INTRODUCTION
Anatomical total shoulder arthroplasty (aTSA) is typically reserved for end-stage glenohumeral arthritis with an intact rotator cuff and adequate glenoid bone stock. The primary objectives of shoulder arthroplasty are to reduce pain, restore range of motion, and improve quality of life. Due to its proven reliability and effectiveness, the demand for primary arthroplasty is anticipated to increase by up to 322% between 2018 and 2050 [1], in part because patients are living longer, and in part because shoulder arthroplasty is also being increasingly used in middle-aged patients.
Given the growing incidence of shoulder arthroplasty surgery, the demand for revision and re-revision surgery is also expected to increase. A substantial socioeconomic and health burden is observed in patients undergoing re-revision surgery [2,3], and it is anticipated that healthcare resources will face increased pressure due to the projected rise in demand for revision surgery in shoulder arthroplasty [4]. Moreover, revision surgery increases the risk of future complications and subsequent re-revisions [5]. Gill et al. [6] conducted an analysis of the Australian National Registry between 2004 and 2018 and reported re-revision rates greater than 20% after revision shoulder arthroplasty. These numbers are comparable to those observed after revisions of total hip and knee arthroplasty [7,8].
Few studies have specifically addressed re-revision rates after an anatomic-to-anatomic revision of aTSA [9-11]. Previous studies that reported the indications and survival of re-revision surgery in revised aTSA had short follow-up periods and were mainly cohort studies [9-11], which are often prone to selection bias, publication bias, and conflicts of interest.
When revision aTSA surgery is required, the surgeon typically has two options: revising the anatomical configuration to a reverse total shoulder arthroplasty (rTSA) or retaining the anatomical configuration. Rotator cuff insufficiency, dislocation, and aseptic glenoid loosening are common indications for revising aTSA to rTSA [12-14]. Although it is less commonly performed, revision to aTSA can be a reasonable option in certain patients. Surgeons might choose to retain the anatomic configuration to preserve bone stock, maintain native joint biomechanics, and potentially optimize range of motion, compared with conversion to rTSA [15].
To be considered for maintained anatomical configuration after revision, patients must have an intact rotator cuff and adequate glenoid bone stock and need revision due to glenoid or humeral component loosening, malalignment, or periprosthetic fractures [16,17]. However, even in that subset of patients, the results remain unpredictable, and data on this option are insufficient [17].
Data from a national registry can offer an accurate report of implant survival in a large, nationwide cohort. In addition, a large national cohort allows surgeries with a very low incidence, such as anatomic-to-anatomic revisions of aTSA, to be analyzed. Therefore, this study uses data from the Dutch National Registry to describe re-revision rates and analyze implant survival after an anatomic-to-anatomic revision of aTSA. Patient characteristics and indications for re-revision surgery are also reported. Given the low incidence of anatomic-to-anatomic revisions, this study is strictly descriptive, focusing on re-revision rates and implant survival, without attempting to identify risk factors or perform predictive analyses.
METHODS
Ethical approval for this study was obtained from the Landelijke Registratie Orthopedische Interventies (LROI; No. LROI2024-131). The LROI operates under an opt-out system for informed consent; therefore, formal informed consent from patients was waived.
National Registry
The Dutch Arthroplasty Register (Landelijke Registratie Orthopedische Implantaten [LROI]) is a nationwide collaboration that has documented shoulder arthroplasty procedures since 2014. Data are collected using a standardized form that is completed after each surgical procedure and submitted to the national registry for all primary and revision shoulder arthroplasties (Supplementary Material 1). From this registry, demographic, surgical, and implant-related data can be extracted.
The overall data completeness is currently 97% for primary shoulder arthroplasty and 95% for revision arthroplasty [18]. Missing records can result from emergency admissions, new registration staff, missing patient IDs, or data entry issues [19]. The vital status of all patients is actively obtained on a regular basis from Vektis, the national insurance database on healthcare in the Netherlands, via a trusted third party.
In the LROI, a revision procedure is defined as a procedure in which at least one implant component is added, removed, or replaced. Consequently, procedures such as cuff repair, tendon transfer, or debridement and antibiotics or implant retention without an implant change following aTSA are not recorded. In cases of re-revision surgery, surgeons may select multiple indications, if necessary.
Study Design
This study is a nationwide, registry-based descriptive study with data from the LROI. After the national registry approved the protocol, the anonymized data were made available for analysis. The provided data cannot be traced back to individual patients, surgeons, or institutions.
For this study, the following data were requested: patient demographics (sex, age, body mass index [BMI], smoking, American Society of Anesthesiologists score, previous surgery), indication for surgery (e.g., infection, periprosthetic fracture, malalignment, instability, component loosening), surgery characteristics (revised prosthesis, type of revision, bone graft, cementing) and implant characteristics (replaced component).
Data
Data on all aTSA procedures performed between 2014 and 2023 were extracted from the LROI. Data were collected until December 2024, which ensured that even procedures performed in the final year of inclusion had a minimum follow-up period of 1 year. All patients who underwent revision surgery for aTSA were included, regardless of which component was revised, provided that the implant remained in an anatomic configuration after the first revision (anatomic-to-anatomic revision).
First revisions in which the aTSA was converted to rTSA were excluded. Revision surgeries for which no data on the primary surgery were available were excluded by the registry, given that the history of those procedures was unknown. For the primary outcome, any surgical arthroplasty-related procedure performed after the first revision of an aTSA was considered to be a re-revision, including cases in which a revised aTSA was converted to an rTSA. Two-stage revision surgery was considered one revision procedure.
Inclusion
The LROI registered 5,547 primary aTSA cases between 2014 and 2023 (Fig. 1). Revision surgery was performed in 281 patients, with conversion to rTSA being performed in 208 cases (74%). Fifteen revision cases were excluded from the analysis because the implanted component was not documented. In total, 58 cases of revision surgery in which the anatomical configuration was preserved are included in this study, including three patients who underwent two-stage revision procedures.
Statistics
Patient and treatment characteristics are reported using descriptive statistics. Binary and categorical variables are presented as numbers and percentages. Continuous variables are presented as means and standard deviations (SDs) for normally distributed data and medians and interquartile ranges (IQRs) for skewed data. For the primary outcome, a survival analysis was performed according to the Kaplan-Meier method. Re-revision was defined as the event, with the time from the first revision surgery to re-revision or the end of the included period (December 2024) as the survival time. Cases in which the patient died were censored at the time of death. Survival data are reported using a survival plot and percentages for the 1-, 3-, 5-, and 8-year survival rates. Reasons for re-revision and the replaced components are also reported. Data entry and statistical analysis were performed using IBM SPSS version 26 (IBM Corp.).
RESULTS
Primary Surgery
The Dutch Arthroplasty Register registered 5,547 primary anatomic shoulder arthroplasties. The mean age at primary surgery was 66 years (SD, 9 years). Most patients were female (68%, n=3,795). Patient demographics for the primary aTSA cohort are presented in Table 1.
Cohort
In total, 58 cases of revision surgery in which the anatomical configuration was preserved are included in this study, including three patients who underwent two-stage revision procedures. The mean age at the time of revision surgery was 60 years (SD, 11 years), with a slight predominance of female patients (55%, n=32). The mean BMI was 28 kg/m2 (SD, 5 kg/m2), and eight patients (14%) were identified as current smokers (Table 2).
The deltopectoral approach was used in 97% of the cases (n=56). Common indications for revision surgery were glenoid component loosening (n=15) and malalignment (n=14). The indications classified as “other” were metal allergy, rotator cuff rupture, humeral head exchange, and liner dislocation (n=2 each), along with arthrofibrosis, glenoid component dislocation, subluxation, hematoma evacuation, overstuffing, pain from a long posterior screw, and one unspecified case (n=1 each). A partial revision was performed in 42 patients, and most components were placed cementless (n=31) (Table 3).
Outcome
Among the 58 included patients, re-revision surgery was performed in 10 patients (17%). The median follow-up of the cohort was 3.0 years (IQR, 1.6–5.3 years). The median time to re-revision was 1.1 years (IQR, 0.2–1.9 years). Implant survival was 91% after 1 year (95% CI, 84%–98%) and 80% after 3 years (95% CI, 72%–93%). All re-revisions occurred within 3 years after revision surgery (Fig. 2). Two patients from the revision cohort died within the follow-up period; 46 patients were alive without undergoing re-revision surgery.
All patients received rTSA during re-revision surgery except for one patient whose prosthesis was removed due to infection without the insertion of a new implant. When documenting re-revision surgeries, surgeons were able to select multiple indications if necessary. The most common indications for re-revision surgery were cuff rupture (n=4) and glenoid component loosening (n=4), followed by infection, contained glenoid defect, humeral component loosening, and instability (Table 4).
DISCUSSION
The purpose of this study was to assess the re-revision rate after an anatomic-to-anatomic revision of aTSA in the Netherlands. Our results show a high rate of re-revision surgery in revised aTSA, with 10 patients (17%) undergoing re-revision. The implant survival of aTSA after revision surgery was 91% at 1 year and 80% at 3 years.
However, it is important to note that the patients selected for the retention of an anatomic configuration likely represented a highly selected and favorable subgroup with adequate bone stock and presumed rotator cuff integrity. Despite that selection, our results show a high early re-revision rate. Technical or indication-related factors might have contributed to these failures, given that the very short median time to re-revision makes wear-related mechanisms unlikely. To date, only a few studies have described implant survival after anatomic-to-anatomic revision [9-11]. In 2008, Cheung et al. [9] reviewed survival rates following revision aTSA due to glenoid component loosening and found the 5-year survival rate to range between 78% and 91%. More recently, Sheth et al. [10] reported a 66% survival rate 4 years after revision aTSA in a cohort of 20 patients. Dillon et al. [11] reported that more than 20% of anatomic-to-anatomic revisions required further revision surgery and concluded that conversion to rTSA might be a more reliable treatment option. In support of that notion, Hao et al. [20] recently reported 2-year implant survivorship of 97% following conversion to rTSA after primary aTSA.
All the re-revisions in our data occurred within 3 years after revision surgery. This finding is consistent with the literature, which reports that most failures after shoulder arthroplasty surgery occur within the first 4 years after the index surgery [17,21]. The most common indications for re-revision were glenoid component loosening (n=4) and rotator cuff rupture (n=4), both of which are common indications for anatomic-to-reversed conversion and align with the findings reported in previous studies [12-14,21,22]. The rotator cuff was likely already compromised, though unrecognized by the surgeon, at the time of the first revision. However, that could not be confirmed here due to data availability. Cuff rupture can cause proximal humerus superior migration, which imbalances shoulder biomechanics, contributing to edge loading and accelerated wear of the glenoid component [23]. Therefore, it is not unexpected that, in our study, all cases initially revised to an anatomic configuration were ultimately converted to rTSA when re-revision surgery was required. The only exception was one patient whose prosthesis was removed without the implantation of a new device.
Compared with the general primary aTSA population, patients undergoing revision surgery were younger (mean age, 60 vs. 66 years) and more frequently had a history of previous surgery (21% vs. 13%). The main indications for revision surgery in our study were glenoid component loosening (n=15) and malalignment (n=14), followed by post-traumatic sequelae (n=7). Those indications have also been reported in the literature as the “select cases” in which anatomic-to-anatomic revision of aTSA can reasonably be considered [15,16]. As such, patients retaining an anatomic configuration in our study likely represented a favorable, selected cohort. Nonetheless, the high early re-revision rate observed (17%) underscores the need for cautious decision-making when considering anatomic-to-anatomic revision surgery.
Because 90% of the patients who underwent re-revision surgery were ultimately converted to rTSA, this study serves as a warning for surgeons considering the retention of an anatomic configuration. The defined failure causes, such as glenoid component loosening and cuff rupture, highlight the underlying mechanisms. These findings underscore the need to consider conversion to reverse shoulder arthroplasty in specific cases, given its potential advantages in improving joint stability and functional outcomes [12-14].
Our study offers valuable insights into the survival rate and indications for re-revision surgery in aTSA. Re-revision surgeries are known to impose physical and psychological burdens on both the individual and the community [24]. In this context, the results derived from our study are important because they warn surgeons of high early failure rates in anatomic-to-anatomic revision surgery. Re-revision surgery in aTSA requires a comprehensive understanding of the underlying causes of failure. Given the rarity of anatomic-to-anatomic revisions, future studies should focus on international registry collaborations to pool data, which could provide sufficient power to investigate true risk factors and improve patient selection.
The results of this study should be interpreted in light of some limitations. The main limitation of this study is the small cohort size, particularly the number of re-revision events (n=10), which was insufficient to perform a regression analysis of factors associated with re-revision. Therefore, this study should be interpreted as descriptive only. Another limitation is potential selection bias because the registry does not capture the clinical reasoning behind retaining an anatomic implant. However, even in this selected cohort, the re-revision rate was high. Also, patient-reported outcome measures were not available in the registry. Although implant survival provides useful information, it is a crude metric and might not reflect functional outcomes or patient satisfaction. For data collection, the registry relies on the accuracy and completeness of the reporting parties, which could introduce reporting or misclassification bias. However, using data from a registry also provides advantages: it allows for the identification of trends and associations in a large cohort, and it increases the generalizability of the results. The results of this study are specific to the Netherlands and might not be generalizable to other countries. Furthermore, it is a retrospective study, and the data were not gathered with a specific hypothesis in mind. Therefore, the conclusions are limited to association and not causation.
Additionally, the follow-up duration was short and varied among patients, which might have resulted in an underestimation of long-term failure rates.
CONCLUSIONS
Anatomic-to-anatomic revision of aTSA was associated with a re-revision rate of 17% and an implant survival rate of 80% at 3 years. The failure causes and numbers of re-revision surgeries in aTSA indicate a high risk of early failure and provide a warning for surgeons considering retention of an anatomic configuration. The interpretation of these findings is limited by the small cohort size and descriptive study design.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement
Print
