Abstract
Aims
The survival of humeral hemiarthroplasties in patients with relatively intact glenoid cartilage could theoretically be extended by minimizing the associated postoperative glenoid erosion. Ceramic has gained attention as an alternative to metal as a material for hemiarthroplasties because of its superior tribological properties. The aim of this study was to assess the in vitro wear performance of ceramic and metal humeral hemiarthroplasties on natural glenoids.
Methods
Intact right cadaveric shoulders from donors aged between 50 and 65 years were assigned to a ceramic group (n = 8, four male cadavers) and a metal group (n = 9, four male cadavers). A dedicated shoulder wear simulator was used to simulate daily activity by replicating the relevant joint motion and loading profiles. During testing, the joint was kept lubricated with diluted calf serum at room temperature. Each test of wear was performed for 500,000 cycles at 1.2 Hz. At intervals of 125,000 cycles, micro-CT scans of each glenoid were taken to characterize and quantify glenoid wear by calculating the change in the thickness of its articular cartilage.
Results
At the completion of the wear test, the total thickness of the cartilage had significantly decreased in both the ceramic and metal groups, by 27% (p = 0.019) and 29% (p = 0.008), respectively. However, the differences between the two were not significant (p = 0.606) and the patterns of wear in the specimens were unpredictable. No significant correlation was found between cartilage wear and various factors, including age, sex, the size of the humeral head, joint mismatch, the thickness of the native cartilage, and the surface roughness (all p > 0.05).
Conclusion
Although ceramic has better tribological properties than metal, we did not find evidence that its use in hemiarthroplasty of the shoulder in patients with healthy cartilage is a better alternative than conventional metal humeral heads.
Cite this article: Bone Joint J 2024;106-B(11):1273–1283.
Take home message
This in vitro study did not reproduce the superior glenoid wear performance observed with ceramic hemiarthroplasties compared to metal hemiarthroplasties, as documented in both clinical and previous in vitro studies.
No significant correlation was found between glenoid wear and various factors, including age, sex, the size of the humeral head, joint mismatch, the thickness of the native cartilage, and the surface roughness.
Introduction
The demand for arthroplasty of the shoulder in young patients is projected to increase by 93% in the next two decades.1 However, the management of osteoarthritis (OA) or fracture of the shoulder in active patients aged ≤ 65 years is difficult and remains controversial. The longer lifespan of these patients after surgery increases the chance of implant failure, particularly as their higher physical demands place greater stress on the joint, increasing the need for early revision procedures. It is thus not surprising that arthroplasties of the shoulder in young patients have been revised at a rate that is two to four times higher than in older patients.2-5
Earlier studies confirmed that the use of a hemiarthroplasty improved the function of the shoulder and reduced pain in young patients.6-9 Hemiarthroplasty also offers the advantages of bone-sparing, less invasive surgery, facilitating later revision. Furthermore, it eliminates the risk of loosening of the glenoid component, a major concern after anatomical total shoulder arthroplasty (TSA),10 as well as scapular notching, which is a major concern after reverse TSA (rTSA).11 Failure of hemiarthroplasty, however, has been reported to result in a high revision rate of 22% in young patients at 15 years postoperatively, of which 71% were for painful glenoid erosion.6 These results are consistent with reports from the National Joint Registry 20th Annual Report in 20232 and other national registries.12-14 This high revision rate is the reason for the declining use of hemiarthroplasty in favour of TSA, particularly rTSA. The current study took the approach that, rather than abandoning the use of hemiarthroplasty, despite the obvious advantages, it may be possible to reduce glenoid erosion and thus the rate of revision.
In addition to causing a high rate of primary revision, glenoid erosion adversely affects the outcome of revision hemiarthroplasty to TSA because the bone loss causes an uneven surface with poor quality of bone, which leads to poorer fixation of the TSA.15-18 This issue negates the great advantage of ease of revision of hemiarthroplasty. Thus, it would be attractive to reduce glenoid erosion after hemiarthroplasty.
The choice of bearing materials for hemiarthroplasty has been widely investigated. Ceramic has become an alternative to metal because it is hard, inert, resistant to abrasion, and is not subject to gradual changes in the roughness of its surface as with metal.19 Ceramic is also more wettable than metal, providing better lubrication which results in less adhesive wear of the glenoid.19 This is supported by the widespread use of ceramic heads in total joint replacements,20,21 with reduced rates of in vivo wear compared with metal heads.20-23 Davies et al24 recently reported that revision of hemiarthroplasty for glenoid erosion was required in ten of the 1,020 patients (0.98%) with metal heads, compared with one of the 343 patients (0.29%) with a ceramic head. Furthermore, superior tribological properties, such as friction and surface roughness, in ceramics compared with metal have been shown in several in vitro studies dealing with hemiarthroplasty.25-28 However, the postoperative association between the materials used in the implants and glenoid erosion was not assessed in any of these studies.
The aim of this study was to compare the in vitro glenoid erosion in specimens from young patients, aged between 50 and 65 years, induced by ceramic or metal humeral heads using a shoulder wear simulator. The hypothesis was that ceramic implants would produce less erosion than metal implants.
Methods
Specimens and specimen preparation
Based on discussions with the surgical team, it was decided that a difference between implant materials of 20% of the thickness of the articular cartilage after five years of use would be a meaningful outcome. A power analysis, and cartilage wear data reported by Patel and Spector29 in 1997, estimated that at least eight samples in each group would be needed to detect this change with 95% power. Thus, 22 fresh-frozen human cadaveric shoulders were acquired from MedCure (USA), following the granting of ethical approval by the Imperial College Healthcare Tissue Bank (ICHTB). These shoulders met the inclusion criteria of being intact right shoulders, in patients aged between 50 and 65 years with no evidence of shoulder pathology or previous surgery. Two shoulders with gross OA were excluded and one shoulder was excluded due to a faulty testing setup that resulted in a severely damaged glenoid. The ceramic group included shoulders from four male and four female cadavers with a mean age of 58.3 years (51 to 63). The metal group included shoulders from four male and five female cadavers, with a mean age of 61.7 years (56 to 65). Two shoulders from male cadavers were used for soak control to account for fluid absorption and creep.
Each specimen was stripped of all soft-tissue, with preservation of the glenoid labrum and cartilage. The glenoid was isolated from the scapular body and then potted in polymethyl methacrylate (PMMA) and encased in a custom-made glenoid pot (Figure 1a). The superior-inferior (SI) and anteroposterior (AP) axes of the glenoid were aligned to the orientation of the simulator, and determined as described by Lee and Lee.30 Specifically, the SI axis is the line between the intersection of the superior rim of the glenoid with the glenoid-coracoid confluence and the intersection of the inferior rim with the inferolateral margin of the scapula. The AP axis is the line representing the widest distance between the anterior and posterior rims, and is perpendicular to the SI axis. Although there may be variation in the orientation of the glenoid, the effect of such variations on the results may be mitigated by the measurement of glenoid wear across four relatively large areas of its face, as is described later.
Fig. 1
a) The glenoid was cemented in a pot. Its superior-inferior (SI) and anteroposterior (AP) axes were aligned to the orientation of the simulator. b) The potted humeral head. c) The six-station shoulder wear simulator, which can apply three glenohumeral (GH) rotations (flexion-extension, abduction-adduction, internal-external rotation) and three GH translations/forces (axial compressive, SI shear, AP shear) to each shoulder. d) Neutral position of GH joint determined by applying compressive load only. e) The GH joint was fully immersed in lubricant within sealed chambers during wear testing.
The glenoid was positioned with a tilt of 5° in both the superior and posterior directions, as determined using a bubble level. This adjustment took into account the mean orientation of the plane of the glenoid relative to the scapula.31 This positioning was consistent with the physiological stability of the shoulder during wear testing. Despite the large variation in version and orientation of the glenoid in normal shoulders,31 we specifically investigated the effect of implant materials on glenoid erosion. Thus, other potential contributing factors which could influence wear were deliberately controlled.
The native humeral head was resected and compared with Mathys Trial Heads (Mathys, Switzerland) in order to determine the size which best matched the Mathys ceramic (aluminium oxide, Al2O3) or metal (cobalt-chrome-molybdenum, CoCrMo) heads. The size of both heads ranges from 39 to 53 mm in diameter, with increments of 2 mm. Both materials have the same geometry of the head for each diameter, eliminating the potential for geometric factors to contribute to glenoid erosion. Each prosthetic head was press-fitted to a cone guide potted in PMMA and encased in a humeral pot (Figure 1b), and carefully located at the centre of rotation of the mounting in the simulator.
Summary of the shoulder wear simulator design
Wear simulation was performed using a six-station shoulder wear simulator (Figure 1c), which can apply three rotations (flexion-extension, abduction-adduction, and internal-external rotation) to the humeral head and three translations (medial-lateral (ML) compressive, SI shear, and AP shear) to the glenoid. The simulator was programmed in LabVIEW to control stepper motors (Arcus Technology, UK) for the rotational movements, while pneumatic cylinders (National Instrument controllers, USA) were used to impose the components of force on the joint. The glenoid was free to translate in the AP, SI, and ML directions, subject to the control of force, the conformity of the joint, and soft-tissue restraints. Once the joints had been mounted in the simulator, their ‘neutral position’ was determined by applying a ML compressive force only and allowing the humeral head to freely settle into its neutral position at the bottom of the concavity of the glenoid (Figure 1d).
Testing conditions
The wear-testing lubricant was prepared according to ASTM F732-17 under a sterile hood to reduce the risk of airborne contamination. Newborn calf serum (16010159; Thermo Fisher Scientific, USA) was diluted with autoclaved deionized water to achieve a protein concentration of 20 g/l,32 mimicking the protein concentration in human synovial fluid. Next, 15 ppm ProClin 300 (48,912 U; Sigma-Aldrich, USA) and 7 g/l ethylenediaminetetraacetic acid (EDTA) (B8R04080; Philip Harris) were added to the calf serum to prevent bacterial growth33 and minimize the development of calcium phosphate films on the surface of the implants. The lubricant mixture was filtered through a 2 µm filter. The wear components were completely immersed in 500 ml of lubricant within sealed chambers during testing (Figure 1e). The lubricant remained at room temperature to avoid overheating the tissues, prevent excessive precipitation, reduce microbial growth, and produce wear characteristics comparable to those in vivo.34,35 Wear was tested for 500,000 cycles at 1.2 Hz (0.8 cycle/s). At every 125,000 cycles, the specimens were removed for wear examination, and the lubricant was renewed.
Determining kinematics and forces on the GH joint
The kinematic and loading profiles for ‘washing the opposite axilla’ were obtained (Figure 2) using a validated musculoskeletal model of the upper limb.36 The ranges of motion per cycle were 25° flexion, 15° abduction, and 15° internal rotation. The maximum loads applied to the glenoid were 340 N compression, 192 N superior shear, and 45 N posterior shear. For the soak control tests, the two specimens were subjected to a 340 N compressive load without shear loads or joint rotations, and were lubricated for the same length of time as for wear testing.
Fig. 2
Kinematics and forces for ‘washing the opposite axilla’ applied to each shoulder during each cycle of wear testing: 0% is at the initial position, 50% is at the washing position, and 100% is back at the initial position.
Cartilage wear analysis
The thickness of the articular cartilage of the glenoid was measured before and after wear testing when the specimens were removed from the simulator, washed and kept in phosphate-buffered saline (PBS) solution for one hour, and then dried with a paper towel. This allowed recovery of the cartilage thickness after being subjected to static compressive loading. Control tests on two specimens showed that this length of time was sufficient to eliminate any substantial changes in cartilage thickness due to creep. Changes in thickness caused by wear were measured from micro-CT scans using a Versa 510 Micro CT scanner (ZEISS, Germany) at 140 kV and 70 μA without a filter. The voxel size was set to 54.6 µm. The images were reconstructed using the Reconstructor Scout-and-Scan software (ZEISS).
The reconstructed 3D scans were imported into Fiji (ImageJ 1.54f; National Institutes of Health, USA) software (Figure 3a). The whole specimen was initially segmented by converting the scans into binary images using the default threshold (Figure 3b). The original scans were then used to segment the bone using the built-in 'Maximum entropy' threshold algorithm (Figure 3c). Noises and outliers were removed. Boolean subtraction was used between the images of the whole specimen and the images of the bone to derive the articular cartilage layer (Figure 3d). The subtracted images were subsequently used to compute the thickness of the cartilage (Figure 4) using Matlab R2019b (MathWorks, USA), adapting the method described by Das Neves Borges et al.37 The SI and AP axes of the glenoid were determined as previously described and the surface was divided into four regions: superior-anterior, superior-posterior, inferior-anterior, and inferior-posterior. As the labrum was preserved during wear testing to maintain the conformity of the joint, the surface of the cartilage was isolated using the ‘drawfreehand’ tool in Matlab R2019b, referencing its outline on the glenoid. The mean thicknesses of the cartilage for each region and across the whole surface (referred to as ‘total’) were calculated. The depth of wear was defined as the difference between the initial thickness of the cartilage (cycle 0) and its thickness in the current test cycle. Soak control tests were used to determine the change in thickness due to creep and swelling.
Fig. 3
a) A micro-CT slice of the glenoid. b) A binary image of the whole specimen. c) A binary image of the bone. d) The articular cartilage layer derived from Boolean subtraction between the whole specimen and the bone.
Fig. 4
Distribution of the thickness and wear depth of the glenoid cartilage.
Joint mismatch
3D models of the glenoid were generated from the reconstructed CT volumes in Materialise Mimics 24.0 software (Materialise, Belgium) to measure the radius of curvature of the cartilaginous surface of the glenoid through sphere fitting. The mismatch was the difference between the radius of curvature of the cartilaginous glenoid and that of the prosthetic humeral head.
Surface roughness of humeral head
At the end of the wear test, all the humeral heads were subjected to a 15-minute ultrasonic acetone bath, then wiped with ethanol and dried with pressurized air. Damage to the articulating humeral head was visually and qualitatively analyzed using a measuring laser microscope (Olympus OLS5000 LEXT; Olympus, Japan). Two untested ceramic heads and two untested metal heads served as controls. The mean roughness (Ra) was measured according to the BS ISO 7206-2−2011+ A1 (2016).38 Five measurement profiles were taken, one at the spherical pole and four at each of the four quadrants, approximately 30° from the spherical pole (Figure 5a), to maximize the cover and to capture the part of the surface which was in contact with the glenoid during testing. In each profile, six series of measurements were performed, with three being taken perpendicular to the others (Figure 5b). Based on the testing standard, the evaluation length for the ceramic head was 0.43 mm with a 0.08 μm cut-off filter (λc), and the evaluation length for the metal head was 1.28 mm with a 0.25 μm cut-off filter (λc). These differences could skew the comparison between the materials. However, the focus was on the change in surface roughness of each material rather than comparing their surface roughness.
Fig. 5
a) Surface profile measurements were taken at the spherical pole and in each of the four quadrants approximately 30° from it. b) Six series of measurements were performed at each of the five locations, with three measurements taken perpendicular to the other three.
Statistical analysis
Welch’s t-test was used to compare the measurements of wear between the two materials. Differences in wear between any of the regions of the glenoid were assessed using the Kruskal–Wallis one-way analysis of variance (ANOVA). Change in surface roughness for each material was assessed using an independent-samples t-test. Wear measurements were also compared between the sexes. Pearson correlation was used to determine statistical correlations between wear measurements and factors including age, humeral head size, joint mismatch, initial cartilage thickness, and the surface roughness of the humeral head. Mean data are presented with the SD. The level of significance was set to 0.05. Analyses were performed using SPSS v. 28 (IBM, USA).
The interobserver reproducibility of the wear measurements was assessed by having a second independent observer (AT) perform the thresholding-based cartilage segmentation and quantified using the intraclass correlation coefficient (ICC), with a 95% CI and an absolute agreement. The ICCs were interpreted following Landis and Koch’s39 classification of the level of agreement. Statistical analysis for the ICC was conducted using SPSS v. 28.
Results
Cartilage wear
The difference in wear between the ceramic and metal groups was not significant (p = 0.606, Welch's t-test) at the completion of the wear tests (Figure 6a). Wear was evident in both groups through the changes in thickness for each glenoid specimen (Figures 6c and 6d). At 500,000 cycles, the mean cartilage thickness had decreased significantly in both the ceramic and metal groups, by 27% (0.53 mm (SD 0.26); p = 0.019) and 29% (0.58 mm (SD 0.16); p = 0.008, both Welch's t-test), respectively. The pattern of wear differed between the groups: ceramic heads caused uniform wear across the surface of the glenoid, whereas metal heads caused the most wear in the superior region, which was 117% higher than that in the inferior region (Figure 6b). The location on the glenoid had no significant effect on the measurement of wear (p = 0.281, Kruskal–Wallis one-way ANOVA). The rate of wear was not significantly different between shoulders from the male and female cadavers (p = 0.196, Welch's t-test). There was no significant correlation between wear and age, humeral head size, joint mismatch, or the initial cartilage thickness (p = 0.138, p = 0.432, p = 0.428, and p = 0.357, respectively; Pearson correlation). The change in thickness due to creep (mean -0.06 mm (SD 0.04)) was considered to constitute only a minor part of the measurements (mean 0.55 mm (SD 0.21)). The interobserver reliability for the measurements of cartilage thickness and, subsequently, wear through threshold-based segmentation was excellent (ICC 0.98).
Fig. 6
a) Total cartilage wear (mm). b) Wear (mm) at each region at 500,000 cycles. c) An example of a glenoid specimen with a changing distribution of cartilage thickness tested against a ceramic head, demonstrating d) progressive depth of wear. IA, inferior-anterior; IP, inferior-posterior; SA, superior-anterior; SP, superior-posterior. Data are shown as mean and SD.
Roughness measurements
The mean Ra of the tested heads of both materials were significantly higher than in the control group (Figure 7a). The mean increase in roughness was 26% in the tested ceramic heads (p = 0.025) and 60% in the tested metal heads (p = 0.046, both independent-samples t-test). Metal heads showed signs of abrasive wear, indicated by scratches and grooves at 10× magnification (Figure 7c), which were not seen on the ceramic heads (Figure 7b) even at higher magnification of 30×. However, there was no significant correlation (p = 0.994, Pearson correlation) between the surface roughness of the humeral head implants and the cartilage wear.
Fig. 7
a) Mean roughness Ra (µm) of untested and tested humeral heads for the ceramic and metal groups (*p ≤ 0.05). Microscopic surface analysis of b) tested ceramic heads at 30× magnification and c) tested metal heads at 10× magnification.
Measurements of natural glenoid
The mean thickness of the native cartilage prior to wear testing was similar for the ceramic and metal groups: 2.2 mm (SD 0.4) and 2.1 mm (SD 0.4), respectively (p = 0.619, independent-samples t-test). The cartilage of all specimens was thickest at the inferior-anterior region (Figure 8). The mean natural radius of curvature for the cartilaginous glenoid was 24.7 mm (SD 2.3) for the ceramic group and 23.8 mm (SD 1.3) for the metal group (p = 0.301, independent-samples t-test). The shoulders of male cadavers had a significantly higher mean radius of curvature than those from female cadavers (25.3 mm (SD 1.9) vs 23.2 mm (SD 1.3); p = 0.018, independent-samples t-test). The radius of curvature of the humeral head was smaller than that of the surface of the glenoid, leading to a mean natural mismatch of 1.6 mm (SD 1.8) for the ceramic group and 0.8 mm (SD 1.0) for the metal group. There was no significant difference in mismatch between the shoulders from male and female cadavers (p = 0.553, Welch's t-test).
Fig. 8
The thickness of the native cartilage (mm) prior to wear testing for both materials. IA, inferior-anterior; IP, inferior-posterior; SA, superior-anterior; SP, superior-posterior. Data are shown as mean and SD.
Discussion
This is the first study in which the wear of the natural concave glenoid cartilage surface against different humeral hemiarthroplasties of different materials was measured using a shoulder wear simulator. Both the ceramic and metal implants induced wear during 500,000 testing cycles, and there was no significant difference between the two materials despite the reasonable pre-wear test power calculation of the number of specimens which were required. Thus, the clinical results which have shown less glenoid wear in hemiarthroplasties,24 and the in vitro studies which have shown better tribological behaviour,25-28 using ceramic rather than metal implants, were not replicated in this in vitro study.
The 60% increase in the surface roughness of metal, as opposed to the 26% increase in ceramic after testing, did not lead to a significant difference in cartilage wear. These results are consistent with those of Jung et al40 and McGann et al,41 who also did not find a correlation between cartilage wear and poorer tribological parameters. Thus, the measurements of surface roughness alone may not represent the entire wear behaviour. Additional factors, including the depth of wear, the properties and composition of the cartilage, and the mechanisms and patterns of wear, may contribute to the wear. There is currently insufficient evidence relating anatomical and concomitant mechanical factors to cartilage wear.42-44
While ceramic has better outcomes than metal in TSAs,22,23,45 no studies could be found in which the effects of implants of different materials on cartilage wear in patients with hemiarthroplasty of the shoulder were compared. Pin-on-disc and ball-on-flat tests have shown ceramic’s superiority over metal in tribological behaviour,25,27-29 but these methods of testing do not simulate physiological conditions. Physiological multidirectional vectors of wear can generate characteristics which align with clinical studies.46,47 Thus, when simulating hemiarthroplasty of the hip, Müller et al,26 in 2004, found a significantly lower coefficient of friction for ceramics compared with metals. However, this finding does not necessarily imply reduced cartilage wear.40,41 The comparative evaluation of biomaterials causing cartilage wear in vivo can only be seen in animal studies. Jung et al40 implanted zirconium oxide ceramic and cobalt-chrome metal into the knees of 27 rabbits, and found more cartilage damage caused by ceramic than by metal. Thus, it remains uncertain whether using ceramic humeral hemiarthroplasties in humans will reduce the rate of glenoid wear.
The location of glenoid erosion in patients who have undergone hemiarthroplasty of the shoulder is predominantly posterior.15 However, we found a varying pattern of overall wear in both the ceramic and metal groups. This is likely to reflect the use of a single wear cycle rather than the range of activities in life. Meyer et al48 reported an associated eccentric posterior glenoid wear, not of dysplastic origin, with a retroverted glenoid, but Herschel et al49 found that glenoid version was not a contributing factor to the degree or orientation of wear.
The glenoid wear in this study could not be quantitatively validated by comparison with other studies because this is the first study in which the cartilage wear was measured on a natural concave glenoid surface. However, the thickness of the native cartilage (mean 2.1 mm (SD 0.4)) was consistent with previous studies (mean 1.7 to 1.9 mm).50-52 Immersing the glenoids in PBS solution for one hour before measuring the wear allowed the cartilage to recover its thickness following compressive loading.53 Moreover, Patel and Spector29 found that swelling and creep contributed only 0.1 mm to the measurements of cartilage wear, of approximately 0.4 mm. We found 0.06 mm creep within a mean wear measurement of ~ 0.5 mm, approximately 12% of the total measurement. Furthermore, the interobserver reliability for determining the thickness of cartilage, and consequently wear measurements, was excellent (ICC ≥ 0.98), indicating the reliability of the method.
The shoulder executes a wide range of complex movements during daily activities and estimating, even approximately, the appropriate activity and the relevant number of cycles to simulate is very difficult, especially given the current lack of standard values in shoulder wear tests. Washing the opposite axilla was chosen as the activity to replicate in the simulator, as it is a commonly undertaken multiaxial activity of daily living and shares a similar magnitude of load with other multiaxial activities such as eating, driving, and perineal care.54 Langohr et al55 measured approximately 800 motions per hour for elevations below 80° in patients, but the frequency of multidirectional motions, as with the activity simulated in this study, was not considered. We simulated more cycles (500,000) than suggested by ASTM F1378-18 (100,000), while applying half of their suggested load (340 vs 750 N). This lower load was judged to be more representative of the activities commonly performed by patients with shoulder prostheses, as measured in vivo by Westerhoff et al.56 Although wear studies of TSAs typically run for at least one million cycles,45,57 the cycle count which we chose was considered adequate for comparing the performance of ceramic and metal implants. This is because noticeable wear for both materials, as well as the largest difference in wear between them, occurred within the first 125,000 cycles. The wear then continued to increase with the passage of time with an approximately linear progression.
A limitation of this study was that it focused on the wear of a glenoid with intact cartilage. Erosion of the glenoid which necessitates revision of a hemiarthroplasty is more severe and includes bone loss in addition to cartilage wear. This could lead to different tribological behaviour and wear outcomes. However, the varying degrees of erosion should not affect the choice of biomaterials.25 Furthermore, this study remains valuable for patients with intact healthy cartilage who require hemiarthroplasty of the shoulder, particularly those with fractures.
In conclusion, although ceramic has better tribological properties than metal, we did not find that a ceramic humeral head in hemiarthroplasty of the shoulder bearing against healthy glenoid articular cartilage caused statistically significantly less cartilage wear than a conventional metal humeral head.
References
1. Rabinowitz J , Kothandaraman V , Lin J , Li X , Friedman RJ , Eichinger JK . Utilization of shoulder arthroplasty in the United States – an analysis of current trends and future predictions . Seminars in Arthroplasty: JSES . 2020 ; 30 ( 3 ): 200 – 209 . Crossref Google Scholar
2. Ben-Shlomo Y , Blom A , Boulton C , et al. The National Joint Registry 20th Annual Report 2023 . London: National Joint Registry . 2023 . https://reports.njrcentre.org.uk/Portals/0/PDFdownloads/NJR%2020th%20Annual%20Report%202023.pdf ( date last accessed 27 August 2024 ). Google Scholar
3. Brewley EE , Christmas KN , Gorman RA , Downes KL , Mighell MA , Frankle MA . Defining the younger patient: age as a predictive factor for outcomes in shoulder arthroplasty . J Shoulder Elbow Surg . 2020 ; 29 ( 7 ): S1 – S8 . Crossref PubMed Google Scholar
4. Dillon MT , Inacio MCS , Burke MF , Navarro RA , Yian EH . Shoulder arthroplasty in patients 59 years of age and younger . J Shoulder Elbow Surg . 2013 ; 22 ( 10 ): 1338 – 1344 . Crossref PubMed Google Scholar
5. Ödquist M , Hallberg K , Rahme H , et al. Lower age increases the risk of revision for stemmed and resurfacing shoulder hemi arthroplasty . Acta Orthop . 2018 ; 89 ( 1 ): 3 – 9 . Crossref PubMed Google Scholar
6. Sperling JW , Cofield RH , Rowland CM . Minimum fifteen-year follow-up of Neer hemiarthroplasty and total shoulder arthroplasty in patients aged fifty years or younger . J Shoulder Elbow Surg . 2004 ; 13 ( 6 ): 604 – 613 . Crossref PubMed Google Scholar
7. Bartelt R , Sperling JW , Schleck CD , Cofield RH . Shoulder arthroplasty in patients aged fifty-five years or younger with osteoarthritis . J Shoulder Elbow Surg . 2011 ; 20 ( 1 ): 123 – 130 . Crossref PubMed Google Scholar
8. Hammond LJ , Lin EC , Harwood DP , et al. Clinical outcomes of hemiarthroplasty and biological resurfacing in patients aged younger than 50 years . J Shoulder Elbow Surg . 2013 ; 22 ( 10 ): 1345 – 1351 . Crossref PubMed Google Scholar
9. Neyton L , Kirsch JM , Collotte P , et al. Mid- to long-term follow-up of shoulder arthroplasty for primary glenohumeral osteoarthritis in patients aged 60 or under . J Shoulder Elbow Surg . 2019 ; 28 ( 9 ): 1666 – 1673 . Crossref PubMed Google Scholar
10. Roberson TA , Bentley JC , Griscom JT , et al. Outcomes of total shoulder arthroplasty in patients younger than 65 years: a systematic review . J Shoulder Elbow Surg . 2017 ; 26 ( 7 ): 1298 – 1306 . Crossref PubMed Google Scholar
11. Bedeir YH , Gawish HM , Grawe BM . Outcomes of reverse total shoulder arthroplasty in patients 60 years of age or younger: a systematic review . J Hand Surg Am . 2020 ; 45 ( 3 ): 254 . Crossref PubMed Google Scholar
12. Critchley O , McLean A , Page R , et al. Reverse total shoulder arthroplasty compared to stemmed hemiarthroplasty for proximal humeral fractures: a registry analysis of 5946 patients . J Shoulder Elbow Surg . 2020 ; 29 ( 12 ): 2538 – 2547 . Crossref PubMed Google Scholar
13. Rasmussen JV , Hole R , Metlie T , et al. Anatomical total shoulder arthroplasty used for glenohumeral osteoarthritis has higher survival rates than hemiarthroplasty: a Nordic registry-based study . Osteoarth Cartil . 2018 ; 26 ( 5 ): 659 – 665 . Crossref PubMed Google Scholar
14. McBride AP , Ross M , Duke P , et al. Shoulder joint arthroplasty in young patients: Analysis of 8742 patients from the Australian Orthopaedic Association National Joint Replacement Registry . Shoulder Elbow . 2023 ; 15 ( 1 Suppl ): 41 – 52 . Crossref PubMed Google Scholar
15. Carroll RM , Izquierdo R , Vazquez M , Blaine TA , Levine WN , Bigliani LU . Conversion of painful hemiarthroplasty to total shoulder arthroplasty: long-term results . J Shoulder Elbow Surg . 2004 ; 13 ( 6 ): 599 – 603 . Crossref PubMed Google Scholar
16. Merolla G , Wagner E , Sperling JW , Paladini P , Fabbri E , Porcellini G . Revision of failed shoulder hemiarthroplasty to reverse total arthroplasty: analysis of 157 revision implants . J Shoulder Elbow Surg . 2018 ; 27 ( 1 ): 75 – 81 . Crossref PubMed Google Scholar
17. Terrier A , Obrist R , Becce F , Farron A . Cement stress predictions after anatomic total shoulder arthroplasty are correlated with preoperative glenoid bone quality . J Shoulder Elbow Surg . 2017 ; 26 ( 9 ): 1644 – 1652 . Crossref PubMed Google Scholar
18. Rasmussen JV , Olsen BS , Al-Hamdani A , Brorson S . Outcome of revision shoulder arthroplasty after resurfacing hemiarthroplasty in patients with glenohumeral osteoarthritis . J Bone Joint Surg Am . 2016 ; 98-A ( 19 ): 1631 – 1637 . Crossref PubMed Google Scholar
19. Davidson JA . Characteristics of metal and ceramic total hip bearing surfaces and their effect on long-term ultra high molecular weight polyethylene wear . Clin Orthop Relat Res . 1993 ; 294 ( 294 ): 361 – 378 . PubMed Google Scholar
20. Greis M , Heubach K , Hoberg M , Irlenbusch U . Proximal humeral bone loss in stemless shoulder arthroplasty: potential factors influencing bone loss and a new classification system . Arch Orthop Trauma Surg . 2023 ; 143 ( 6 ): 3085 – 3090 . Crossref PubMed Google Scholar
21. Jordan RW , Kelly CP , Pap G , et al. Mid-term results of a stemless ceramic on polyethylene shoulder prosthesis - a prospective multicentre study . Shoulder Elbow . 2021 ; 13 ( 1 ): 67 – 77 . Crossref PubMed Google Scholar
22. Bell SN , Christmas M , Coghlan JA . Proximal humeral osteolysis and glenoid radiolucent lines in an anatomic shoulder arthroplasty: a comparison of a ceramic and a metal humeral head component . J Shoulder Elbow Surg . 2020 ; 29 ( 5 ): 913 – 923 . Crossref PubMed Google Scholar
23. Clarke IC , Gustafson A . Clinical and hip simulator comparisons of ceramic-on-polyethylene and metal-on-polyethylene wear . Clin Orthop Relat Res . 2000 ; 379 ( 379 ): 34 – 40 . Crossref PubMed Google Scholar
24. Davies A , Sabharwal S , Liddle AD , Zamora Talaya MB , Rangan A , Reilly P . Revision rate in metal compared to ceramic humeral head total shoulder arthroplasty and hemiarthroplasty . Bone Joint J . 2024 ; 106-B ( 5 ): 482 – 491 . Crossref PubMed Google Scholar
25. Ajdari N , Tempelaere C , Masouleh MI , et al. Hemiarthroplasties: the choice of prosthetic material causes different levels of damage in the articular cartilage . J Shoulder Elbow Surg . 2020 ; 29 ( 5 ): 1019 – 1029 . Crossref PubMed Google Scholar
26. Müller LP , Degreif J , Rudig L , Mehler D , Hely H , Rommens PM . Friction of ceramic and metal hip hemi-endoprostheses against cadaveric acetabula . Arch Orthop Trauma Surg . 2004 ; 124 ( 10 ): 681 – 687 . Crossref PubMed Google Scholar
27. Wimmer MA , Pacione C , Yuh C , et al. Articulation of an alumina-zirconia composite ceramic against living cartilage - an in vitro wear test . J Mech Behav Biomed Mater . 2020 ; 103 : 103531 . Crossref PubMed Google Scholar
28. Chan SMT , Neu CP , Komvopoulos K , Reddi AH , Di Cesare PE . Friction and wear of hemiarthroplasty biomaterials in reciprocating sliding contact with articular cartilage . J Tribol . 2011 ; 133 ( 4 ). Crossref Google Scholar
29. Patel AM , Spector M . Tribological evaluation of oxidized zirconium using an articular cartilage counterface: a novel material for potential use in hemiarthroplasty . Biomaterials . 1997 ; 18 ( 5 ): 441 – 447 . Crossref PubMed Google Scholar
30. Lee YS , Lee TQ . Specimen-specific method for quantifying glenohumeral joint kinematics . Ann Biomed Eng . 2010 ; 38 ( 10 ): 3226 – 3236 . Crossref PubMed Google Scholar
31. Lawrence RL , Roseni K , Bey MJ . Correspondence between scapular anatomical coordinate systems and the 3D axis of motion: a new perspective on an old challenge . J Biomech . 2022 ; 145 : 111385 . Crossref PubMed Google Scholar
32. Saikko V . Effect of lubricant protein concentration on the wear of ultra-high molecular weight polyethylene sliding against a CoCr counterface . J Tribol . 2003 ; 125 ( 3 ): 638 – 642 . Crossref Google Scholar
33. No authors listed . For In Vitro Diagnostic Materials: Efficacy vs. Thimerosal and Sodium Azide . Sigma-Aldrich . 2020 . https://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/marketing/global/documents/179/781/proclin-preservatives-for-in-vitro-diagnostic-materials-flyer-mk.pdf ( date last accessed 12 September 2024 ). Google Scholar
34. Liao YS , McKellop H , Lu Z , Campbell P , Benya P . The effect of frictional heating and forced cooling on the serum lubricant and wear of UHMW polyethylene cups against cobalt-chromium and zirconia balls . Biomaterials . 2003 ; 24 ( 18 ): 3047 – 3059 . Crossref PubMed Google Scholar
35. Saikko V . A hip wear simulator with 100 test stations . Proc Inst Mech Eng H . 2005 ; 219 ( 5 ): 309 – 318 . Crossref PubMed Google Scholar
36. Wu W , Lee PVS , Bryant AL , Galea M , Ackland DC . Subject-specific musculoskeletal modeling in the evaluation of shoulder muscle and joint function . J Biomech . 2016 ; 49 ( 15 ): 3626 – 3634 . Crossref PubMed Google Scholar
37. Das Neves Borges P , Forte AE , Vincent TL , Dini D , Marenzana M . Rapid, automated imaging of mouse articular cartilage by microCT for early detection of osteoarthritis and finite element modelling of joint mechanics . Osteoarth Cartil . 2014 ; 22 ( 10 ): 1419 – 1428 . Crossref PubMed Google Scholar
38. No authors listed . Implants for surgery — Partial and total hip joint prostheses — Part 2: Articulating surfaces made of metallic, ceramic and plastics materials (ISO 7206-2:2011/Amd 1:2016) . International Organization for Standardization (ISO) . 2016 . https://www.iso.org/standard/67929.html ( date last accessed 25 September 2024 ). Google Scholar
39. Landis JR , Koch GG . The measurement of observer agreement for categorical data . Biometrics . 1977 ; 33 ( 1 ): 159 – 174 . Crossref PubMed Google Scholar
40. Jung M , Wieloch P , Lorenz H , et al. Comparison of cobalt chromium, ceramic and pyrocarbon hemiprostheses in a rabbit model: ceramic leads to more cartilage damage than cobalt chromium . J Biomed Mater Res . 2008 ; 85B ( 2 ): 427 – 434 . Crossref PubMed Google Scholar
41. McGann ME , Vahdati A , Wagner DR . Methods to assess in vitro wear of articular cartilage . Proc Inst Mech Eng H . 2012 ; 226 ( 8 ): 612 – 622 . Crossref PubMed Google Scholar
42. Peshkova M , Lychagin A , Lipina M , et al. Gender-related aspects in osteoarthritis development and progression: a review . Int J Mol Sci . 2022 ; 23 ( 5 ): 2767 . Crossref PubMed Google Scholar
43. Tschon M , Contartese D , Pagani S , Borsari V , Fini M . Gender and sex are key determinants in osteoarthritis not only confounding variables. a systematic review of clinical data . J Clin Med . 2021 ; 10 ( 14 ): 3178 . Crossref PubMed Google Scholar
44. Hanna FS , Teichtahl AJ , Wluka AE , et al. Women have increased rates of cartilage loss and progression of cartilage defects at the knee than men: a gender study of adults without clinical knee osteoarthritis . Menopause . 2009 ; 16 ( 4 ): 666 – 670 . Crossref PubMed Google Scholar
45. Mueller U , Braun S , Schroeder S , et al. Influence of humeral head material on wear performance in anatomic shoulder joint arthroplasty . J Shoulder Elbow Surg . 2017 ; 26 ( 10 ): 1756 – 1764 . Crossref PubMed Google Scholar
46. Wang A , Polineni VK , Essner A , et al. The significance of nonlinear motion in the wear screening of orthopaedic implant materials . J Test Eval . 1997 ; 25 ( 2 ): 239 – 245 . Crossref Google Scholar
47. Bragdon CR , O’Connor DO , Lowenstein JD , Jasty M , Syniuta WD . The importance of multidirectional motion on the wear of polyethylene . Proc Inst Mech Eng Part H J Eng Med . 1996 ; 210 ( 3 ): 157 – 165 . Crossref PubMed Google Scholar
48. Meyer DC , Riedo S , Eckers F , Carpeggiani G , Jentzsch T , Gerber C . Small anteroposterior inclination of the acromion is a predictor for posterior glenohumeral erosion (B2 or C) . J Shoulder Elbow Surg . 2019 ; 28 ( 1 ): 22 – 27 . Crossref PubMed Google Scholar
49. Herschel R , Wieser K , Morrey ME , Ramos CH , Gerber C , Meyer DC . Risk factors for glenoid erosion in patients with shoulder hemiarthroplasty: an analysis of 118 cases . J Shoulder Elbow Surg . 2017 ; 26 ( 2 ): 246 – 252 . Crossref PubMed Google Scholar
50. Zumstein V , Kraljević M , Conzen A , Hoechel S , Müller-Gerbl M . Thickness distribution of the glenohumeral joint cartilage: a quantitative study using computed tomography . Surg Radiol Anat . 2014 ; 36 ( 4 ): 327 – 331 . Crossref PubMed Google Scholar
51. Graichen H , Jakob J , von Eisenhart-Rothe R , Englmeier K-H , Reiser M , Eckstein F . Validation of cartilage volume and thickness measurements in the human shoulder with quantitative magnetic resonance imaging . Osteoarthritis Cartilage . 2003 ; 11 ( 7 ): 475 – 482 . Crossref PubMed Google Scholar
52. Schleich C , Bittersohl B , Antoch G , Krauspe R , Zilkens C , Kircher J . Thickness distribution of glenohumeral joint cartilage: a normal value study on asymptomatic volunteers using 3-Tesla magnetic resonance tomography . Cartilage . 2017 ; 8 ( 2 ): 105 – 111 . Crossref PubMed Google Scholar
53. Schmid TM , Li J , Muehleman C , Wimmer M , Irving TC . Cartilage compression changes collagen fiber orientation as measured by small angle x-ray diffraction . 50th Annual Meeting of the Orthopaedic Research Society . 2004 . Google Scholar
54. Klemt C , Prinold JA , Morgans S , et al. Analysis of shoulder compressive and shear forces during functional activities of daily life . Clin Biomech (Bristol, Avon) . 2018 ; 54 : 34 – 41 . Crossref PubMed Google Scholar
55. Langohr GDG , Haverstock JP , Johnson JA , Athwal GS . Comparing daily shoulder motion and frequency after anatomic and reverse shoulder arthroplasty . J Shoulder Elbow Surg . 2018 ; 27 ( 2 ): 325 – 332 . Crossref PubMed Google Scholar
56. Westerhoff P , Graichen F , Bender A , et al. In vivo measurement of shoulder joint loads during activities of daily living . J Biomech . 2009 ; 42 ( 12 ): 1840 – 1849 . Crossref PubMed Google Scholar
57. Smith SL , Kennard E , Joyce TJ . Shoulder simulator wear test of five contemporary total shoulder prostheses with three axes of rotation and sliding motion . Biotribol . 2018 ; 13 : 36 – 41 . Crossref Google Scholar
Author contributions
H. Mahmud: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing
D. Wang: Methodology, Writing – review & editing
A. Topan-Rat: Validation, Writing – review & editing
A. M. J. Bull: Supervision, Writing – review & editing
C. H. Heinrichs: Conceptualization, Resources, Writing – review & editing
P. Reilly: Conceptualization, Funding acquisition, Methodology, Supervision, Writing – review & editing
R. Emery: Conceptualization, Supervision, Writing – review & editing
A. A. Amis: Methodology, Writing – review & editing
U. N. Hansen: Conceptualization, Funding acquisition, Methodology, Supervision, Writing – review & editing
Funding statement
The authors disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: this study was funded by Orthopaedic Research UK (ORUK; ref: 539) and through the Brunei UBD Chancellor’s PhD Scholarship, as reported by U. N. Hansen and H. Mahmud, respectively. The cost of the implants was funded by Mathys Ltd Bettlach, as reported by C. H. Heinrichs and U. N. Hansen. The wear simulator was developed within the Wellcome Trust and Engineering & Physical Sciences Research Council (EPSRC) funded Medical Engineering Solutions in Osteoarthritis Centre of Excellence at Imperial College London, grant 088844/Z/09/Z, as reported by A. A. Amis. Consulting fees related to the study were also received from Imperial College London, as reported by D. Wang.
ICMJE COI statement
H. Mahmud reports funding from the Brunei UBD Chancellor’s PhD Scholarship, related to this study. D. Wang reports research consulting fees from Imperial College London, related to this study. C. H. Heinrichs and U. N. Hansen report the provision of implants by orthopaedic manufacturer Mathys Ltd Bettlach, related to this study. A. A. Amis and R. Emery report an institutional grant (paid to Imperial College London) from the Wellcome Trust to fund the project to build the shoulder wear simulator, related to this study.
Data sharing
The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request.
Ethical review statement
Infrastructure support for this research was provided by the National Institute for Health and Care Research (NIHR) Biomedical Research Centre (BRC), based at Imperial College Healthcare NHS Trust and Imperial College London, and is approved by REC3 Wales (17/WA/0161; Project R20052-2A; HTA license 12275).
Open access funding
The authors report that they received open access funding for their manuscript from Orthopaedic Research UK (ORUK; ref: 539).
Open access statement
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives (CC BY-NC-ND 4.0) licence, which permits the copying and redistribution of the work only, and provided the original author and source are credited. See https://creativecommons.org/licenses/by-nc-nd/4.0/
This article was primary edited by J. Scott.
