was 75, with 56% of patients reporting good or excellent outcomes. The mean post-operative range of motion was 97°.

Discussion

Elbow prosthesis reconstruction after tumour resection can provide good functional outcomes at mid-term follow-up. The complication and revision rates are comparable to other indications for elbow replacement surgery. Further prospective studies are required to compare outcomes between different elbow arthroplasty options after tumour resection.

Keywords

Total elbow arthroplasty endo-prosthetic replacement custom prosthesis elbow tumour outcomes

Introduction

Tumours of the elbow are extremely rare and account for 1% of all tumours presenting to orthopaedic surgeons.¹ The limited soft tissue envelope coupled with the proximity of neurovascular structures makes the surgical management of elbow tumours a lot more challenging than other anatomical regions. These challenges have previously resulted in amputation being an accepted surgical option.² However, limb salvage has been reported to have improved functional outcome and cosmesis.³ The evolution of chemotherapy and radiotherapy have resulted in limb salvage procedures being increasingly considered for treatment of these challenging cases with primary bone and soft tissue tumours. Furthermore, in patients with metastatic disease in the elbow, pathological fracture and limited life expectancy, limb salvage seems more appropriate than ablative surgery.

A multidisciplinary approach must be adopted in the decision-making process between resection with reconstruction or amputation. Key factors considered include tumour size, location, extramedullary extension, metastatic disease, life expectancy and patient factors. Reconstructive options include standard total elbow arthroplasty (TEA), modular endo-prosthetic replacements (EPR), resection arthroplasty, interposition arthroplasty, arthrodesis, elbow osteoarticular allograft reconstruction, allograft-prosthesis composite (APC) arthroplasty and vascularized fibular grafts.^1,4,5 Of these options, elbow reconstruction in the form of a prosthesis or APC is more attractive as it has been reported to restore bone defects whilst providing acceptable analgesia, function and complication rates.⁶ An inappropriate selection of a standard TEA in patients with large defects after tumour resection could lead to instability with complications like implant loosening and failure. On the other hand, an EPR using a megaprosthesis after significant soft-tissue stripping is unlikely to allow as good a functional outcome as a TEA.⁷

It is important to acknowledge that as part of the ‘Get It Right First Time’ approach (GIRFT) to Orthopaedic Practice, Elbow Replacement Surgery has come under particular scrutiny. A robust multidisciplinary approach has been developed by the British Elbow and Shoulder Society (BESS), and from here-on, these management principles will be as important to tumour practice as they are to elbow replacement in non-tumour patients.⁸ It is important that an elbow arthroplasty specialist from a designated centre should be involved throughout the management process including surgery.

In addition, each patient should be actively involved in decision making over the surgical reconstruction options. This requires informed consent and discussion of the benefits, potential risks, complications and long-term implications of surgical intervention. However, the current evidence available on patients undergoing elbow replacement surgery for tumours is extremely limited. The aim of this systematic review is to document the functional outcomes, complications and revision rates in this patient group to aid decision making for these complex cases.

Methods

Search strategy

A systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.⁹ An electronic search was performed on MEDLINE and EMBASE databases for all English language articles with no time filter restriction. The search was performed independently by two authors on 01 September 2020 and repeated on 10 December 2020 to ensure accuracy. The authors also performed searches in the Cochrane library and manual searches in Google using the keywords ‘elbow tumours’, ‘elbow replacement’, ‘elbow arthroplasty’ and ‘endo-prosthetic’. The reference lists of included articles were also screened to identify any articles not initially identified in the preliminary searches. Any discrepancies were resolved through discussion between these two authors, with the senior author resolving any residual differences.

Inclusion and exclusion criteria

Clinical studies published in English were considered for eligibility. Studies could be either case series or comparative studies reporting results after elbow replacement for elbow tumours. Case reports were excluded.

For inclusion, the study must have reported functional outcomes, complications or revision rates. Elbow tumours could be involving either the distal humerus or proximal ulna. The elbow prosthesis used could be TEA, EPR or custom prosthesis, with or without additional use of allografts. Only primary research was considered for review with any abstracts, comments, review articles, case reports and technique articles excluded.

Data extraction

Information was extracted regarding the number of patients in each included study, along with their age, diagnosis, procedure, follow-up, outcomes (range of motion (ROM), post-operative functional scores), complications and revision rates.

Statistical analysis

Descriptive statistics such as mean and range are reported where applicable and percentages were calculated for qualitative data. The heterogeneity and frequent absence of crucial data made comparative statistical analysis difficult. The assessment of methodological quality of the studies was undertaken using the methodological index for non-randomised studies (MINORS) score, where non-comparative studies achieve a maximum score of 16 and comparative studies achieve a maximum score of 24.¹⁰

Results

The search strategy identified 11 studies eligible for inclusion (n = 134): 3 studies on TEA with or without allograft (n = 38),^1,11,12 4 studies on EPR (n = 56)^7,13–15 and 4 studies on custom replacements (n = 40).^16–19 A summary of the included studies is presented in Tables 1 to 3. The demographics according to surgical intervention are provided in Table 4. A flow chart of the search strategy is shown in Figure 1. Table 5 outlines the reasons for article exclusion after full-text review.^20–30

Table 1.

Summary of the included articles for patients receiving TEA.

Case series	Patients			Surgery		Outcomes	Results
Case series	Number	Mean age [range] (years)	Diagnosis	Prosthesis	Follow-up in months (range)	Outcomes	Results
Sperling et al.¹	13	65 [43–83]	Primary – 7 (lymphoma, GCT, chondrosarcoma, MM, osteochond-romatosis) Metastases – 6 (renal, breast, rectal)	Version of Coonrad prosthesis	30	ROM arc (degrees)	73 post-op (29 pre-op)
						Scores	MEPS – 69% excellent/good Pain score 1.6 (3.3 pre-op)
						Complications	23.1% (1 HO excision, 1 ulnar paraesthesia, 1 screw replacement) [3 local recurrence – 1 amputation]
						Revisions (average time to)	0% (23.1% reoperation due to local recurrence)
						Mortality Time to death (months)	15.4% (3.5)
Athwal et al.¹¹	20	66 [38–83]	Primary – 10 (chondrosarcoma, lymphoma, MM, recurrent GCT) Metastases – 10 (cervical, breast, lung, renal, colon)	19 Standard linked semi-constrained Coonrad-Morrey TEA (3 with APC) 1 Custom humeral component with an extended flange and standard ulnar Coonrad-Morrey component	34 (1–192)	ROM arc	48 pre-op 92 post-op
						Scores	Pain 32 (pre-op 2) MEPS 75 [30% E/G] (pre-op 22)
						Complications	35% (5 nerve, 1 healed fracture, 1 CRPS) [5 local recurrence – 1 amputation]
						Revisions (average time to)	0% (10% reoperation due to local recurrence)
						Mortality Time to death (months)	70% (16)
Schwab et al.¹²	5	40 [23–58]	Primary – 5 (synovial sarcoma, malignant histiocytoma, Ewings) Metastases – 0	Coonrad morrey semi-constrained cemented (4 APC)	60 (20–105)	ROM arc	Post-op 106
						Scores	MEPS – 100% excellent/good
						Complications	40% (1 radial neuropraxia recovered, 1 humeral loosening)
						Revisions (average time to)	20% (1 loosening humeral)
						Mortality Time to death (months)	1 (25%) 61

TEA: total elbow arthroplasty; GCT: giant cell tumour; MM: multiple myeloma; APC: allograft-prosthesis composite; ROM: range of motion; MEPS: Mayo Elbow Performance Score; HO: Heterotopic Ossification; E/G: excellent/good; CRPS: Chronic regional pain syndrome.

Table 2.

Summary of the included articles for patients receiving modular endo-prosthetic replacement.

Case series	Patients			Surgery		Outcomes	Results
Case series	Number	Mean age [range] (years)	Diagnosis	Prosthesis	Follow-up in months (range)	Outcomes	Results
Kulkarni et al.⁷	9 (1 paeds excluded)	51.1 [29–76]	Primary – 7 (osteosarcoma, liposarcoma, leiomyosarcoma, plasmacytoma, pleomorphic sarcoma, GCT) Metastases – 2 (renal, breast)	Stanmore prosthesis, constrained hinged-cemented	96.7 (9–372)	ROM arc (degrees)	Post-op 100
						Scores	TESS 69% (from five patients)
						Complications	22.2% Loosening humeral (2)
						Revisions (average time to)	22.2% 2 Loosening (56, 366)
						Mortality Time to death (months)	44.4% (35.3 mean)
Peterson et al.¹³	9	57 [36–78]	Primary – 7 (CCS, lymphoma, MM, GCT, synovial sarcoma, plasmacytoma) Metastases – 2 (renal, breast)	Endoprosthesis modular hinge semiconstrained (Stryker)	34 (3–75)	ROM arc	110 post-op
						Scores	MSTS 81%
						Complications	33% (1 median nerve entrapment, 1 Infection, 2 bushing failure) [1 local recurrence]
						Revisions (average time to)	33%
						Mortality Time to death (months)	67% 17
Capanna et al.¹⁴	27 (once filtered for age and other procedures)	66 [24–83]	Primary – 9 (GCT, chondrosarcoma, osteosarcoma, STS, plasmacytoma) Metastases – 18 (unknown primary)	HMRS megaprosthesis (Howmedica) and discovery SRS prosthesis (Biomet)	20.2 (2–144)	ROM arc	Post-op over 100 (67% patients)
						Scores	MSTS 78%
						Complications	14.8% (2 permanent radial nerve palsies, 1 ulnar nerve palsy, 1 structural failure)
						Revisions (average time to)	5 year implant survival 93.7%
						Mortality Time to death (months)	74% from disease 11
Henrichs et al.¹⁵	11 (1 lost to f/u)	46 [21–73]	Primary – 9 (osteosarcoma, GCT, Osseus UPS, ABC, STS, MM, Ewing’s sarcoma) Metastases – 2 (renal)	Modular universal tumour and revision stem	91 (13–229)	ROM arc	Post-op 108 FFD 18 (55% patients)
						Scores	MSTS post-op 80%
						Complications	45% (3 aseptic humeral loosening, 1 aseptic ulna loosening, 1 infection, 1 wound healing problem) [3 local recurrences]
						Revisions (average time to)	36% at five years [25% - 3 amputation for local recurrence]
						Mortality Time to death (months)	27%

ABC: aneurysmal bone cyst; GCT: giant cell tumour, UPS: undifferentiated pleomorphic sarcoma, STS: soft tissue sarcoma; MM: multiple myeloma, CCS: clear cell sarcoma; HMRS: Howmedica Modular Resection System; SRS: Segmental Revision System; ROM: range of motion; TESS: Toronto Extremity Salvage Score; MSTS: Musculoskeletal Society Tumour Score; FFD: fixed flexion deformity.

Table 3.

Summary of the included articles for patients receiving custom prosthesis.

Case series	Patients			Surgery		Outcomes	Results
Case series	Number	Mean age [range] (years)	Diagnosis	Prosthesis	Follow-up in months (range)	Outcomes	Results
Rolf et al.¹⁶	4	65 [57–73]	Primary – 0 Metastasis – 4 (renal-solitary)	Custom-made GSB-1 prosthesis, sloppy hinge constraint	36.8 (12–61)	ROM arc	Post-op 93
						Scores	–
						Complications	25% Loosening ulna (1)
						Revisions (average time to)	25% 1 Loosening (24)
						Mortality Time to death (months)	50% (57)
Hanna et al.¹⁷	15 (3 paeds excluded)	56.1 [25–87]	Primary – 10 (Ewings sarcoma, chondrosarcoma, melanoma, HGSSS, myeloma, GCT) Metastases – 5 (renal, breast, laryngeal)	Custom-made constrained hinged implant Stanmore	52.8 (12–218.4)	ROM arc	–
						Scores	MSTS 75% TESS 72%
						Complications	26.7% (1 radial neuropraxia recovered, 1 Loosening, 2 infection) [2 local recurrence – 1 amputation]
						Revisions (average time to)	20% (2 infection, 1 loosening) [6.7% reoperation for recurrence]
						Mortality Time to death (months)	46.7% (7/15 patients dead at mean 27.9 months after surgery)
Tang et al.¹⁸	15 (10 paeds/ total humerus excluded)	45.8 [22–71]	Primary – 9 (Ewings sarcoma, MFH, synovial sarcoma, lymphoma, desmoid, GCT) Metastases – 6 (renal, lung, rectal, ovarian)	Custom-made constrained hinged implant (Chinese)	40 (12–98)	ROM arc	Post-op over 100 (for 10 patients)
						Scores	MSTS 83% MEPS 84 (80% good/ excellent)
						Complications	20% (3 Loosening,) [2 Local recurrence]
						Revisions (average time to)	20% (55 months)
						Mortality Time to death (months)	53.3% (21.9 mean)
						Mortality Time to death (months)	74% from disease 11
Balasubramanian et al.¹⁹	6	34.3 [23–45]	Primary – 6 (GCT)Metastases – 0	Custom prosthesis – stainless steel locally made	26 (18–33)	ROM arc	Post-op 92 Fixed flexion deformity (post-op) 10-20
						Scores	DASH 77
						Complications	30% 2 Iatrogenic ulna neuropraxia – spontaneous recovery by 10 weeks
						Revisions (average time to)	0
						Mortality Time to death (months)	0

GSB: Gschwend/Scheier/Bähler; GCT: giant cell tumour; MFH: malignant fibrous histiocytoma; HGSSS: high grade soft tissue sarcoma; ROM: range of motion; TESS: Toronto Extremity Salvage Score; MSTS: Musculoskeletal Society Tumour Score; MEPS: Mayo Elbow Performance Score; DASH: Disabilities of the Arm, Shoulder and Hand score.

Table 4.

Summary of outcomes from different types of prosthesis.

			Prosthesis type
			TEA	Endoprosthesis	Custom	Total/mean
Number of studies			3	4	4	11
Number of patients			38	56	40	134
Mean age (years)			62.2	58	50	56.8
Mean follow-up (months)			36.1	49.4	42.4	43.6
Outcomes	ROM arc (degrees)		90	105	95
	Score	MEPS % E/G	66.3	–	80
		MSTS	–	80	79
		TESS	–	69	72
		DASH	–	–	76
	Complication rate (%)		32.7	28.8	25.4	28.4
	Revision rate (%)		2.6	19.6	17.5	14.2

ROM: range of motion; MEPS % E/G: Mayo Elbow Performance Score % of Excellent/Good; MSTS: Musculoskeletal Society Tumour Score; TESS: Toronto Extremity Salvage Score; DASH: Disabilities of the Arm, Shoulder and Hand score; TEA: total elbow arthroplasty.

Table 5.

Reason for exclusion of articles after full-text review.

Study	Reason for exclusion
Ross et al.²⁰	Collated data for all indications with no individual data or stratification for tumours
Weber et al.²¹	Collated data for entire humerus with no stratification
Bickels et al.²²	Collated data for entire humerus with no stratification
Torbert et al.²³	Collated data for upper and lower limb endoprosthesis with no stratification
Funovics et al.²⁴	Collated data for both total humeral and distal humeral replacements with no stratification
Hossein et al.²⁵	Total elbow allograft transplantation
Sewell et al.²⁶	Only one adult patients in the series
Sorensen et al.²⁷	Collated data for upper and lower limb joint replacements with no stratification
Pradhan et al.²⁸	Collated data for entire upper limb (including metacarpals and distal radius) with no stratification
Pruksakorn et al.²⁹	Collated data for distal humeral, proximal humeral and total humeral prosthesis with no stratification
Casadei et al.⁶	Collated data for total and distal humerus, paediatric cases and revision prosthesis.
Chebli et al.³⁰	Indications for surgery were not restricted to tumours and no stratification of the collated data.

The mean age at surgery for all patients was 56.8 years (21 to 87 years), and the overall mean follow-up duration was 44 months (1 to 372 months), as outlined in Table 4. Seventy-nine patients (59%) receiving elbow arthroplasty were diagnosed with primary tumours, while 55 patients (41%) had a metastatic deposit in the elbow. Primary tumours included aneurysmal bone cyst, giant cell tumour of bone, undifferentiated pleomorphic sarcoma, soft tissue sarcoma, malignant fibrous histiocytoma, high grade soft tissue sarcoma, multiple myeloma and clear cell sarcoma. Metastatic lesions included primary lesions from breast, lung, kidney, colon, uterus, ovary, cervix and larynx.

Functional outcomes

A summary of outcomes from the included studies including ROM (through flexion/extension), functional scores, complications, revision rates and mortality are provided in Tables 1 to 3. These data have been collated in Table 4 into overall means using each type of implant when available.

Range of motion

The average post-operative ROM arc for the three implant types was 97°. The average ROM arc for patients with an EPR was 15° greater than patients with a TEA and 10° greater than those with a custom prosthesis.

Athwal et al. reported the largest series using TEA and demonstrated a statistically significant improvement in ROM from 48° pre-operatively to 92° post-operatively (p < 0.05).¹¹ Similarly, Capanna et al. presented the largest case series using EPR and reported that two-thirds patients had over 100° arc of motion.¹⁴ The largest such series of custom prosthesis reported by Tang et al. also demonstrated that two-thirds of patients achieved a post-operative ROM of over 100°.¹⁸

Only two studies documented both pre-operative and post-operative ROM, both reporting on TEA and demonstrating statistically significant improvements in post-operative ROM.^1,11 Hence, no meaningful statistical change in arc of motion for comparison of different prosthetic types after surgical intervention can be presented. However, it is important to consider that for tumour-related surgery, function is often sacrificed for survival.

Outcome scores

The Mayo Elbow Performance Score (MEPS) was reported in four studies (n = 53). Overall, 70% of the patients had excellent or good scores as demonstrated in Table 4; 80% of patients who received a custom prosthesis reported excellent or good MEPSs, compared to the 66% receiving TEA. There were no MEPSs for patients receiving an EPR. Five studies reported the Musculoskeletal Society Tumour Score (MSTS) (n = 77). The overall mean score for all prosthesis was 80% and was similar after custom prosthesis (79%) and EPR (80%) as illustrated in Table 4. There were no MSTSs for TEA patients. Toronto Extremity Salvage Score (TESS) and Disabilities of the Arm Shoulder and Hand scores were both reported in some studies and the details of these are presented in Tables 1 to 3.

Complications

All included studies reported complications. Overall, there were 38 complications in the 134 patients (28%). The complication rates showed no statistical difference between the groups (p = 0.82); TEA (33%), EPR (29%) and custom prosthesis (25%). Furthermore, there were 17 additional local recurrences during follow-up. The most common overall complications were aseptic prosthesis loosening (11%), nerve injury (10%) and infection (2%). Aseptic prosthesis loosening was due to humeral loosening (7%), ulna loosening (2%) or bushing failure (2%). The other less frequent complications (each at n = 1/134) included heterotopic ossification, united peri-prosthetic fracture, chronic regional pain syndrome, wound healing complication and screw replacement.

The most common complication in the group of patients receiving a TEA was nerve injury (18%). Aseptic loosening occurred in one patient (3%) and there were no reported infections.

The most common complication in the modular EPR group was aseptic prosthesis loosening (16%), followed by nerve injury (7%) and infection (4%). Permanent nerve palsies were only recorded in patients who underwent implantation of endoprosthesis (n = 2, both radial nerve) and they were both in the same study.¹⁴

The most common complication in the custom prosthesis group was aseptic prosthesis loosening (13%), followed by nerve injury (8%) and infection (5%).

Revisions

The overall revision rate for all 134 patients was 14% with a mean follow-up of 44 months. The revision rate for TEA was the lowest at 3% (follow-up 36 months), custom prosthesis was higher at 18% (follow-up 42 months) and modular EPR was the highest at 20% (follow-up 49 months). Despite the apparent difference in revision rates, this was not statistically significant (chi-square test; p = 0.555).

In the larger case series, Hanna et al. and Tang et al. reported 20% revision rates (n = 15, follow-up 53 months and 55 months, respectively) for custom prosthesis, Athwal et al. reported no revisions for TEA (n = 20, follow-up 34 months), and Capanna et al. reported five-year implant survival of 94% (n = 27) for EPR.^14,17,18

Re-operation for local recurrence

The local recurrence rate for all patients was 13% (n = 17). The recurrence rate was highest for TEA (21%), followed by custom prosthesis (13%) and EPR (7%). The incidence of reoperation amongst those cases of local recurrence was 65% (n = 11) with six cases (36%) requiring an above elbow amputation. The alternative treatment options included proximal humeral allografts (12%), soft tissue excisions when recurrence was in the soft tissues (12%), prosthetic replacement (6%) or non-operative palliative management (35%).

Study quality

All studies in this systematic review were non-comparative case series with evidence level 4. The studies were assessed for level of quality using the MINORS criteria. The mean MINORS score for all 11 studies was 9.8 (range 7–11) suggesting that these studies were of poor methodological quality.

Discussion

The most important finding of this study is that although the overall revision rate was 14% at a mean of 44 months, the revision rate for implant failure after TEA appeared to be lower (3%) than after modular EPR (20%) and custom implants (18%) despite similar follow-up periods.

TEA is indicated for alternative pathologies like rheumatoid arthritis, osteoarthritis, acute trauma and post-traumatic arthritis, and revision rates of 15% have been reported in the literature at a 7.8-year follow-up.³⁰ The highest revision rates were reported after trauma sequelae (30%), while the more common indications like inflammatory arthropathy (11%–13%) and acute trauma (10%–11%) were lower.³¹ The literature is generally scarce for EPR and custom prosthesis undertaken for non-oncological aetiology of the elbow joint, as a primary procedure. Hence, it is difficult to make direct comparisons of revision rates after EPR and custom implants for tumours against non-tumour indications.

Cross et al. published their results of 14 custom-made elbow replacements (10 Juvenile Rheumatoid Arthritis[JRA] and 4 Rheumatoid Arthritis [RA]) with mean age 28 years and mean follow-up of 18 years.³² They noted no revisions but recorded bushing changes and implant removal for infection. Ross et al. performed nine EPRs for fractures with a mean follow-up of six years.²⁰ Although no revisions were performed, implant removal was recorded for infection (33%). Funovics et al. performed EPR in 38 tumour patients and 15 non-tumour patients.²⁴ The tumour group saw a 5% revision rate, while the non-tumour group experienced a 20% revision rate. They concluded that EPR was a viable option in patients with a tumour but highlighted the need for careful patient selection in non-tumour patients. It is noteworthy that the non-tumour group included five patients with failed primary TEA.

The familiarity of using a standard TEA implant may be one potential reason for a reduced revision rate with this implant, and although the UK National Joint Registry only reports an average of 467 cases performed annually over the last five years, it is likely that the BESS-GIRFT initiative will improve the experience of surgeons.^8,33 In addition, tumour cases which are deemed amenable to TEA rather than EPR or custom implants may require less soft tissue resection and present less challenging surgical procedures.

The local recurrence rate is reported to be higher after TEA (21%) when compared to modular EPR (7%) and custom implant (13%) in this review. This raises concerns that in an effort to use a standard TEA implant the margins of resection maybe compromised, increasing risk of local recurrence. However, the heterogeneity of patients included within these studies and the lack of comparative studies means this is only a hypothesis based on the current evidence.

The overall complication rate for all types of implants in this review was 28% (TEA 33%, modular EPR 29% and custom 25%), with the most common complications being aseptic loosening (11%) and nerve injury (10%). Humeral loosening was three times more common than ulna loosening. 14% of nerve injuries were permanent. The literature suggests that these complication rates are comparable to those reported for alternative indications. The complication rates after TEA for inflammatory arthropathy is reported at 5.2%–30.9%, and after acute trauma or trauma sequelae is at 14.2%–50%.³⁴ The commonest complications after TEA in inflammatory arthropathy are bushing wear (23%) and deep infection (8%),³³ while in acute trauma they are nerve injury (8%) and aseptic loosening (2.3%), and for trauma sequelae, they are aseptic loosening (18.9%) and infection (5.7%).³⁵ Cross et al. reported complication rates of 36% for custom elbow prosthesis, with the commonest being bushing wear (28.6%) and deep infection (7%).³² Ross et al. reported complication rates of 56% for their series of EPRs, with the commonest being deep infection (33.3%) and aseptic loosening (22.2%).²⁰ Funovics et al. reported similarly high complication rates of 60% for their non-tumour patient group who received EPRs compared to their tumour group who experienced 24% complication rates.²⁴

This review of tumour patients showed an overall post-operative arc of motion of 90° in the TEA group, 105° in the EPR group and 95° in the custom group. This is compared to the literature which reports 115° for rheumatoid arthritis, 97° for acute trauma and 94° for trauma sequelae when considering TEA prosthesis.^35,36,37 With regards to non-oncological indications, the mean arc of motion for custom prosthesis was reported at 111° and EPR was 81°.^20,32 Funovics et al.’s comparative study showed no significant difference in arc of motion between their tumour group at 97° and their non-tumour group at 96°.²⁴

Similarly, when considering the post-operative functional scores in this review, the authors report an MEPS of 75, with 56% of patients reporting good or excellent outcomes after TEA for tumours. The literature reports similar functional outcome scores for TEA with MEPS of 81–94 for rheumatoid arthritis, MEPS of 83–99 for acute trauma and MEPS of 80–94 for trauma sequelae.³⁴ Tumour patients receiving EPR in this review reported MEPS of 84, MSTS of 80 and TESS of 69. The Inglis–Pellicci score reported by Funovics et al. on EPR for their tumour group was 84, better than the score for their non-tumour group at 76.²⁴ The authors were unable to find any other scores for EPR in non-tumour patients in the literature, to compare with the tumour cohort in this review. The MEPS of 91 for custom prosthesis in non-tumour patients was better than the custom prosthesis MEPS of 84 in this review. Other functional scores recorded for custom prostheses from this review were MSTS of 79 and TESS of 72, which are comparable to the scores for the EPR cohort in this review.

Limitations of this systematic review are acknowledged and highlighted by the MINORS scores of studies ranging from 7 to 11. The included studies provide only level IV evidence with common weaknesses being the low study numbers, retrospective nature of the studies and the lack of comparative groups. Heterogeneity of the underlying tumour diagnosis and reporting of outcome measures precludes data synthesis and restricts direct comparison of results. Determining post-operative improvements proved difficult as this requires pre-operative measurements which were commonly unavailable. The studies included were published from 1999 to 2019; no time filter was applied to the search in order to improve the volume of studies and patients available for analyses. However, this meant that the effect of evolving and improving surgical technique is difficult to factor into the overall outcomes. The range of patient follow-up is also very large (1–372 months) and the true long-term outcomes of elbow arthroplasty for tumours are therefore hard to quantify.

This systematic review has identified a significant demand for high-quality prospective studies in order to better report and compare the outcomes of different elbow arthroplasty options in oncological patients.

Conclusions

Elbow prosthesis reconstruction after tumour resection can provide good functional outcomes at mid-term follow-up. However, there is a high overall complication rate of 28% and a revision rate of 14%. Although high, the complication and revision rates are comparable to other indications for elbow arthroplasty including rheumatoid arthritis (which is the commonest indication), acute trauma and traumatic sequelae. However, in order to effectively, scientifically compare the functional outcome of different elbow arthroplasty options amongst both the tumour and non-tumour groups, we conclude that more robust prospective studies are required.

It is recommended that all patients are discussed at a Multi-Disciplinary Meeting that includes a specialist elbow reconstructive surgeon as well as a surgeon with an oncological interest. There should be robust surveillance methods and audits with data collection that includes outcome scores.

Footnotes

Acknowledgements

The authors would like to thank the Frances-Costello library for their assistance with the electronic literature search.

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.

Guarantor

SMH.

Contributorship

GM –PRISMA process, data extraction and manuscript first draft; RWJ – manuscript first draft; GO – PRISMA process; MSC – data extraction from articles; PC – statistics, revision manuscript and SMH – conception and manuscript final review

ORCID iDs

Gopikanthan Manoharan

Robert W Jordan

References

Sperling

Pritchard

Morrey

. Total elbow arthroplasty after resection of tumors at the elbow. Clin Orthop Relat Res 1999; 367: 256–261.

Savvidou

Koutsouradis

Chloros

, et al. Bone tumours around the elbow: a rare entity. EFORT Open Rev 2019; 4: 133–142.

Wong

Abraham

. Upper extremity considerations for oncologic surgery. Orthop Clin North Am 2014; 45: 541–564.

Scaglioni

Chang

Gur

, et al. The role of the fibula head flap for joint reconstruction after osteoarticular resections. J Plast Reconstr Aesthet Surg 2014; 67: 617–623.

Urbaniak

Black

Jr . Cadaveric elbow allografts: a six-year experience. Clin Orthop Relat Res 1985; 197: 131–140.

Casadei

De Paolis

Drago

, et al. Total elbow arthroplasty for primary and metastatic tumor. Orthop Traumatol Surg Res 2016; 102: 459–465.

Kulkarni

Fiorenza

Grimer

, et al. The results of endoprosthetic replacement for tumours of the distal humerus. J Bone Joint Surg Br 2003; 85: 240–243.

Hay

Kulkarni

Watts

, et al. The provision of primary and revision elbow replacement surgery in the NHS. Shoulder Elbow 2018; 10: S5–S12.

Moher

Liberati

Tetzlaff

, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009; 6: e1000097–e1000097.

10.

Slim

Nini

Forestier

, et al. Methodological index for non-randomized studies (MINORS): development and validation of a new instrument. ANZ J Surg 2003; 73: 712–716.

11.

Athwal

Chin

Adams

, et al. Coonrad-Morrey total elbow arthroplasty for tumours of the distal humerus and elbow. J Bone Joint Surg Br 2005; 87: 1369–1374.

12.

Schwab

Healey

Athanasian

. Wide en bloc extra-articular excision of the elbow for sarcoma with complex reconstruction. J Bone Joint Surg Br 2008; 90: 78–83.

13.

Peterson

Villalobos

Zamora

, et al. Limb sparing resection for tumors involving the distal humerus and reconstruction with a modular endoprosthesis. Bull Hosp Jt Dis 2015; 73: 190–197.

14.

Capanna

Muratori

Campo

, et al. Modular megaprosthesis reconstruction for oncological and non-oncological resection of the elbow joint. Injury 2016; 47: 78–83.

15.

Henrichs

Liem

Gosheger

, et al. Megaprosthetic replacement of the distal humerus: still a challenge in limb salvage. J Shoulder Elbow Surg 2019; 28: 908–914.

16.

Rolf

Gohlke

. Endoprosthetic elbow replacement in patients with solitary metastasis resulting from renal cell carcinoma. J Shoulder Elbow Surg 2004; 13: 656–663.

17.

Hanna

David

Aston

, et al. Endoprosthetic replacement of the distal humerus following resection of bone tumours. J Bone Joint Surg Br 2007; 89: 1498–1503.

18.

Tang

Guo

Yang

, et al. Custom-made prosthesis replacement for reconstruction of elbow after tumor resection. J Shoulder Elbow Surg 2009; 18: 796–803.

19.

Balasubramanian

Gnanasundaram

Prakasam

. Endoprosthetic reconstruction of distal humerus following resection of distal humeral giant cell tumours in six patients in rural India. Malays Orthop J 2017; 11: 25–29.

20.

Ross

Sneath

Scales

. Endoprosthetic replacement of the humerus and elbow joint. J Bone Joint Surg Br 1987; 69: 652–655.

21.

Weber

Lin

Yasko

. Complex segmental elbow reconstruction after tumor resection. Clin Orthop Relat Res 2003; 415: 31–44.

22.

Bickels

Kollender

Wittig

, et al. Function after resection of humeral metastases: analysis of 59 consecutive patients. Clin Orthop Relat Res 2005; 437: 201–208.

23.

Torbert

Fox

Hosalkar

, et al. Endoprosthetic reconstructions: results of long-term followup of 139 patients. Clin Orthop Relat Res 2005; 438: 51–59.

24.

Funovics

Schuh

Adams

Jr , et al. Modular prosthetic reconstruction of major bone defects of the distal end of the humerus. J Bone Joint Surg Am 2011; 93: 1064–1074.

25.

Hossein

Ashraf

Peivandi

. Total allograft transplantation of the elbow joint after wide resection of synovial cell sarcoma: a case series. Transplant Proc 2011; 43: 652–654.

26.

Sewell

Hanna

Pollock

, et al. Proximal ulna endoprosthetic replacement for bone tumours in young patients. Int Orthop 2012; 36: 1039–1044.

27.

Sorensen

Gregersen

Grum-Schwensen

, et al. Patient and implant survival following joint replacement because of metastatic bone disease. Acta Orthop 2013; 84: 301–306.

28.

Pradhan

Reddy

Grimer

, et al.

Osteosarcomas in the upper distal extremities: are their oncological outcomes similar to other sites?

Eur J Surg Oncol 2015; 41: 407–412.

29.

Pruksakorn

Chantarapanich

Arpornchayanon

, et al. Rapid-prototype endoprosthesis for palliative reconstruction of an upper extremity after resection of bone metastasis. Int J Comput Assist Radiol Surg 2015; 10: 343–350.

30.

Chebli

Crimi

Letson

, et al. Factors influencing outcomes of endoprosthetic reconstruction of the distal end of the humerus. Curr Orthop Pract 2019; 30: 123–128.

31.

Welsink

Lambers

KTA

van Deurzen

DFP

, et al. Total elbow arthroplasty: a systematic review. JBJS Rev 2017; 5: e4–e4.

32.

Cross

Cicalese

Nam

, et al. Results of custom-fit, noncemented, semiconstrained total elbow arthroplasty for inflammatory arthritis at an average of eighteen years of follow-up. J Shoulder Elbow Surg 2014; 23: 1368–1373.

33.

National Joint Registry. National Joint Registry for England, Wales and Northern Ireland; annual reports, https://reports.njrcentre.org.uk/elbows-primary-procedures-activity (2020, accessed 10 January 2021).

34.

Samdanis

Manoharan

Jordan

, et al. Indications and outcome in total elbow arthroplasty: a systematic review. Shoulder Elbow 2019; 12: 353–361.

35.

Morrey

Adams

Bryan

. Total replacement for post-traumatic arthritis of the elbow. J Bone Joint Surg Br 1991; 73: 607–612.

36.

Sanchez-Sotelo

Baghdadi

Morrey

. Primary linked semiconstrained total elbow arthroplasty for rheumatoid arthritis: a single-institution experience with 461 elbows over three decades. J Bone Joint Surg Am 2016; 98: 1741–1748.

37.

Mansat

Nouaille Degorce

Bonnevialle

, et al. Total elbow arthroplasty for acute distal humeral fractures in patients over 65 years old – results of a multicenter study in 87 patients. Orthop Traumatol Surg Res 2013; 99: 779–784.