Mid-term Outcomes of Shoulder Resurfacing Arthroplasty for Glenohumeral Osteoarthritis
Abstract
Introduction: Shoulder resurfacing arthroplasty has gained popularity as a bone-preserving alternative to conventional total or hemi shoulder arthroplasty in patients with glenohumeral osteoarthritis. By maintaining native humeral head geometry and bone stock, resurfacing facilitates easier future revisions and potentially restores more natural joint kinematics. This study evaluates the mid-term clinical, functional, and radiological outcomes of humeral head resurfacing using the Global® CAP® (DePuy Synthes, Warsaw, IN, USA) prosthesis performed at a single centre, with a average follow-up of about 6 years.
Materials and Methods: A retrospective observational study was conducted of 18 shoulders (15 patients) who underwent shoulder resurfacing arthroplasty between 2014 and 2020 for primary glenohumeral osteoarthritis. Inclusion criteria included intact rotator cuff, adequate humeral head bone stock (>60%), and minimum 4 year follow-up. All procedures were performed through a deltopectoral approach by a single surgeon. The Global® CAP® humeral resurfacing prosthesis was implanted; no glenoid component was used. Clinical evaluation included active range of motion (ROM), Visual Analogue Scale (VAS) for pain, and Constant–Murley score (CMS). Radiographs were assessed for component positioning, loosening, and glenoid erosion. Statistical analysis was performed using the SPSS software.
Results: The mean patient age was 55.4 years (range 38–70). The average follow-up duration was 6.2 years (range 4–9). Mean active forward flexion improved from 94° ± 16° preoperatively to 150° ± 12° at final follow-up (p <0.05); abduction increased from 87° ± 18° to 140° ± 15° (p <0.05); and external rotation from 10° ± 8° to 45° ± 10° (p < 0.05). Mean VAS pain score decreased from 7.1 ± 1.2 to 1.3 ± 0.7 (p < 0.05). Mean CMS improved from 31.2 ± 7.8 to 78.5 ± 9.2 (p < 0.05). Radiographically, all implants demonstrated stable fixation without radiolucent lines or loosening. Mild varus positioning (mean neck-shaft angle 133° ± 3°) was noted in approximately one-third of cases, but without clinical consequence. No patient required revision surgery, and there were no cases of infection, periprosthetic fracture, or glenoid medialization.
Conclusion: Mid-term outcomes following shoulder resurfacing arthroplasty using the Global® CAP® prosthesis for glenohumeral osteoarthritis demonstrated good pain relief, functional improvement, and implant survivorship. The procedure offers predictable results when performed in appropriately selected patients with preserved rotator cuff function and adequate humeral bone stock. Humeral resurfacing remains a reliable, bone-conserving alternative to stemmed arthroplasty for younger and active patients, with the added benefit of simple future conversion if required.
Keywords: Shoulder resurfacing arthroplasty, Global CAP, Glenohumeral osteoarthritis, Bone preservation
Introduction
The practice of shoulder arthroplasty has evolved markedly over more than a century. The first shoulder prosthesis was implanted by French surgeon Péan in 1893, but it was Neer’s pioneering work in the 1950s that popularized shoulder replacement for fractures and osteoarthritis (1). Early designs were simple hemiarthroplasties or total shoulder replacements (TSA), each with drawbacks such as glenoid wear or component loosening. These limitations spurred interest in joint-surface-preserving implants (surface replacements) to better restore anatomy and preserve bone stock (2). Humeral head resurfacing arthroplasty (HRA) involves reaming only the epiphyseal surface of the humeral head and capping it with a prosthetic “cap”, avoiding a stemmed humeral implant (3). In contrast to stemmed arthroplasties, resurfacing retains the native humeral neck and most of the head, preserving bone and the native head-shaft angle (1) (4). Modern resurfacing designs have benefited from advances in hip resurfacing: first-generation models used cement fixation, followed by implants with a central stem and porous or hydroxyapatite-coated surfaces for bone ingrowth (5). Contemporary devices (e.g. Copeland, Global CAP) offer various sizes and offsets, with improved fixation. The goal of shoulder resurfacing is to reproduce the patient’s native anatomy (head size, version, offset) and kinematics. Biomechanical studies suggest resurfacing can restore normal rotation centers and avoid joint “overstuffing” if correctly sized (6).
The theoretical advantages of humeral resurfacing arthroplasty include preservation of proximal humeral bone stock (beneficial for younger patients) and simpler revision if needed (7). Resurfacing avoids complications of stems (e.g. humeral shaft fracture) and may yield faster recovery with less blood loss. Moreover, resurfacing allows easier conversion to anatomic or reverse TSA if glenoid arthritis or cuff failure later develops. However, shoulder resurfacings have a reported high revision rate: they account for about one-third of shoulder arthroplasties and have the highest revision rates of any shoulder procedure (6).
Mid-term outcome data for resurfacing arthroplasty are still emerging. In a multicenter review, Ingoe et al. found that Global® CAP® (DePuy) hemi-resurfacing had 80% implant survival at 7 years, comparable to other resurfacing results (6). Giannotti et al. (2023) recently reported excellent functional recovery and minimal loosening over 11-year follow-up in 33 HRA cases (8). To our knowledge, no study has yet reported outcomes of Global® CAP® resurfacing from our region. We therefore performed a single-center retrospective cohort analysis of patients with glenohumeral osteoarthritis treated with Global® CAP® resurfacing. We hypothesized that, at mid-term (mean >5 years) follow-up, patients would show significant pain relief and functional gains without major implant problems.
Materials and Methods
We retrospectively reviewed all patients who underwent humeral head resurfacing arthroplasty for primary glenohumeral osteoarthritis at our institution between January 2014 and December 2020. The study was approved by the institutional review board (IRB approval was waived given the retrospective design). Informed consent for data collection was obtained from all patients.
Inclusion criteria were: (1) primary glenohumeral osteoarthritis (Walch A1/A2 type glenoid or B1, according to preoperative imaging), (2) intact rotator cuff, (3) sufficient humeral head bone stock (>60% of articular surface) for resurfacing, (4) age 18–70 years, and (5) available follow-up of 4≥ years. Exclusion criteria included inflammatory arthritis, large rotator cuff tears, glenoid erosion requiring anatomic TSA, prior shoulder arthroplasty, or insufficient bone stock. We identified 15 patients (18 shoulders) meeting these criteria. Of these, 8 were male and 7 female, with mean age 55.4 years (range 38–70). Three patients had bilateral procedures (one simultaneously in a staged manner, two sequentially).
Surgical Technique: All operations were performed by a single shoulder surgeon. A standard deltopectoral approach was used in all cases. The subscapularis tendon was exposed, tagged and tenotomized approximately 1 cm from its insertion on the lesser tuberosity. The long head of the biceps was tenotomized in all cases without tenodesis. The humeral head was dislocated anteriorly and surrounding osteophytes were removed to outline the neck and anatomic landmarks. Care was taken to achieve appropriate version and inclination matching the patient’s anatomy. With the arm in neutral rotation, a guide pin was placed centrally through the humeral head parallel to the native neck. A cannulated reamer was then used over the guide pin to sequentially remove articular cartilage and shape the residual subchondral bone. A central peg hole was drilled. Trial cups were placed to verify sizing, offset, and stability of trial reductions. The definitive Global CAP (DePuy Synthes) humeral resurfacing prosthesis was then inserted. The glenoid was denervated and all glenoid osteophytes were removed. The subscapularis tendon was re-approximated with multiple non-absorbable sutures. The wound was closed in layers. A schematic of these operative steps is summarized in Figure 1.
Postoperatively, the arm was supported in a shoulder sling in internal rotation for 3 weeks. Pendulum and passive motion exercises of flexion and abduction were initiated on postoperative day one under sling protection. Active-assisted elevation and abduction were allowed after 3 weeks, progressing gradually. External and internal rotations were delayed until after 6 weeks. Active strengthening was begun around 6–8 weeks post-op. Patients were advised to avoid heavy lifting or strenuous activities for 3 months. Weight training was started at 3 months. All patients were evaluated preoperatively and at regular intervals postoperatively (6 weeks, 3 months, 6 months, 1 year, and then annually). Clinical assessment included active range of motion (ROM) measured with a goniometer (forward flexion, abduction, external rotation at side, internal rotation recorded as the highest vertebral level reached). Pain was quantified by a 10-point visual analogue scale (VAS). Shoulder function was assessed by the Constant–Murley score (absolute score out of 100). Pre and postoperative Constant scores, VAS and ROM were recorded and compared. Standard anteroposterior and axial shoulder radiographs were obtained at each follow-up. Humeral component position was assessed by measuring the neck-shaft angle to detect varus or valgus inclination. Central peg seating and any radiolucent lines were noted. The glenoid was inspected for erosion or medialization of the humeral head. Radiographic osteolysis or loosening (periprosthetic lucencies) was defined if present.
Statistical Analysis: Continuous variables are presented as means with standard deviations or ranges. Paired comparisons of pre- and post-operative scores (Constant, VAS) and ROM were performed comparison Student’s t-test, with p<0.05 considered significant. No patients were lost to follow-up for functional assessment; all were contacted for any history of further shoulder procedures.
Results
The study cohort included 15 patients (18 shoulders; 11 right, 7 left). The mean patient age was 55.4±9.8 years (range 38–70) at the time of surgery. Indication was primary glenohumeral osteoarthritis in all cases. The mean follow-up duration was 6.2±0.8 years (range 4.0–9.0 years), with 14 shoulders (78%) followed >5 years.
All patients demonstrated substantial improvements in shoulder mobility. Mean active forward flexion increased from 94°±16° preoperatively to 150°±12° at final follow-up (p<0.05). Active abduction improved from 87°±18° to 140°±15° (p<0.05). External rotation at the side improved from a mean of 10°±8° pre-op to 45°±10° post-op (p<0.05). Internal rotation (measured by highest vertebral level) improved from the sacrum (mean L5 level) up to approximately T8–T7 in most shoulders. All ROM gains were statistically significant. Patients reported marked pain relief. The mean VAS pain score decreased from 7.1±1.2 (severe pain) preoperatively to 1.8±0.7 (mild pain) postoperatively (p<0.05). Correspondingly, the mean Constant–Murley shoulder score improved significantly from 31.2±7.8 (range 18–45) before surgery to 78.5±9.2 (range 65–95) at final follow-up (p<0.05). At the latest follow-up, all implants appeared well-positioned without signs of loosening or migration. No radiolucent lines were visible around any humeral component. The mean humeral neck-shaft angle was 132°±3° postoperatively (6° varus from the standard 138°), indicating that a subset of implants (approximately 6/18, 33%) were placed in slight varus (neck-shaft <130°). This varus placement rate is similar to what has been noted by others. No progressive glenoid erosion or medialization was observed on serial radiographs. One patient had a subtle 3 mm medial shift of the center of rotation at 5-year film, without clinical symptoms.
There were no intraoperative complications such as fracture or neurovascular injury. Postoperatively, one patient developed a wound hematoma that resolved with aspiration. There were no infections, no glenoid bone loss requiring treatment, and no postoperative scapular fractures. Importantly, no shoulder required reoperation or revision arthroplasty during the follow-up period. One patient sustained a minor subscapularis strain 2 years post-op during heavy lifting, managed conservatively, but with stable function thereafter.
Discussion
This single-center series demonstrates that shoulder resurfacing arthroplasty with the DePuy Global CAP prosthesis yields good mid-term results in carefully selected patients. By 4–9 years postoperatively, patients had dramatic improvements in pain and function, comparable to those reported in the literature for humeral resurfacing. Mean Constant scores rose from poor preoperative levels (~31) into the “good”/“excellent” range (~79), and VAS pain scores showed substantial pain relief, indicating good quality-of-life gains. These outcomes align with previous reports: for example, Giannotti et al. noted an increase in Constant score from 29.9 to 81.2 (p<0.05) in their HRA series (8), and Chillemi et al. reported significant improvement in Constant, VAS, and ROM in mid-term follow-up (9).
The improvements in range of motion also confirm the effectiveness of preserving native joint anatomy. Forward flexion and abduction gains (>50° increase) restored near-normal overhead motion, and external rotation averaged 45° post-op, sufficient for most daily activities. These ROM gains exceeded those reported for stemmed hemiarthroplasty in some series, suggesting that resurfacing may better maintain anatomy (10). We did not formally measure scapulothoracic motion, but no patient exhibited significant scapular dyskinesia at follow-up. Preoperatively, impaired internal rotation (reaching only the sacrum) improved to around T7–T8 level in most patients, suggesting relief of posterior cuff tightness and anterior pain.No cases required revision surgery, conversion to TSA, or suffered catastrophic implant failures. Ingoe et al. found a 7-year survival of 80% (20% revision) for Global CAP, and Giannotti et al. (mean 11-year follow-up) reported only 2 of 78 shoulders revised (3%) (6) (8). Our lack of revisions likely reflects both the younger patient age (mean 55) and rigorous selection (intact cuff, adequate bone stock). A notable concern with resurfacing is glenoid degeneration over time. We saw no clinically significant glenoid wear yet, though one shoulder showed minimal medial shift. This mirrors Giannotti’s finding of no “clinically relevant central migration” in any implant at 11 years (8). Ongoing surveillance is needed, as long-term glenoid arthrosis could emerge. Radiographically, all humeral components remained well-fixed with no osteolysis or radiolucencies. We did observe a tendency for slight varus inclination in some implants (average neck-shaft angle ~133°). Varus positioning has been described as a risk factor for later failure. Lebon et al. noted a “tendency towards varus positioning” and increased offset as possible prognostic factors for failure (11). In our series, despite mild varus in ~30% of cases, none have failed within 5 years. We attribute this to overall surgical technique and patient selection, but we remain cautious. Surgeons performing resurfacing must ensure accurate sizing and alignment to avoid ‘overstuffing’ the joint. Our results support the notion that shoulder resurfacing is especially advantageous in younger or middle-aged patients with isolated humeral-sided arthritis. By preserving bone stock, these patients retain the option of future revision to stemmed TSA or reverse arthroplasty if needed (12). Both Giannotti et al. and others emphasize that conversion from a resurfacing to a stemmed arthroplasty is relatively straightforward because the humeral canal is untouched (8). Compared to stemmed hemiarthroplasty or TSA, resurfacing may reduce certain risks. We observed negligible blood loss and none required transfusion. Our operative times averaged around 60 minutes, which is on par or shorter than reported for stemmed implants (13). The absence of a stem also avoids stress risers in the humeral shaft (14). In this series, no intraoperative humeral fractures occurred, whereas fracture rates of ~2–5% have been reported for some stemmed procedures (15). Furthermore, the bone-preserving nature may explain the absence of humeral osteolysis or subsidence on follow-up X-rays. From a literature standpoint, our findings are generally concordant with prior reports on humeral resurfacing. Levy and Copeland’s classic series reported excellent pain relief and satisfaction with the Copeland resurfacing (the predecessor to Global CAP) (16). Long-term studies (20+ years) of Copeland resurfacings have shown ~96% prosthesis survival and high patient satisfaction, underscoring durability (17). Chillemi et al. (9) also concluded that mid-term revision rates for uncemented resurfacing are relatively low, suggesting it as a viable alternative to TSA in selected cases. Notably, our implant survival (no failures at 5+ years) is as good as most published cohorts. This may reflect shorter follow-up (mean 5.2y) or smaller sample, but it is encouraging.
Limitations of this study include its retrospective design and small sample size. Being a single-center case series, our findings may not generalize to all populations. Also, follow-up (mean 6 years) is still mid-term; longer-term surveillance is required to detect late failures or glenoid wear. Despite these limitations, the consistency of outcomes across patients and alignment with existing literature suggest the results are robust.
Conclusions
In this series of 15 patients (18 shoulders) who underwent shoulder resurfacing arthroplasty for glenohumeral osteoarthritis, we found significant improvements in shoulder pain, range of motion and function at mid-term follow-up (mean 6.2 years). Resurfacing with the Global® CAP® (DePuy Synthes) prosthesis provided durable fixation with no aseptic loosening or implant revisions observed. These results align with recent literature on humeral resurfacing. Humeral head resurfacing should be considered a viable bone-conserving option for younger patients with isolated shoulder arthritis and intact soft tissues. Careful surgical technique (correct component sizing and alignment) is critical to avoid issues like varus malposition. Longer-term studies are required to assess the ultimate durability of these implants and the fate of the glenoid joint surface.
Ethical approval for this retrospective study was waived by our IRB.
Funding: None.
Conflict of Interest: The authors declare no conflicts of interest.
References
1. Levine WN, Ahmad CS, Bigliani LU, Jobin CM, Knudsen ML, Luzzi AJ. "From Neer to now - 80 years of shoulder arthroplasty". JSES Int. 2024 Apr 25;9(3):934-943. doi: 10.1016/j.jseint.2024.04.007. PMID: 40486816; PMCID: PMC12144970.
2. Twomey-Kozak J, Adu-Kwarteng K, Lunn K, Briggs DV, Hurley E, Anakwenze OA, Klifto CS. Recent Advances in the Design and Application of Shoulder Arthroplasty Implant Systems and Their Impact on Clinical Outcomes:. A Comprehensive Review. Orthop Res Rev. 2024 Jul 26;16:205-220. doi: 10.2147/ORR.S312870. PMID: 39081796; PMCID: PMC11288362.,
3. Kerzner B, Mehta N, Khan ZA, Bonadiman JA, Damodar D, Dasari SP, Fortier LM, Verma NN. Humeral Head Resurfacing for Isolated Primary Humeral Osteoarthritis With a Large Chondral Defect. Arthrosc Tech. 2022 Jun 14;11(7):e1223-e1231. doi: 10.1016/j.eats.2022.03.004. PMID: 35936842; PMCID: PMC9353195.,
4. Bidwai R, Kumar K. Outcomes of different stem sizes in shoulder arthroplasty. J Orthop. 2022 Oct 30;35:37-42. doi: 10.1016/j.jor.2022.10.009. PMID: 36387763; PMCID: PMC9647328.
5. Schmidutz, Florian & Agarwal, Yash & Müller, Peter & Gueorguiev, Boyko & Richards, Robert & Sprecher, Christoph. (2014). Stress-shielding induced bone remodeling in cementless shoulder resurfacing arthroplasty: A finite element analysis and in vivo result. J Biomech. 2014 Nov 7;47(14):3509-16. doi: 10.1016/j.jbiomech.2014.08.029. Epub 2014 Sep 9. PMID: 25278045.,
6. Ingoe H, Holland P, Tindall E, Liow R, Mcvie JL, Rangan A. Seven-year survival analysis of the Global® CAP® (Conservative Anatomic Prosthesis) shoulder resurfacing. Shoulder Elbow. 2018 Apr;10(2):87-92. doi: 10.1177/1758573217704818. Epub 2017 May 3. PMID: 29560033; PMCID: PMC5851121.,
7. Widnall JC, Dheerendra SK, Macfarlane RJ, Waseem M. The use of shoulder hemiarthroplasty and humeral head resurfacing: a review of current concepts. Open Orthop J. 2013 Sep 6;7:334-7. doi: 10.2174/1874325001307010334. PMID: 24082971; PMCID: PMC3785053.,
8. Giannotti, S., Troiano, E., Saviori, M., Crippa Orlandi, N., Greco, T., Perisano, C., Colasanti, G. B., & Mondanelli, N. (2023). Humeral Resurfacing Arthroplasty: Clinical, Functional and Radiological Assessment at a Mean Follow-Up of 11 Years. Prosthesis, 5(4), 1312-1321. https://doi.org/10.3390/prosthesis5040090,
9. Chillemi C, Paglialunga C, De Giorgi G, Proietti R, Carli S, Damo M. Outcome and revision rate of uncemented humeral head resurfacing: Mid-term follow-up study. World J Orthop. 2021 Jun 18;12(6):403-411. doi: 10.5312/wjo.v12.i6.403. PMID: 34189078; PMCID: PMC8223723.,
10. Zhang B, Chen G, Fan T, Chen Z. Resurfacing hemiarthroplasty versus stemmed hemiarthroplasty for glenohumeral osteoarthritis: a meta-analysis. Arthroplasty. 2020 Sep 1;2(1):25. doi: 10.1186/s42836-020-00045-5. PMID: 35236453; PMCID: PMC8796565.
11. Lebon J, Delclaux S, Bonnevialle N, Rongières M, Bonnevialle P, Mansat P; French Society for Shoulder, Elbow (SOFEC). Stemmed hemiarthroplasty versus resurfacing in primary shoulder osteoarthritis: a single-center retrospective series of 78 patients. Orthop Traumatol Surg Res. 2014 Oct;100(6 Suppl):S327-32. doi: 10.1016/j.otsr.2014.05.012. Epub 2014 Aug 12. PMID: 25130762.,
12. Awad G, Boutros M, Asmar C, Yammine K. Stemless vs. stemmed reverse total shoulder arthroplasty as primary treatment in the elderly: A meta-analysis. Shoulder Elbow. 2025 Oct 22:17585732251388447. doi: 10.1177/17585732251388447. Epub ahead of print. PMID: 41142069; PMCID: PMC12546121.,
13. Nixon RA, Dang KH, Haberli JE, O'Donnell EA. Surgical time and outcomes of stemmed versus stemless total shoulder arthroplasty. J Shoulder Elbow Surg. 2022 Jun;31(6S):S83-S89. doi: 10.1016/j.jse.2022.01.129. Epub 2022 Feb 13. PMID: 35172208.
14. Ryan WK, Vander Voort WD, Saad MA, Wu E, Garcia-Nolen TC, Bayne CO, Szabo RM. The effect of shoulder prosthesis stem length on failure due to torsional loading. A biomechanical study in composite humeri. JSES Int. 2023 May 24;7(5):819-826. doi: 10.1016/j.jseint.2023.04.011. PMID: 37719832; PMCID: PMC10499865.,
15. Cowling PD, Holland P, Kottam L, Baker P, Rangan A. Risk factors associated with intraoperative complications in primary shoulder arthroplasty. Acta Orthop. 2017 Dec;88(6):587-591. doi: 10.1080/17453674.2017.1362155. Epub 2017 Aug 7. PMID: 28782448; PMCID: PMC5694801.,
16. Ofer Levy, Stephen A Copeland,Cementless surface replacement arthroplasty (Copeland CSRA) for osteoarthritis of the shoulder,. Journal of Shoulder and Elbow Surgery, Volume 13, Issue 3, 2004, Pages 266-271, ISSN 1058-2746, https://doi.org/10.1016/j.jse.2004.01.005.,
17. Rai P, Davies O, Wand J, Bigsby E. Long-term follow-up of the Copeland mark III shoulder resurfacing hemi-arthroplasty. J Orthop. 2015 Oct 1;13(1):52-6. doi: 10.1016/j.jor.2015.09.003. PMID: 26955225; PMCID: PMC4761623.