INTRODUCTION
Total elbow arthroplasty (TEA) is used to address various arthritic conditions as well as acute fractures in the elderly that are not suitable for fixation and can provide satisfactory results for properly selected, low-demand patients [1-3]. When TEAs fail, they are often accompanied by bone loss, necessitating more extensive bony reconstruction during revision surgeries [4,5]. Addressing this bone loss in revision TEAs has remained a challenge due to the limited availability of viable reconstructive options and the relative scarcity of focused literature evaluating techniques designed to manage extensive bone loss.
In patients with severe humeral or ulnar bone loss, the use of an allograft–prosthesis composite (APC) has been used as a reconstructive option. As an APC provides a tailor-made reconstruction that biologically integrates with the patient’s anatomy, it remains a viable reconstructive option for long-term stability in TEA with large bone defects [6-8]. However, multiple small series have reported high failure rates when using APCs [9,10]. Given that an APC is often necessary for addressing bone loss in TEA, further investigation is needed to evaluate its effectiveness and the outcomes associated with this technique. The aim of this review was to summarize outcomes of APC use in revision TEA with significant bone loss.
METHODS
We conducted this systematic review under the guidelines set by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). This study did not involve human participants or identifiable data; therefore, institutional review board approval and informed consent were not required.
Eligibility Criteria
We included original studies reporting clinical outcomes in patients undergoing TEA with the use of an APC construct. Studies that were duplicates, reviews or meta-analyses, a case report or that reported fewer than five patients, a commentary or editorial, or a biomechanical or cadaveric study, along with those that had only their abstract available; included non-adult (<18 years) patients; reported no clinical outcomes; included patients undergoing elbow hemiarthroplasty without differentiating from APC TEA cases; included oncologic disorders as an indication for TEA; or that reported standard non-APC TEA cases without differentiating from APC cases were excluded.
Search Strategy
The databases PubMed/Medline, Embase, Web of Science, and the Cochrane Database of Systematic Reviews were queried from inception to October 2023 for articles evaluating TEA and documenting postoperative clinical outcomes. A title search was used with the following terms: (“total elbow” OR “elbow replacement” OR “elbow arthroplasty” OR “total elbow arthroplasty” OR “elbow prosthesis”). The specific search strategies used for each database can be found in the Supplementary Material 1. Initially, articles were screened by title and abstract for exclusion by two authors (TRB and VEB). Then, full-text screenings for exclusion were conducted by three authors (AHAK, TRB and VEB). Finally, further text screenings of all TEA articles were performed by a single author to isolate articles assessing TEA with the use of an APC (TRB). If inclusion eligibility was questionable, authors erred on the side of inclusion. A pair of senior shoulder and elbow surgeons (JOW and JJK) offered expert guidance to assist in resolving inclusion decisions when screeners faced uncertainty.
Outcome Measures and Data Extraction
Data extraction was conducted by one author (AHAK) using a standardized data-collection form. The extracted article characteristics included the country of the study and the level of evidence. Patient demographics were collected, including age at the time of surgery, sex, mean follow-up, and the number of elbows involved. Surgical characteristics were also extracted, including indication for primary TEA and cause of bone loss necessitating the use of an APC. Details of the APC were reported, including allograft interface (humeral versus ulnar); union rates (defined by each individual study); and failures, including allograft non-union or loosening. The outcomes included in this study were the Mayo Elbow Performance Score (MEPS) [11], elbow flexion–extension arc, and the incidence of complications associated with allograft non-union or loosening.
Risk-of-Bias Assessment
The methodological quality of included studies was assessed by a single investigator (AHAK) using the Methodological Index for Non-Randomized Studies (MINORS) criteria. Although interobserver reliability was not formally calculated, the criteria were applied according to standard MINORS guidelines to ensure methodological consistency across included studies. For the MINORS scores, refer to Supplementary Table 1.
RESULTS
The search yielded 2,070 publications, of which 833 unique publications remained after removing duplicates (Fig. 1). Following abstract and full-text screening—during which 411 and 417 articles were excluded, respectively—five studies met the inclusion criteria for final analysis (Fig. 1).
Comparative Clinical Studies
The five included articles included 85 elbows that were treated with TEA using an APC for reconstruction due to extensive bone loss [9,10,12-14]. Across these 85 elbows, there were 58 humeral allografts and 32 ulnar allografts, totaling 90 allografts, with five of the included elbows having both humeral and ulnar allografts. Of the 85 TEAs, 77 APCs had radiographic follow-up data available following evaluations of 50 out of 58 humeral and all 32 ulnar allografts, with union achieved in 60% of the humeral and 84% of the ulnar reconstructions, respectively (Table 1) [9,10,12-14]. The most common complications found in TEA APC reconstructions were loosening and infections (Table 2) [9,10,12-14]. Postoperative range of motion (ROM) values were reported by four of the included studies for a total of 70 elbows and were calculated to be a mean active extension of 24° (standard deviation [SD], 19°) and mean active flexion of 120° (SD, 20°) (Table 3) [9,10,12,13]. Postoperative MEPS values were reported by four of the included studies, accounting for 70 elbows, yielding a mean of 64.2 points (SD, 21.8 points) (Table 3).
Amirfeyz et al. [9] studied 11 elbows in 10 patients who underwent APC reconstructions of prior failed TEAs with massive structural bone loss. Union was observed in one of the six humeral allografts and in seven of the eight ulnar allografts. The 11 revision TEA APC reconstructions yielded six postoperative complications (Table 2). Of these complications, only the periprosthetic joint infection required reoperation, and implant removal was performed 26 months after the index procedure. The mean postoperative extension and flexion of the 11 elbows was 8°±13° and 100°±31°, respectively. Separately, the mean postoperative MEPS of this cohort was 74.1±19.2 points.
Burnier et al. [12] studied 10 elbows in 10 patients with failed prior TEAs who were treated with an APC technique that employed a structural proximal ulnar allograft with its attached triceps as an APC to reconstruct the proximal ulna and preserve the extensor mechanisms. At a mean radiographic follow-up of 41 months, eight of the 10 ulnar APCs showed allograft union. Meanwhile, the 10 revision TEA APC reconstructions yielded six postoperative complications, with all six requiring additional surgery (Table 2). The mean postoperative extension and flexion of nine elbows was 31°±17° and 122°±11°, respectively, and the mean postoperative MEPS of the same nine elbows was 73.3±18.2 points. Cheema et al. [13] studied 41 revision TEAs in 41 patients, performed with humeral APC reconstruction. Of these, 33 elbows made it to radiographic follow-up, and 20 showed allograft union. The 41 revision TEA APC reconstructions yielded 13 postoperative complications of aseptic loosening (12 humeral, 1 ulnar), all of which required revision surgery. The mean postoperative extension and flexion of 37 elbows were 26°±23° and 124°±23°, respectively, and the mean postoperative MEPS of the same 37 elbows was 53±23 points.
Mansat et al. [10] reported 13 failed TEAs in 13 patients as a result of ulnar or humeral component loosening that underwent reconstruction with the use of an APC. Allograft union was reported in two of four humerus allografts and in all nine of the included ulnar allografts. The 13 revision TEA APC reconstructions yielded five postoperative complications, four of which were chronic infections (30.8%); all ultimately required surgical intervention (Table 2). The mean postoperative extension and flexion of the 13 revised elbows were 28°±17° and 125°±12°, respectively, and their mean postoperative MEPS was 67±24.5 points.
Renfree et al. [14] studied 10 elbows in 10 patients that underwent composite TEA. At radiographic follow-up, allograft union was reported in all seven humeral allografts and in three of the five ulnar allografts. The 10 revision TEA APC reconstructions yielded four postoperative complications (Table 2). One elbow showed failure following the APC reconstruction and required implant removal due to infection 26 months after the index procedure. This study did not report any ROM or functional outcome scores.
DISCUSSION
APC reconstruction is a useful technique in revision TEA with extensive bone loss, as this systematic review of five articles shows: four reported mean postoperative ROM and patient-reported outcome measures across 70 TEA APC reconstructions, yielding an average active elbow arc of 24°–120° and MEPS of 64 points. Despite functional ROM, across all included articles, the complication rate was 39%, and the overall non-union rate was 31%.
The total mean ROM figures across the four studies were 24° (active extension) and 120° (active flexion) (Table 3) [9,10,12,13]. These values produce a similar flexion–extension arc to what is reported in the broader literature on TEA revisions. In a systematic review by Geurts et al. [4], which included 21 articles and 532 cases of revision TEAs, the final postoperative ROM figures were 29° of active extension and 128° of active flexion [4,9,10,12,13]. However, the APC TEA flexion–extension arc is notably smaller than that reflected in the literature for primary TEA. In contrast, Davey et al. [15], in a systematic review looking at primary TEA outcomes across 23 studies that included 1429 elbows, found a mean active extension of 24° and active flexion of 135°. This comparison underscores the greater challenges TEA APC revisions face in restoring ROM, likely due to significant bone loss and the complexity of revision surgeries involving APCs [16]. However, although less than that achieved with primary TEA, the reported TEA APC ROM demonstrates sufficient extension–flexion for most activities of daily living, aligning with literature suggesting that a 100° flexion–extension arc (approximately 30°–130°) is necessary for most functional daily tasks [17].
Separately, the average postoperative MEPS after APC reconstruction for revision TEA was 64 points (Table 3) [9,10,12,13]. The MEPS score reflects four key patient outcomes: pain, ROM, stability, and function. A score below 60 points indicates severe deficiencies across all these postoperative measures, while a score between 60 and 74 points, as observed in the TEA APC elbows, suggests some improvement but persistent limitations in pain, function, and mobility [18]. However, this average MEPS for APC TEA revisions is significantly lower than what is seen in the broader TEA revision literature; in the previously described systematic review by Geurts et al. [4], the mean MEPS was 80 points. Furthermore, the mean MEPS of APC TEA demonstrates an even greater disparity when compared to the MEPS of primary TEA. In the previously mentioned study by Davey et al. [15], primary TEA yielded a mean MEPS of 89 points. The lower MEPS score in the APC group suggests that, while this technique effectively addresses the structural challenges posed by significant bone loss, it may fall short in fully restoring functional capacity and full ROM [6-8]. Certainly, patients should be counseled regarding realistic expectations of functional outcomes and pain relief after revision TEA with an APC construct.
All five included studies reported postoperative complications following APC TEA revisions. Across these studies, 33 out of 85 elbows (39%) experienced complications (Table 2) [9,10,12-14]. This complication rate is consistent with prior reports of TEA revisions. In the systematic review by Geurts et al. [4], 232 elbows (44%) had at least one postoperative complication. However, the APC TEA complication rate was markedly higher than that seen in the 1,429 primary TEAs, reported by the previously mentioned study by Davey et al. [15] documenting at least one complication in 16% of elbows. The most common complications in the current APC TEA review were aseptic loosening (19 elbows, 22%) and infection (7 elbows, 8.2%)[9,10,12-14]. Comparatively, Geurts et al. [4] also identified aseptic loosening (232 elbows, 22%) as the single most frequent complication. The persistent issue of aseptic loosening across both groups highlights a common mechanical challenge in revision TEA, regardless of the reconstruction method [19,20]. While loosening was a persistent issue in both studies, higher infection rates appear to be a unique challenge for APC TEA reconstructions. The higher infection rate among APC TEA revisions (8.2%) is potentially due to the use of allografts and because the increased operative time associated with APC TEA may increase the risk of postoperative infection [21-23].
This study revealed a clear trend across all included reports, with ulnar allograft non-union occurring in 16% of cases and humeral allograft non-union occurring in 40%, revealing a significantly higher non-union rate among humeral reconstructions (Table 1) [9,10,12-14]. This suggests that graft incorporation is more challenging in the humerus, possibly due to differences in bone density with aging. Biological studies have hypothesized that, with aging, the bone mineral density of the humerus often declines more sharply than that of the ulna [24-26]. Non-union in these cases represents a critical complication, as it can lead to compromised implant stability, persistent pain, and reduced overall function [5,16,27]. These differences in ulnar and humeral allograft non-union rates highlight the importance of tailored approaches to revision TEA, depending on the site and extent of bone loss, to optimize outcomes and minimize complications. In a study examining surgical techniques and outcomes of proximal humeral osteoarticular allografts, Farfalli et al. [28] performed 19 allograft reconstructions using a combination of a long lateral plate and short anterior plate and reported no non-union cases in their cohort. While Farfalli et al. [28] performed allografts without a prosthesis, a similar dual-plating system should be strongly considered to improve union rates in distal humeral APC. Additionally, biological augmentation may support union rates in APC reconstructions, Cheema et al. [13] proposed either incorporating vascularized fibular autografts in an intramedullary capacity, similar to the Capanna tumor-reconstruction technique, or using a vascularized medial femoral condyle osteoperiosteal transfer wrapped around the APC–host junction to support union in APC reconstructions [29,30]. However, it is currently unknown whether the marginal benefits achieved through these techniques would outweigh the extended surgical time and increased risk of infections that may be encountered. Nevertheless, these augments to the APC reconstruction could help to achieve union for humeral allografts in appropriately selected cases but require careful consideration and discussion with patients.
When interpreted collectively, the findings from this current systematic review suggest that APC reconstruction should be reserved for revision TEA cases with severe bone loss in which alternative fixation options are not feasible, given its high complication and non-union rates. Additionally, preoperative planning should prioritize infection-mitigating strategies and the optimization of host bone quality to improve graft incorporation. Intraoperatively, fixation methods may need to be tailored to the graft site, particularly for humeral reconstructions, where additional plating or biologic augmentation may reduce non-union risk. Surgeons should anticipate modest functional recovery and counsel patients accordingly, emphasizing that the primary goal is pain reduction and structural preservation rather than full restoration of motion or strength. Overall, this study endorses a selective, individualized approach to APC TEA revision with careful patient selection and transparent communication about expected outcomes and potential complications.
This descriptive systematic review has inherent limitations. Most of the included studies were retrospective, allowing opportunity for compounding and reporting bias as well as publication bias. Furthermore, with five included studies, heterogeneity in study methodology and surgical techniques make it challenging to directly compare results. With a total sample size of only 85 elbows, the findings of this study may also lack generalizability when drawing conclusions. Despite our efforts to perform broad searches in largely accessible literature databases, it is possible that some relevant literature pertaining to APC TEA was not included. Notwithstanding these limitations, this systematic review provides insight into the characterization of APC TEA revision outcomes, a subject currently not covered in the literature. Future prospective comparative studies are needed to further evaluate the role of standardized fixation techniques and biologic augmentation methods to further optimize union rates and improve long-term functional outcomes in APC reconstruction for revision TEA.
CONCLUSIONS
This systematic review found that, while APC reconstruction offers a viable option for managing extensive bone loss in revision TEA, it remains burdened by substantially high complication rates, notably non-union, especially for humeral allografts. Thus, a selective and individualized approach is required for APC reconstruction in revision TEAs, with personalized surgical planning and realistic patient counseling regarding expected outcomes and complication risk.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement
Print
