Results of a differential approach to surgical treatment of proximal humerus fractures in patients with osteoporosis

Objective. To conduct a comparative retrospective analy - sis of the anatomical and functional results of surgical treatment of proximal humerus fractures AO/OTA 11-B, 11-C using different methods in patients over 50 years old with osteoporosis to justify a differential approach. Methods. The study included 102 patients aged 50 years and older with osteoporosis and proximal humerus fractures AO/OTA 11-B, 11-C. The patients were divided into three groups: I — 50 (16 men, 34 women) whom were performed open reduction and internal fixation (ORIF) with the PHILOS plate; II — 44 (8 men, 36 women) — ORIF with the PHILOS plate and using 3D polylactide (PLA) porous implants were applied; III — 8 (2 men, 6 women) whom were performed primary reversed total shoulder arthroplasty (RTSA) with developed total reversible endoprosthe - sis. The results of treatment were evaluated according to the Con - stant-Murley Shoulder Score system after 3, 6, 12 months. The re - sults. Positive results


Introduction
Proximal humerus fractures (PHF) make up 5-6 % of injuries of the musculoskeletal system [1][2][3]. It is known that 80 % of AO/OTA A2/A3 type PHF are stable, and conservative treatment in such cases results in good functional outcomes [4][5]. But in unstable three-and four-fragment PHF according to Neer or type AO/OTA 11-B, 11-C this can be achieved only with the help of surgical intervention [4-6]. For patients with osteoporosis, there are no standards for the selection of surgical treatment methods in types B and C PHF, and the issue remains debatable. The main types of surgical intervention for them are open reduction and internal fixation (ORIF) using various plates with angular stability and the mandatory use of various auto-or allografts, bone cement [7-13], including structural porous implants made of polylactide (PLA) [14-16]. Hemiarthroplasty of the shoulder joint for three-and fourfragment PHF is now used much less often compared to reverse total shoulder arthroplasty (RTSA). This is due to a significant number of unsatisfactory results and complications [17][18][19]. In recent years, RTSA of the shoulder joint, proposed by Grammont [20-21], is the main method of surgical treatment in patients with types AO/OTA 11-B, 11-C PHV secondary to osteoporosis [22][23][24][25]. However, there are few studies comparing the clinical results of RTSA and ORIF in patients with type B and C proximal humeral fractures (AO/OTA).
The purpose of the study: to conduct a comparative retrospective assessment of the anatomical and functional results of surgical treatment of types AO/ OTA 11-B, 11-C fractures of the proximal humerus using different methods in patients over 50 with osteoporosis to substantiate a differential approach. -III -8 individuals with osteoporosis (2 men, 6 women, from 50 to 76 years old) in the period from 2020 to 2022 underwent primary RTSA with the elaborated total reversible endoprosthesis [30]. In all cases, the indication for RTSA was fractures of type AO/OTA 11-B (2 subjects) or AO/OTA 11-C (6 subjects) secondary to severe osteoporosis.

Material and methods
In the pre-and postoperative period, all patients underwent a standard clinical examination and radiography. Spiral computed tomography was performed in some cases in multifragmentary fracture-dislocations to determine the defect of the articular surface. The severity of osteoporosis was assessed by X-rays of the humerus in the anterior-posterior projection with the calculation of the cortical index (CI) [31,32].
Radiographic standard examination of the proximal part of the humerus in 2 projections was performed in 1, 3, 6, 12 months. The results of treatment were evaluated according to the Constant-Murley Shoulder Score system within 3, 6, 12 months after surgery [33].
Operations were performed in the «beach chair» position under general anesthesia and nerve block. Intraoperative X-ray control was provided using C-arm.
Deltopectoral access was performed in 85 patients, anterolateral one in 23.
In Group II patients, two porous 3D PLA implants with a diameter of 8 or 10 mm and a length of 20 to 45 mm were mostly used. In the case of severe osteoporosis (CI ≤ 0.29) [31, 32] and a large diameter of the humeral canal (more than 15 mm), a third PLA 3D implant with a diameter of 12-15 mm was previously installed in the canal as a support for the other two. In some cases, in a significant post-traumatic defect of the cancellous tissue of the humeral head, the space between the 3D PLA implants was additionally filled with allobone, a cortical-spongy granulate OMS-A-KGG produced at the State Institution Professor M. I. Sytenko Institute of Spine and Joint Pathology of the National Academy of Medical Sciences of Ukraine.
In primary or revision RTSA, a standard deltopectoral access was used. Immobilization in a Dezo bandage lasted 4-6 weeks with the possibility of oscillating and pendulum movements in the shoulder and elbow joints with a small amplitude of 10-20°. Vacuum drainage was left for 24-48 hours. Active movements in the elbow joint were allowed on the 2 nd to 3 rd day after surgery, passive movements in the shoulder joint on the 3rd to 5th day. Active movements in the shoulder joint were usually allowed 4 weeks after the operation and in significant post-traumatic defects of the humeral head and severe osteoporosis (CI ≤ 0.29) in 6 weeks.
Group III patients underwent surgical intervention using the elaborated reversible total modular endoprosthesis of the shoulder joint [30]. The following tasks were set while improving the design of the endoprosthesis: to upgrade the stability of fixation in the bone; to enhance the durability of the device; to increase the life of the friction pair to prevent the pos-sible development of aseptic instability due to wear of the polyethylene liner.
The use of additive technologies makes it possible to print parts of endoprostheses of complex geometric shapes with the required porosity. In the proprietary endoprosthesis, the glenoid support plate, printed on a 3D printer from titanium powder (similar to VT-5), has a central rod covered with conical ribs and pores with a pore rib length of 350-500 μm. The surface of the glenoid component is also completely porous (pore edge length 350-500 μm) to a depth of 2 mm. The outer surface of the glenoid support plate is made in the form of a Morse cone for the fitting of the glenoid head, which also has a Morse cone surface on the inner side (Figs. 1, 2).
The proposed endoprosthesis contains a glenoid head (hemisphere) made of a titanium alloy covered with a diamond-like coating (Fig. 2) or made of polyether ether ketone (PEEK). It is fixed to the glenoid support plate with a screw through the central hole.
Coating of the external friction surfaces of the glenoid head with a diamond-like carbon coating from a filtered vacuum-arc cathode plasma with the application of high-voltage pulse potentials of negative shear to the deposited surfaces ensure high adhesion of these coatings to the surfaces, a low level of internal stresses against the background of ensuring a high hardness of not less than 40 GPa and a low Fig. 1. Reversible modular endoprosthesis of the shoulder joint: 1 -a leg made in the form of a conical three-lobed rod; 2 -porous proximal part, printed on a 3D printer from titanium powder (similar to VT-5); 3 -an PEEK insert in the form of a conical cylinder; 4 -concave spherical surface of the liner; 5 -Morse cone; 6 -the proximal part of the endoprosthesis is fixed using a Morse cone to the proximal part of the leg; 7 -locking nut; 8 -glenoid head; 9 -glenoid support plate; 9 -the outer edge of the support plate, made in the form of a Morse cone; 10 -fixing screw; 11 -central rod; 12 -cone-shaped ribs with a pore rib length of 350-500 μm; 13 -central screw; 14 -the porous surface of the glenoid support plate with a pore edge length of 350-500 μm to a depth of 2 mm; 15 -the outer surface of the glenoid support plate in the form of a Morse cone; 16 -spherical holes of the glenoid support; 17spherical washers; 18 -screws; 19 -proximal part with ribbed porous outer surface [30]; a) appearance; b) frontal projection; c) side view а b c coefficient friction ~ 0.1 [34,35]. When this coating slides over the titanium nitride coating, the coefficient of friction between them decreases several times and becomes less than 0.1. The thickness of protective coatings, made of diamond-like carbon coating is not less than 1.5 μm, provides not only reliable protection of the metal surface from electrochemical corrosion, but also a sufficiently high service life due to a significant reduction in the coefficient of friction of these surfaces during their mutual sliding.

Results and their discussion
Results of treatment of Group I patients according to the Constant-Murley Shoulder Score system 3, 6, 12 months after the operation averaged 71.3; 76.6; 78.4 points respectively. In 3 patients, varus migration of the head occurred without significant impairment of the function of the shoulder joint, in 2 patients complete destabilization of the metal structure occurred, resulting in its removal 3 months after surgery. In 5 operated patients, a secondary displacement of the greater tubercle occurred up to 10-15 mm one month after surgery. Avascular necrosis of the head of the humerus was determined in 4 patients within 6 to 12 months, resulting in the removal of metal structures and implementation of RTSA [28]. The conducted evaluation showed that the use of PHILOS plates made it possible to obtain positive results (excellent and good) in 36 Group I patients (72.0 %) within 6 to 12 months after surgery, 8 (16.0 %) were satisfactory, 6 (12.0 %) were unsatisfactory.
In Group II patients the average indicator according to the Constant-Murley Shoulder Score system within 3 months after the operation was an average of 78.0 points, 81.8 points within 6 months and 88.0 points within 12 months. Fragment consolidation was achieved in all patients. No signs of avascular necrosis of the humeral head were observed. One woman experienced perforation of a fragment of the head with a screw 6 weeks after surgery, which was removed under C-arm control through a skin puncture under local anesthesia [28]. 2 patients deve-loped a superficial infection, which was eliminated by local and systemic antibacterial therapy. One patient in 3 months was found to have a secondary displacement of the lesser tubercle up to 8-10 mm, 8 subjects had a varus displacement up to 5 mm without functional impairment. In all patients of this group, a plate of the PHILOS type, 2-3 PLA implants were used, and in 14, cortical-spongy granulate OMS-A-KGG was additionally employed. The conducted analysis showed that the use of PHILOS plates using additional reinforcement with 3D PLA implants allowed to obtain positive results (excellent and good) in 36 (81.8 %) patients, satisfactory results in 8 (18.2 %) and to reduce the number of complications (88.0 points by the Constant-Murley Shoulder Score) within 6 to 12 months after surgery.  (Fig. 3). On 01.03.2022 an open reposition under general anesthesia was performed followed by nerve block using osseous metal-osteosynthesis with the PHILOS plate and two porous 3D PLA implants. She was immobilized with a Dezo type bandage for 4 weeks. Active movements in the left elbow joint were allowed on the 2 nd day after surgery, passive movements in the shoulder joint under the control of a physical therapy instructor in the first week after surgery, and active movements in 4 weeks. The function of the right shoulder joint was restored according to the Constant-Murley system 78 points within 3 months, 85 points within 6 months (Fig. 4).
In Group III patients the average score according to the Constant-Murley Shoulder Score system within 3 months after the operation was 76 points, 80 points within 6 months, 82 points within 12 months. Restoration of the shoulder joint function was achieved in all of them in the absence of signs of instability of the endoprosthesis components. One patient developed neuropathy of the radial nerve, the function of which was restored in 3 months following neurotropic therapy. In the period from 6 to 12 months after surgery, 6 patients (75 %) had good results, 2 (25 %) had satisfactory results. Patient O. was immobilized with a Dezo bandage for 4 weeks. She began active movements in the right elbow joint on the 3rd day after the operation, pas-sive movements in the shoulder joint under the control of a physical therapy instructor on the 4 th week, and active movements in 6 weeks. The function of the right shoulder joint was restored according to the Constant-Murley system: 75 points within 3 months, 78 points within 6 months, 80 points within 12 months (Fig. 7).
Discussion PHF makes up almost 30 % of all fractures in the elderly. Due to the aging of the population, the incidence of these fractures will continue to increase [36]. Although the majority of such injuries are treated conservatively, some patients with type AO/ OTA 11-B, 11-C fractures are recommended surgical treatment. However, debate continues regarding its indications and ideal strategy. The Section of Skeletal Trauma and Sports Injuries of the European Society of Traumatology and Emergency Surgery (ESTES) appointed a working group to reach a consensus among European countries on PHF management. The recommendations compare conservative and four possible options for surgical treatment -ORIF, intramedullary blocked osteosynthesis, hemi-and total reversible endoprosthesis [37]. People over 50 with osteoporosis are recommended to undergo either hemiarthroplasty or RSTA due to the impossibility of osteosynthesis. Hemiarthroplasty has been shown to result in worse functional outcomes compared to RTSA. Secondary RTSA after other surgical treatment also results in outcomes comparable to primary RTSA [37]. A retrospective assessment of the treatment outcomes of patients who underwent shoulder arthroplasty due to PHF from 2010 to 2019 showed no significant differences in the function of the shoulder joint after RTSA and hemiarthroplasty. The authors concluded that this could be explained by the restrictive indications for hemiarthroplasty [38]. In another retrospective study [39] of the results of treatment of patients with PHF from 2010 to 2017, the authors differentially analyzed osteosynthesis or endoprosthetic repair. Osteosynthesis of type AO/OTA 11-B, 11-C fractures was carried out when technically possible. Either a deltopectoral access or minimally invasive plate osteosynthesis (MIPO) was used. Hemiarthroplasty was used if it was possible to reconstruct the tubercles and there were no signs of osteoarthritis. In all other cases, RTSA was performed. The authors believe that both osteosynthesis and primary arthroplasty can lead to a good or even excellent functional result, noting a high frequency of revisions after osteosynthesis. Revision procedures, which involve endoprosthetic repair after failed osteosynthesis, provide much lower functional results [39]. Thus, it has been proven that surgical treatment of PHF in the elderly and senile secondary to osteoporosis creates problems for decision-making regarding the choice of a surgical intervention method. The use of RTSA has led to a paradigm shift in the treatment of PHF in the elderly. The unique biomechanical principles and design features of RTSA make it a suitable treatment option for PHF in the elderly with osteoporosis and tubercle defects. The method is considered a reliable approach after failed hemiarthroplasty or failed ORIF [40]. RTSA has clear advantages over hemiarthroplasty and internal fixation, providing pain relief and reliable recovery or significant improvement in functional outcomes. As a result, an increase in the number of RTSAs in elderly patients has been observed over the last decade [41].
Our evaluation of surgical treatment of patients with PHF over 50 showed that the use of 3D PLA implants as a reinforcing material during ORIF results in positive outcomes with a minimum number of complications in the majority of patients. In an unfavorable result of ORIF, repeated intervention involves revision, removal of metal structures, arthrolysis and reversible total endoprosthetic repair of the shoulder joint. Our experience proves that the initial use of RTSA in chronic type AO/ OTA 11-B, 11-C fracture-dislocations and the technical impossibility of conducting ORIF results in positive outcomes up to 3 years after the operation.
Thus, on the basis of a retrospective assessment of our own results and special literature, we offer a differentiated approach to the selection of surgical treatment for patients aged 50 years and older with type AO/OTA 11-B, 11-C fractures: