https://orcid.org/0000-0001-8805-9848
Abstract
We report a case of a naturally-developed reverse shoulder joint following a proximal humeral fracture sustained at the age of 16 years and managed non-operatively. Over time, progressive post-traumatic remodelling resulted in an inversion of the native glenohumeral anatomy, resembling the biomechanical configuration of a reverse shoulder arthroplasty as originally described by Grammont. The humeral head remodelled into a concave socket-like structure, while the glenoid progressively assumed a convex articulation. At 46 years of follow-up, the patient demonstrated an unusual pattern of glenohumeral osteoarthritis but remained pain-free, with a functional shoulder (Constant score 69). To our knowledge, this is the first reported case of a post-traumatic native reverse shoulder joint resulting from an adolescent proximal humeral fracture with long-term follow-up.
Introduction
Proximal humeral fractures comprise ~0.45% to 2% of all fractures and ~14% of upper extremity fractures in the paediatric population. An understanding of proximal humeral anatomy is critical to comprehending fracture displacement and its implications for fracture remodelling [1, 2].
The vascularity of the proximal humerus is derived from the terminal branches of the axillary artery. Damage to these blood vessels during severe trauma or aggressive surgery can cause avascular necrosis [3]. Potential complications of non-operative treatment include malunion, stiffness, and physeal growth arrest causing limb-length discrepancies [4, 5].
Treatment selection is based on patient age, fracture displacement, and remodelling capacity. Non-operative management is successful in younger patients or less displaced fractures. Operative management is usually considered in older paediatric patients with more severe fracture displacement [5].
In this report, we present an exceedingly rare case of a long-term sequelae following a non-operatively treated proximal humeral fracture in a 16-year-old patient. A natural reverse shoulder joint developed, mimicking the biomechanical configuration of a reverse shoulder prosthesis described by Grammont.
Case report
A 16-year-old male sustained a high-energy proximal right humeral fracture during a motor vehicle rollover accident. Initial conservative management consisted of simple sling immobilization. Although surgical correction of the residual deformity was subsequently contemplated, non-operative care was maintained. After treatment completion, the patient was lost to follow-up. He subsequently worked as a woodcutter and reported no functional limitations during his professional duties. At age 53, an occupational accident involving a chainsaw resulted in a severe degloving injury to the contralateral wrist and hand, leading to permanent work disability. He currently works as a light vehicle driver.
In November 2020, ~40 years after the proximal humeral fracture, the patient was referred to our orthopaedic outpatient department because of right shoulder pain and stiffness. The patient reported seasonal mechanical pain during colder months, which he managed without regular analgesics, as well as restricted internal and external rotation. Imaging evaluation revealed a post-traumatic reverse shoulder deformity.
In 2026, right shoulder radiographs and computed tomography (CT) imaging demonstrated marked proximal humeral dysmorphia and a reverse shoulder deformity with joint space obliteration. The CT scan also highlighted supraspinatus muscle belly atrophy, with fatty infiltration of the infraspinatus and subscapularis muscles (Figs 1 and 2).
Anteroposterior right shoulder radiograph showing a reverse shoulder deformity.
CT of the right shoulder with 3D reconstruction (A), alongside sagittal (B), and coronal (C) views, showing a reverse shoulder joint deformity associated with a specific osteoarthritis phenotype.
On physical examination, atrophy of the right supraspinatus and infraspinatus muscles was noted. Range of motion testing registered active abduction of 130°, internal rotation of 60°, external rotation of 30°, flexion of 130°, and extension of 25° (Fig. 3). The Constant score was 69/100, lower than the expected age- and sex-matched normative value (83 ± 4.2 points). He reported occasional mild pain and high satisfaction with his functional status.
Clinical photographs showing painless range of motion of the right shoulder.
Discussion
Proximal humeral fractures and epiphyseal separations in skeletally immature children and adolescents are traditionally managed conservatively using a simple sling and swathe system. Conversely, older children (>13 years) with severely displaced fractures may benefit from anatomical reduction and surgical stabilization [2, 4, 6].
The present clinical case was potentially driven by avascular necrosis of the humeral head, which subsequently triggered the development of a naturally occurring reverse shoulder deformity, in accordance with Wolff’s law. Boileau et al. analysed proximal humeral fracture sequelae and classified them into four distinct types, distinguishing between intra-articular (Types 1 and 2) and extra-articular (Types 3 and 4) lesions. Type 1 sequelae are characterized by avascular necrosis, collapse of the humeral head, and resorption of the greater tuberosity [7, 8]. Characterizing the aetiopathogenesis of our case, the humeral head collapsed and spontaneously healed with reverse polarity. Essentially, the resorbed and collapsed native humeral head morphed into a concave, socket-like structure, while the glenoid progressively rounded into a convex ‘ball’.
A few limitations should be acknowledged. Unfortunately, it was not possible to retrospectively trace the complete clinical and radiological evolution of the fracture. Although initial conservative management was pursued, surgical intervention was subsequently recommended, likely due to pain and fracture displacement. Nevertheless, the patient declined surgery, and no intermediate follow-up records are available. Despite this, he successfully worked as a woodcutter without functional limitations until his contralateral hand injury.
In 2026, at the 46-year follow-up, the patient (now 66 years old) reported only occasional mild pain, demonstrated a functional shoulder (Constant score 69), and expressed high satisfaction. Consequently, a conservative, watchful-waiting strategy was maintained. This deformity could potentially have been prevented had the initial fracture been managed with open reduction and internal fixation. Conversely, a reverse shoulder arthroplasty may be indicated in the future should the patient develop painful, end-stage osteoarthritis.
Only three cases of a naturally occurring reverse shoulder joint have been documented in the literature [9–11]. Two of these patients had congenital deformities, specifically isolated congenital proximal humeral hypoplasia and thalidomide-induced dysmelia. The third case involved a reverse shoulder joint secondary to a humeral neck fracture in a 70-year-old woman.
To our knowledge, our case represents the first clinical report of an acquired reverse shoulder joint secondary to a proximal humeral fracture sustained during adolescence, wherein the humeral head remodelled into a socket-like configuration and the glenoid transformed into a convex articulating surface.
When a proximal humeral fracture occurs during adolescence, the humeral head may undergo necrosis and resorption. In an attempt to restore a functional new articulation, the remaining bone can undergo extensive remodelling. This biological cascade may invert the native anatomy of the shoulder through a purely non-surgical process, thereby mimicking a reverse shoulder arthroplasty. This spontaneous biological reverse arthroplasty may provide a long-term functional joint with minimal pain.
Conflicts of interest
None declared.
Funding
None declared.
