Optimized prosthetic design, surgical technique, and cosmetic outcome in shoulder reconstruction with double-constrained implant following extra-articular shoulder resection: a proof-of-concept

Abstract

Reconstruction of the anatomic defect following extra-articular shoulder resection is a challenging problem, particularly in cases when function of the deltoid muscle and rotator cuff are compromised. Standard reconstruction techniques often result in either instability or rigidity. Constrained implants have been used to overcome these problems; however, they have been associated with a high rate of aseptic loosening. Recently, a novel double-constrained implant has been introduced, yielding promising functional results. Nonetheless, this implant exhibited a cosmetic defect related to protrusion of the humeral component that becomes apparent with time as result of surrounding muscle atrophy. An updated improved design of the implant has been developed to counteract this.

We report the case of a 15-year-old patient who underwent an extra-articular (Malawer type V) shoulder resection due to osteosarcoma and received an innovated custom-made double-constrained implant. Moreover, we describe a new modification of the Malawer utilitarian approach to the shoulder girdle that enhances tumor visibility and allows safer dissection. The patient recovered well with satisfactory outcomes at 18 months follow-up, highlighting the potential benefits of this implant design and surgical approach.

Introduction

Extra-articular resections of the shoulder girdle are complex procedures aimed at achieving adequate margins in tumors invading the shoulder joint [1, 2]. These surgeries frequently result in loss of deltoid muscle and rotator cuff function as a consequence of the radical surgical treatment [3, 4]. Endoprosthetic implants have become a benchmark in limb-sparing procedures for reconstructing the anatomic defect after an intricate shoulder resection [5], although the limb function may irreversibly be compromised following these procedures. Constrained implants provide greater shoulder joint stability [6, 7], however, a high degree of aseptic loosening has been reported [6, 8]. Recently, supported by biomechanical analyses, our team introduced a novel reconstructive option with a custom-made double-constrained implant designed to mitigate these complications [7]. Despite the reported promising functional outcomes, the implant exhibited a tendency to protrude anteriorly over time as atrophy of the denervated deltoid muscle developed, leading to cosmetic deformity.

In this report, we present an innovated updated implant design that was used in a young male patient with osteosarcoma of the proximal humerus. The improved design aims to resolve the cosmetic concerns while maintaining its original biomechanical advantages. Additionally, we discuss a modification of the Malawer approach employed for extra-articular shoulder resections [9].

Case report

Patient history and examination

A 15-year-old male presented to our clinic with a growing mass and worsening pain of the right shoulder over the past two months. The obtained radiographs showed a mixed (sclerotic and lytic) lesion of the proximal humerus with periosteal reaction and ossifying extraosseous component of the tumor (Fig. 1A). Subsequent magnetic resonance imaging (MRI) revealed a large mass originating in the proximal humerus, invading the neighboring muscles, including the rotator cuff tendons and muscles, deltoid muscle, and encasing the axillary nerve and surrounding the glenoid part of the scapula (Fig. 1B). An open biopsy confirmed the diagnosis of high-grade osteosarcoma. Consequently, the patient was referred to the oncology department for staging and neoadjuvant chemotherapy.

Fig. 1

Radiographs (A) and magnetic resonance imaging (B, C) of the right shoulder showing the extent of the tumor

Preoperative planning and implant design

Due to the extent of the tumor, an extensive extra-articular resection of the proximal humerus and adjacent glenoid of the scapula was necessary (i.e., type V resection according to Malawer) [4]. Considering the large soft tissue component, sacrifice of the axillary nerve and transection of the rotator cuff tendons was mandatory to achieve safe oncologic margins (i.e., subtype B resection according to Malawer) [4]. The anticipated absent function of both the deltoid muscle and rotator cuff met the criteria for reconstruction using a double-constrained implant [7]. The endoprosthesis consisted of a custom-made three dimensional (3D) printed scapular component and a new type of a modular humeral component, which were serially connected via a dumbbell piece connector (Fig. 2). The constrained joint design was also improved, allowing for a shorter middle dumbbell piece. The scapular component had a trabecular metal surface to enhance osseous integration and fixation. The new humeral component was stripped of any anchoring parts as there were no structures to reattach in this patient. These modifications from the original design permitted a sleek less protruding implant. Additionally, a newer design of the constraining mechanism was used, allowing for a shorter connecting dumbbell piece that still retained the same range of motion as in the original design (Fig. 3). A preliminary assessment of the endoprosthesis was carried out with the use of a 3D printed plastic dummy, and after approval of its feasibility the definitive implant was manufactured in six weeks. The final implant was 3D printed from a Ti64Al4V titanium alloy with a diamond-like carbon (DLC) coating (ProSpon Company, Czech Republic). Socket inlays were fabricated with Polyether ether ketone (PEEK) polymer and the connecting dumbbell was made of polished steel.

Fig. 2

Illustration of the new implant design consisting of custom scapular component (A), modular humeral component (B), and connecting dumbbell piece (C)

Fig. 3

Graphic model of the innovated double-constrained implant comparing the volume of the older (above) design versus the new design (below). *Theor COR: theoretical center of rotation

The new design was completely stripped of all non-essential parts such as anchoring holes with the intent to reduce its final volume. The new locking mechanism allowed a shortening of the dumbbell piece from 28 mm to 22 mm (center to center) (Fig. 3). The distance from the medial center of rotation to the most protruding part of the implant was reduced by 8.8 mm from 48.8 mm to 40.0 mm (18.0% reduction). The total length of the dumbbell design was reduced by 7.5 mm from 46.4 mm to 38.9 mm (16.2% reduction). The stripped humeral component had a 36.0% reduction in its final size (22,450.1 mm³ vs. 35,076.5 mm³). The mass and volume reduction resulted in the component approximating a conical shape extending from the medial center of rotation, effectively reducing the total component volume by 40.0% (Fig. 3).

Surgical procedure

The surgery was performed under general anesthesia with the patient in a beach chair position. A modified Malawer approach was used to permit visualization and safe dissection of the tumor [9]. An anterior incision was made over the deltopectoral groove extending distally along the anterior aspect of the arm. The biopsy site was excised and left in continuity with the specimen. The incision was further extended posteriorly following the lateral border of the scapula, approximately up to the middle of its length. Next, the coracoid process and lateral part of the clavicle were exposed. The coracoid process was drilled from the apex to the base, and a subsequent osteotomy was performed through the middle of its length (Fig. 4). An osteotomy of the clavicle diaphysis was also performed. We then exposed the isthmus of the scapular spine in a similar fashion, after trialing and pre-bending a locking compression plate (LCP); an osteotomy at the base of acromion process was also carried out. The osteotomies allowed ample and safe visualization of the dorsal aspect of the shoulder joint (Fig. 4). Muscles of the rotator cuff were dissected at the level of scapular neck exposing the dorsal aspect of the scapula. The axillary neurovascular bundle was identified anteriorly in close proximity to the tumor. The circumflex bundle could not be safely dissected off the tumor, therefore the axillary nerve had to be sacrificed. Subsequently, an osteotomy of the scapular neck was carried out using a custom-made disposable cutting guide. After the standard dissection of the humeral site, the specimen was mobilized and resected.

Fig. 4

Illustration of the new modified approach demonstrating the line (dotted line) of the extended skin incision (A), the additional osteotomies to the clavicle, coracoid and acromion allowing an enhanced visualization of the tumor area (B), and final fixation of the clavicle and acromion osteotomies with plate and screws (C). © Copyright 2025 by The Curators of the University of Missouri, a public corporation

The reconstructive phase began with preparing the infraspinous fossa and placement of the custom-made scapular component, which was then fixed with locking screws achieving good primary stability. The humeral component was then cemented with the standard technique. The stem was anchored with two interference screws through a short plate to improve rotational resistance. PEEK inlays were inserted into the articulating sites, and the dumbbell was locked in place according to the manufacturer’s manual. Open reduction and internal fixation of the clavicle and acromion was achieved with the use of pre-drilled and pre-bend LCPs with perfect anatomic restoration (Fig. 4). The coracoid process was reinserted with a cancellous screw into a pre-drilled hole. The procedure lasted 240 min, with a total estimated blood loss of 600 ml.

Follow-up

The final pathology report confirmed negative resection margins and the diagnosis of conventional osteosarcoma, that exhibited over 95% necrosis following neoadjuvant chemotherapy. The patient was periodically followed with radiographic imaging (Fig. 5). Passive range of motion exercises started after six weeks. There were no healing complications, and the distal extremity was neurovascularly intact. At three-month follow-up the implant demonstrated solid fixation with the adjacent bones and the osteotomy sites were fully healed. At 18-month follow-up the patient was able to use his right upper extremity for daily living activities and remained disease- and pain-free. On exam, there was no sign of deformity at the anterior shoulder region, a commonly observed cosmetic problem with the older implant design. (Fig. 6). The patient had a stable shoulder joint allowing for full strength in elbow flexion and he was able to freely position the hand in space. Passive arm flexion and abduction were achieved with the assistance of the contralateral limb. Musculoskeletal Tumor Society (MSTS) score was 22/30; and Disabilities of Arm, Shoulder, and Hand (DASH) score was 30.8/100 at last follow-up.

Fig. 5

Radiographs obtained immediately after surgery (A) and at one year follow-up (B)

Fig. 6

Photographs of the scar three weeks (A, B) and one year after the surgery (C, D)

Discussion

Shoulder girdle is a common site for malignant and aggressive benign tumors [1, 5]. Depending on the extent of the tumor, extra-articular shoulder resection may be indicated in as many as one third of the cases frequently resulting in severely compromised shoulder function due to the impairment of the abductor muscle and shoulder stabilizers [1, 3, 4]. The goal of the reconstructive strategies following an extra-articular resection is to provide a stable and painless shoulder that can support a fully functional forearm and hand. Various options for limb-salvage surgery have been proposed, including endoprosthetic reconstruction, autograft, massive allograft, allograft-prosthesis composite, arthrodesis, or clavicula pro humero [1, 2, 5, 6, 10, 11]. Each technique presents a unique set of limitations and potential complications, with shoulder instability being the most common concern [3, 10]. To address this challenge, constrained shoulder implants were developed to counteract dislocation forces [6, 8]. However, these implants demonstrate a higher-than-expected risk of revision due to their susceptibility for aseptic loosening of the glenoid component [6]. In line with the presented case, recent evidence suggests that double-constrained implants may be a solution to this problem due to its unique biomechanical properties that shield the glenoid component from transferred forces from the arm, rendering them less prone to loosening [7].

The dumbbell double bearing component translates the sheering forces to only frictional forces, hence, effectively reducing the acting torque on the glenoid component and compensating for the deficient soft tissue stabilizers [7]. This underscores the utility of the double-constrained implant in improving the stability and longevity of shoulder reconstruction whilst preserving free passive motion allowing for ample positioning of the hand in space improving the function of upper extremity in daily activities.

Custom 3D printing is the cutting edge of modern orthopaedic oncology, offering innovative solutions for endoprosthetic reconstructions when combined with advanced materials. By tailoring the design of the endoprosthesis to the patient’s unique anatomy and anticipated soft tissue and bony defects, we can achieve a precise fit that surpasses the limitations of standard off-the-shelf implants with an improved function. The glenoid component can be then tailored to achieve maximum purchase in the remaining bone further improving the durability of such reconstruction. Additionally, the bone-contacting surfaces were fabricated with a porous coating material to promote robust biological integration. The optimized design of the proximal humeral component and locking mechanism of the dumbbell piece minimized implant protrusion and resulted in a more aesthetically pleasing outcome to the patient, compared to the original prosthetic design [7]. Furthermore, the application of DLC coating reduces the risk of adverse biological reactions, and its smooth surface inhibits bacterial adhesion, thereby limiting the ability of pathogens to colonize and form biofilms, which is highly beneficial for oncologic patients [7, 12].

The utilitarian approach proposed by Malawer is commonly utilized for the resection of shoulder girdle tumors [9]. In this paper, we present a modified version of this approach, that employs osteotomies of the clavicle, acromion, and coracoid process, ensuring wide exposure of the tumor and safe dissection especially from the posterior side (Fig. 3). The rationale behind our technique is grounded in the differential healing capabilities of various tissues. Specifically, bones after osteotomies are capable of swift and complete tissue regeneration [13]. In contrast, ligaments heal with often inferior fibrous tissue and over a prolonged time [13]. Thus, the implementation of osteotomies followed by rigid fixation promotes a faster recovery. Furthermore, the perioperative application of the LCPs prior to performing the osteotomy, along with preparation of screw holes, assures excellent anatomic reduction throughout the subsequent reconstructive phase.

This case report is limited by the absence of long-term outcomes, as the follow-up period extends only beyond one year. While this duration is adequate for assessing the cosmetic benefits of the innovated implant, it falls short of providing comprehensive insights into its long-term properties. The design improvements were based on prior validated mechanical principles and earlier clinical experiences. These refinements, such as volume reduction and locking mechanism optimization, aimed to address previously observed issues such as anterior implant protrusion. Whereas this study presents only a single case, it is intended as a proof-of-concept to demonstrate the feasibility, surgical technique, and early cosmetic benefits of an optimized double-constrained implant following extra-articular shoulder resection. The primary objective of this case report is to highlight technical modifications and promising early cosmetic outcomes observed in this complex oncologic setting, rather than to draw generalizable conclusions regarding functional efficacy or long-term implant performance. Importantly, since the initial report of this implant design (7), our experience now includes a cohort of four patients who underwent shoulder reconstruction with the double-constrained implant. Of these, three—including the present case—remain stable at latest follow-up (ranging from 18 to 45 months) with no evidence of glenoid loosening or implant dislocation. One patient required implant removal due to local tumor recurrence. Although biomechanical testing and long-term data are still lacking, these early clinical results reinforce the potential of this design and support continued study in a larger, comparative cohort. Future studies involving larger sample sizes and extended follow-up periods are crucial to thoroughly evaluate the routine utility of the double-constrained implant for shoulder reconstruction.

Conclusion

Shoulder reconstruction using an innovated double-constrained implant yielded a satisfactory cosmetic outcome while maintaining original properties in terms of durability and range of motion. Furthermore, the implementation of our modified Malawer approach for extra-articular resection enhanced visualization of the tumor, thereby facilitating the achievement of oncologic margins. This combination underscores the potential for improved patient outcomes in shoulder reconstruction.