Allograft-prosthetic composite versus megaprosthesis for proximal humerus reconstruction after tumor resection: a meta-analysis of clinical outcomes

Article information

Clin Shoulder Elb. 2025;28(3):298-305

Publication date (electronic) : 2025 August 26

doi : https://doi.org/10.5397/cise.2025.00388

Mohammad Daher¹, Mohamad Y. Fares¹, Steve S Gill², Peter Boufadel¹, Andrew R Jensen³, William C Eward⁴, Adam Z Khan⁵, John G Horneff⁶, Joseph A. Abboud^,1

¹Division of Shoulder and Elbow Surgery, Rothman Orthopaedic Institute, Thomas Jefferson Medical Center, Philadelphia, PA, USA

²Department of Emergency Medicine, University of California, Los Angeles, Los Angeles, CA, USA

³David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA, USA

⁴Department of Orthopedic Surgery, Duke University, Durham, NC, USA

⁵Department of Orthopedic Surgery, Southern California Permanente Medical Group, Panorama City, CA, USA

⁶Division of Shoulder and Elbow Surgery, Department of Orthopedics, University of Pennsylvania, Philadelphia, PA, USA

Corresponding Author: Joseph A. Abboud Division of Shoulder and Elbow Surgery, Rothman Orthopaedic Institute, 925 Chestnut St, Philadelphia, PA 19107, USA Tel: +1-610-547-8351, Email: abboudj@gmail.com

Received 2025 April 7; Revised 2025 May 14; Accepted 2025 May 22.

Abstract

INTRODUCTION

The proximal humerus is the most common site for primary bone tumors in the upper limb and the fourth most common site in the human body for primary bone tumors [1]. Tumors in the proximal humerus range from simple bone cysts, which can be treated conservatively, to high-grade bone sarcomas that might require complex oncological limb-sparing surgeries or even amputation. The shoulder region comprises 15% of all primary sarcomas and is the third most common site for such tumors, following the hip-pelvis and knee [2]. Tumors of the proximal humerus can present with non-specific symptoms, which leads to an average 6-month delay in diagnosis of the primary condition [2]. Once diagnosed, the managing team should work with the patient to establish realistic goals and expectations, which tend to coalesce around the following objectives: removal of the tumor mass with wide margins, spare the limbs, and maintain length and shape to allow optimal functionality [1].

Although there are numerous reconstruction options after surgical resection in the proximal humerus, two of the mainstay options are allograft prosthetic composite (APC) and megaprosthesis (MP). While APC replaces resected or compromised bone with a combination of shoulder prosthesis and bulk allograft [3], reconstructions involving MP commonly utilize modular implants to restore anatomy and functionality after massive bone loss. MP is typically the reconstructive technique of choice due to its low complication rate of reconstruction [4]. Nevertheless, MP can limit shoulder function. However, in recent years, MP reconstruction of the proximal humerus in combination with a reverse total shoulder design has addressed this issue. The APC reconstruction can improve function by repairing the rotator cuff and restoring the deltoid wrap [4]. However, subluxation is common with the latter [4].

Consensus has not been reached regarding the ideal method of reconstruction, with reviews of the literature showing that differing reconstruction techniques lead to similar functional results [5,6]. Though there have been efforts to compare these two methods, the occurrence is rare, and most studies have been underpowered. While there was a previous meta-analysis performed by Gautam et al. [7], the study evaluated several locations instead of focusing on the proximal humerus alone. Furthermore, only six of their included studies compared APC to MP at the proximal humerus. For this reason, a meta-analysis of the existing literature should compare APC to MP following tumor resection of the proximal humerus. Based on the findings in the literature, we hypothesize that APC will have better functional outcomes and a higher rate of revisions than MP.

METHODS

Search Strategy

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, PubMed, Cochrane, and Google Scholar (pages 1–20) were searched through September 2024 [8] to find articles comparing APC to MP for proximal humerus reconstruction after tumor resection. The following keywords and Boolean terms were used: “Shoulder,” “Proximal Humerus,” “Allograft prosthetic composite,” “Endoprosthesis,” and “Megaprosthesis.” Supplementary articles were added based on reference lists from articles and Internet searches. One author extracted the data and another confirmed the choice of the included articles. The process is summarized in the PRISMA flowchart (Fig. 1).

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart for article selection.

The inclusion criteria consisted of studies comparing APC to MP for proximal humerus reconstruction after tumor resection. Studies including either reverse shoulder arthroplasty (RSA) or hemiarthroplasty constructs will be included in this study. However, there is an insufficient number of patients to include a subanalysis of these two techniques. The exclusion criteria consisted of non-comparative studies or studies comparing these two procedures in the setting of proximal humerus reconstruction performed for non-oncologic etiologies.

RESULTS

Characteristics of the Included Studies

Ten retrospective studies met the inclusion criteria [4,10-18]. These studies included 400 patients, with 115 in the APC group (29%) and 285 in the MP group (71%). Four studies reported the age of each of their two groups, with the APC group having a mean age of 45.2 years and the MP group having a mean of 53.2 years. The main characteristics of the included studies are summarized in Table 1.

Table 1.

Characteristics of the included studies

Adverse Events

Eight studies comprising 294 patients reported complication data (98 in the APC group and 198 in the MP group), and nine studies comprising 398 patients reported reoperation data (118 in the APC group and 280 in the MP group). A higher rate of reoperations (OR, 2.50; 95% CI, 1.40–4.45; P=0.002) (Fig. 2A) was observed in the APC group at an average follow-up of 7.0 years. However, there was no difference in the risk of complications (OR, 1.99; 95% CI, 0.43–9.34; P=0.38) (Fig. 2B). Furthermore, when looking at specific complications, there was no difference in number of infections (OR, 1.05; 95% CI, 0.45–2.42; P=0.92) (Fig. 2C), subluxations (OR, 1.03; 95% CI, 0.62–1.73; P=0.90) (Fig. 2D), or dislocations (OR, 0.98; 95% CI, 0.28–3.44; P=0.97) (Fig. 2E) between the groups.

Fig. 2.

Forest plots showing the differences in rates of (A) reoperations, (B) complications, (C) infections, (D) subluxations, and (E) dislocations. APC: allograft prosthetic composite, MP: megaprosthesis, M-H: Mantel-Haenszel method.

Patient-Reported Outcome Measures

Six studies including 206 patients reported MSTS scores (73 in the APC group and 133 in the MP group). Higher MSTS scores were seen in the APC group compared to the MP group (MD, 3.73; 95% CI, 1.37– 6.08; P=0.002) (Fig. 3).

Fig. 3.

Forest plot showing the difference in Musculoskeletal Tumor Society (MSTS) score. APC: allograft prosthetic composite, MP: megaprosthesis, SD: standard deviation, IV: inverse variance method.

Range of Motion

Two studies including 97 patients reported postoperative FF (34 in the APC group and 63 in the MP group). A better FF was seen in the APC group compared to the MP group (MD, 10.11; 95% CI, 5.33– 14.90; P<0.001) (Fig. 4).

Fig. 4.

Forest plot showing the difference in forward flexion. APC: allograft prosthetic composite, MP: megaprosthesis, SD: standard deviation, IV: inverse variance method.

DISCUSSION

Several surgical approaches exist to reconstruct the proximal humerus after resection due to primary or secondary tumors. Two of the most commonly used surgeries are APC and MP [4]. MP is typically the reconstructive technique of choice due to its low complication rate of reconstruction [4]. However, MP can limit shoulder function, especially in its traditional form in which the humerus is reconstructed as a “suspended hemiarthroplasty.” This limitation has resulted in an interest in APC reconstructions, which can improve function by repairing the rotator cuff [4]. Nevertheless, subluxation was common with APC due to the deficient soft tissue and bony constraints [4]. The present meta-analysis shows that APC leads to better FF and MSTS than does MP. However, APC also had a higher reoperation rate with no difference in the rate of complications compared to those with MP at a mean follow-up of 7.0 years.

There was no difference in overall or specific complications including infections, subluxation, and dislocations between the APC and MP groups. However, the APC group had a higher rate of reoperations than did the MP group. These findings are in line with those of Gautam et al. [7], who found that deep infections were the most common complications. However, the most common complication in the present APC group was subluxation (22.7%), followed by graft resorption (17.0%), dislocations (9.1%), and periprosthetic fractures (6.8%). The most common complications in the MP group also was subluxations (21.0%), though they were followed by dislocations (9.7%), cortical resorption (5.1%), and infections (4.6%), with the latter having a lower rate than in the APC group (5.7%) (Table 2). These findings could explain the higher rate of reoperations in the APC group, since the APC group had high rates of graft resorptions and periprosthetic fractures, both of which require intervention.

Table 2.

Number of complications

As for functional outcomes, the APC group had better MSTS scores than did the MP group, which was also reported by Gautam et al. [7]. A potential explanation could be preservation of the rotator cuff and the deltoid wrap and their insertion to the allograft [3,11]. In fact, biological reinsertion of periarticular muscles on an allograft is important since insertion on a metal body in MP is often unsuccessful [19]. Another explanation of the better MSTS scores in the APC cohort could be that the patients were younger (by a mean of 10 years) compared to those in the MP cohort. The latter also could justify the better FF achieved in the APC group. However, not only were two studies included in this analysis, but the MD in FF between the two groups (10°) did not exceed the minimum clinically important difference of 12° [20].

The clinical choice between APC and MP should be carefully individualized. For younger, high-demand patients in whom maximum function is prioritized and the patient can accept the possibility of future revision surgery, APC might offer superior functional results. In contrast, for older or lower-demand patients or those in whom surgical durability and fewer reoperations are the priority, MP can be the more appropriate option. The emerging role of RSA combined with MP designs also deserves emphasis. It is suggested that RSA-MP constructs may combine the durability of MP with improved functional restoration. As surgical technology evolves, this hybrid approach may eventually bridge the gap between APC and MP.

Future studies should focus on large, prospective, multi-center registries or randomized controlled trials to mitigate the selection bias inherent to retrospective designs. Subgroup analyses based on patient age, tumor type, soft tissue involvement, and use of RSA constructs are essential to better guide patient-specific recommendations. Additionally, longer-term follow-up is necessary to assess the survivorship of both implants and functional outcomes over decades rather than years. Finally, standardized reporting of complications and outcomes across future studies would greatly enhance the quality of evidence and the ability to draw clinically actionable conclusions. As the field moves toward personalized oncologic reconstruction, incorporating patient-reported outcomes and quality-of-life measures will be crucial to truly define the "best" reconstruction strategy.

The present study has several limitations. As a systematic review, findings depend on the variables examined by the studies meeting inclusion criteria. Furthermore, the results of this meta-analysis are pooled, as granular data were not available; therefore, we were unable to perform subgroup analysis based on demographics such as age. However, despite being older, the MP group had a lower rate of reoperations. A unique limitation of this study, and knowledge on this topic in general, is the inability to perform subgroup analysis on patients who undergo a megaprosthetic reconstruction with a reverse total shoulder construct. Surgeons performing these operations are hopeful that this technique will allow the benefits of MP reconstruction with the added ROM and functional benefits typically associated with APC reconstruction. As the number of studies of this specific reconstruction type increases, such an analysis will be performed. Another limitation would be the quality of the included studies as they were all retrospective and non-randomized; therefore, bias from surgeon selection or patient resection length could have played a role in the differences in outcomes between groups. Last, we were only able to assess FF and were not able to look into other ROM measurements, as these were not reported in the included studies.

CONCLUSIONS

The present meta-analysis showed that patients who underwent APC for proximal humerus reconstruction had better MSTS scores and FF compared to patients who underwent MP. However, those who underwent APC had a higher reoperation rate than those who underwent MP. This study demonstrates the efficacy of the two options for proximal humerus reconstruction. Nevertheless, the clinical superiority of APC regarding MSTS score has not been demonstrated, indicating MP as a better option due to its lower rate of reoperations.

Article information Continued