Continuous clinical and radiological follow-up of isainlay reverse shoulder prostheses: results from the isa registry with a minimum of 1-year follow-up

Abstract

Background

Reverse total shoulder arthroplasty (rTSA) has been the subject of numerous studies reporting promising clinical outcomes which has led to its widespread use as the most common implant in shoulder arthroplasty today. However, few mid-term studies have reported outcomes of newer stem designs. This study evaluates the clinical and radiographic outcomes of a reverse shoulder prosthesis with an inlay stem design, through a multicenter, retrospective cohort analysis.

Methods

This retrospective study included 856 patients who underwent primary implantation of the isainlay, a reverse total shoulder prosthesis with an inlay stem design, between March 2019 and October 2024. Data were collected at pre-operative and post-operative time points (1 year, 3 years, and 5 years). The clinical outcomes were assessed using: Constant-Murley Score, American Shoulder and Elbow Surgeons score, Simple Shoulder Test, Subjective Shoulder Value, and visual analog scale for pain. Radiological assessments, including the presence of osteolysis around the humeral stems of the prostheses, were also performed at the key follow-up visits.

Results

At each follow-up, patients implanted with the isainlay stem design prosthesis demonstrated significant improvements in both pain and shoulder function (P < .0001) across all evaluated clinical scores. The primary outcome measure, the Constant-Murley Score, improved from 28 pre-operatively to 72 at 5 years. Radiographic analysis showed a low incidence of osteolysis, with no cases involving all stem zones.

Conclusion

Over the 5-year results, this reverse prosthesis design demonstrated favorable clinical and radiographic outcomes, comparable to those of other rTSA implants.

Since his introduction by Grammont at the end of the 20th century,¹¹ reverse shoulder arthroplasty has revolutionized the treatment of rotator cuff-deficient by medializing and lowering the center of rotation of the shoulder, which allows the restitution of active elevation.^2,11 Initially developed for cuff tear arthroplasty, this prosthesis has expanded to include complex proximal humeral fractures, failed conventional shoulder arthroplasties, and advanced glenohumeral arthropathies.^6,17 Numerous studies have confirmed favorable clinical outcomes^3,6,9,14 and recent national registry data reveal that reverse total shoulder arthroplasty (rTSA) is the most commonly used type of prosthesis compared with hemiarthroplasty or anatomic total shoulder arthroplasty.²

In 2019, a new reverse shoulder arthroplasty system featuring an inlay humeral stem design was introduced. Unlike the onlay design, the inlay-type prosthesis is buried below the proximal cutting level of humeral proximal epiphysis. This configuration allows a more anatomic positioning of the humerus relative to the glenoid, which may contribute to improved soft tissue tensioning. The excess thickness above the proximal humeral cut, which generates humeral lateralization, is thus reduced. To date, no mid-term study has assessed clinical and radiographic outcomes of the isainlay-specific design.

The purpose of this multicenter, nonrandomized, retrospective study is to evaluate clinical results of this stem design reverse prosthesis in a real-world setting, offering insights into its effectiveness and potential for widespread use in shoulder arthroplasty. This study aims to contribute to the growing body of evidence on reverse shoulder prostheses, particularly on the isainlay implants. We hypothesized that reverse shoulder arthroplasty using this inlay design would provide satisfactory clinical outcomes with complication rates in line with previously published data.

Methods

This study is multicenter, nonrandomized, open-label, and retrospective. Between March 2019 and October 2024, 856 isainlay reverse prostheses (Move-Up SAS, Alixan, France) featuring an inlay stem design were implanted as primary shoulder arthroplasties in 18 hospitals across France, including both academic and nonacademic institutions. All patients provided nonopposition to the use of their data, in accordance with the MR004 methodology. The study received institutional review board approval under reference number RnIPH 2025-131. Patients were followed as part of standard clinical practice, with data collected pre-operatively and at follow-up visits conducted at 1, 3, and 5 years.

Inclusion and exclusion criteria

Patients were eligible for inclusion if they underwent a primary rTSA using the isainlay stem between March 2019 and October 2024. Only patients with complete clinical data at baseline and at least one follow-up visit within the study period were included. Patients were excluded from the study if they had undergone previous shoulder arthroplasty or if the primary indication for arthroplasty was a fracture regardless of its type or location.

Collected variables and statistical analysis

Baseline demographic data included age, sex, body mass index, medical history, and pre-operative diagnosis. The primary outcome measure is the Constant-Murley Score (CMS), assessed pre-operatively and at all follow-up visits. Secondary outcome measures include the American Shoulder and Elbow Surgeons (ASES) score, Simple Shoulder Test (SST), Subjective Shoulder Value (SSV), and pain evaluated using a visual analog scale (VAS).

Radiographic evaluations were also performed at each follow-up. Anteroposterior and axillary views were reviewed to assess the presence of osteolysis around the humeral stem. All images were interpreted by the operating surgeons at their respective centers as part of routine clinical follow-up. Evaluations were performed using standardized criteria to ensure consistency in the identification of osteolytic regions. Osteolysis was defined as any new or progressive area of focal radiolucency exceeding 2 mm in width at the bone-implant or bone-cement interface, characterized by the disappearance of previously visible trabeculae. Five zones were defined, and the presence or absence of osteolysis in each zone was recorded (Fig. 1, A and B).

Figure 1.

Radiological zones for osteolysis assessment around the isainlay stem: (a) anteroposterior view and (b) axial view.

Statistical analysis aimed to compare post-operative clinical data at each timepoint with baseline scores. Given the sample size (n > 30), a normal distribution was assumed. Comparisons between timepoints were conducted using repeated measures analysis of variance. A P value less than < .05 was considered statistically significant.

Implant description

The reverse shoulder prosthesis evaluated in this study is the isainlay system (Move-Up, SAS, Alixan, France), a modular 4-component design composed of a humeral stem, humeral liner, glenosphere, and glenoid baseplate impacted by its press-fit peg and secured with peripheral screws (Fig. 2). The configuration assessed corresponds to an inlay design, in which the polyethylene liner is aligned with the humeral resection plane, thereby limiting lateral extension beyond the cortical margin and potentially facilitating more anatomic soft tissue tensioning. The monoblock humeral stem combined with the liner provides a neck-shaft angle of 145°, a design feature intended to reduce the risk of scapular notching by decreasing inferior impingement. Surface finishes differ depending on the fixation method: in cemented configurations, components are available in titanium-aluminum-vanadium alloy (Ti6Al4V), with an optional titanium nitride coating for patients with metal hypersensitivity, whereas the cementless humeral stem features a proximal porous titanium coating to promote osseointegration and metaphyseal biological fixation. All components are assembled using a Morse taper junction.

Figure 2.

Overview of the isainlay shoulder prosthesis and its 4 components: the isasphere glenosphere and isareverse glenoid baseplate; the isainlay humeral stem; the isaliner humeral insert.

Surgical technique

All patients underwent general or local anesthesia. Patients were positioned in a beach-chair setup, with the torso elevated to approximately 30° for the deltopectoral approach and 60° for the lateral approach. The upper limb was positioned freely within the surgical field, with careful attention to ensuring the shoulder extended beyond the edge of the table, allowing for retropulsion and adduction during the humeral stages of the procedure.

Surgeries were performed using either the deltopectoral or lateral approach. Tenotomy or peel-off was performed according to the surgeon's preference. After release of the subscapularis, osteophytes were removed and the anatomical neck was fully exposed. The humeral head was resected using a centromedullary guide (Fig. 3, A and B). A humeral cutting block was positioned and equipped with an “angel wing” to precisely control the level and orientation of the cut. Following humeral cut, glenoid exposure was achieved using retractors. After cartilage removal, the glenoid baseplate pin guide with 4 screws was placed (Fig. 4). Reaming was performed over the central guide pin, with the goal of preserving as much subchondral bone as possible. The baseplate was then implanted without cement (Fig. 5), using press-fit and the glenosphere was positioned. Finally, the definitive humeral implant was placed (Fig. 6). This surgical technique, particularly the deltopectoral approach combined with cementless baseplate fixation and anatomical humeral preparation, reflects current standard practice in shoulder arthroplasty and is widely adopted in routine clinical settings.¹²

Figure 3.

Resection of humeral head (b) with a centromedullary guide (a).

Figure 4.

Placement of the glenoid baseplate pin guide.

Figure 5.

Baseplate implantation using press-fit technique.

Figure 6.

Definitive humeral implantation.

Post-operative rehabilitation was at the surgeon's discretion, in line with what is achieved in routine monitoring.

Results

Baseline data

Among the 856 patients who received the reverse isainlay prosthesis in primary indication, there were 595 females and 261 males. The mean age was 74 ± 7 years and the mean body mass index was 28 kg/m². Given the retrospective nature of the study and the recent launch of the isainlay prosthesis, follow-up data were available for 779 patients at 1 year, 259 at 3 years, and 104 at 5 years (Fig. 7).

Figure 7.

Patient flow diagram.

rTSA was most commonly performed for eccentric omarthrosis, primary osteoarthritis, and massive rotator cuff tear; less frequent etiologies are summarized in Table I.

Table I.

Demographic and pre-operative characteristics in all patients

Baseline characteristics	Number of patients (n = 856)
Age at surgery, mean (SD)	74 (7)
BMI, mean (SD)	28 (6)
Gender
Female	595
Male	261
Operated shoulder
Right	536
Left	320
Indications
Eccentric omarthrosis	311
Primary osteoarthritis	273
Massive rotator cuff tear	240
Subluxating osteoarthritis	73
Fracture sequelae	28
Inflammatory arthritis	15
Post-traumatic necrosis	14
Post-traumatic arthrosis	13
Post-instability arthrosis	13
Atraumatic osteonecrosis	7
Chronic dislocation	6
Osteochondromatosis	6
Walch classification (n = 722)
A1	349
A2	140
B1	49
B2	118
B3	37
C	17
D	12

SD, standard deviation; BMI, body mass Index.

Pre-operative glenoid morphology was evaluated for 722 patients using the Walch classification.¹⁶ Most patients exhibited centered osteoarthritis (n = 489), classified as type A. Additionally, 204 patients exhibited posteriorly eccentric glenoids (type B), predominantly characterized by posterior erosion with a double glenoid line and posterior subluxation of the humeral head (n = 118), and to a lesser extent by narrowing of the posterior joint space, subchondral sclerosis, and osteophytes (n = 49). A smaller number of patients were classified as types C and D.

Clinical results

Due to variability in the availability of clinical scores at each follow-up point, the number of patients included in the analysis differs across scores and timepoints. The corresponding sample sizes (n) are reported in Table II. Clinical scores (CMS, ASES, SST, SSV, and VAS) were collected pre-operatively and at each of the 3 post-operative follow-up timepoints. At each post-operative visit, the mean values of all scores demonstrated significant improvement compared to baseline (P < .0001). At 1 year, clinical results for the CMS were available for 759 patients and for 257 patients at 3 years. At the 5-year mark, clinical results were available for 99 patients for CMS and for SST, 102 patients for ASES, 104 patients for SSV, and 103 patients for VAS. Therefore, at 5 years and compared to pre-operative data, the CMS improved by 44 points, the ASES by 61 points, the SST by 7 points, the SSV by 56 points, and the VAS has decreased by 7 points. Over the course of the 5 follow-up visits, 284 patients were lost to follow-up and 5 patients died. None of the reported deaths were related to prosthesis complications. More broadly, the significant improvement in clinical scores was consistent across all post-operative follow-up visits when compared to pre-operative values. Overall, the evolution of the scores followed a typical pattern: a marked and rapid improvement was observed immediately after surgery, followed by a slower but steady progression, eventually reaching a plateau. This trajectory suggests that most functional recovery and pain relief are achieved early, with long-term stability of the results.

Table II.

Mean values, standard deviations, and ranges (min-max) of PROMs (CMS, ASES, SST, SSV, and VAS) at each follow-up timepoint, with corresponding sample sizes (n) varying by score and timepoint

	Baseline			Year 1
PROMs	n	Mean (SD)	Min; Max	n	Mean (SD)	Min; Max
CMS	815	28 (12)	2; 76	759	71 (12)	15; 99
ASES	720	27 (13)	0; 72	630	87 (14)	17; 100
SST	714	3 (2)	0; 10	638	9 (2)	4; 12
SSV	854	29 (16)	0; 80	763	83 (14)	20; 100
VAS	856	7 (2)	0; 10	750	1 (1)	0; 8

	Year 3			Year 5
PROMs	n	Mean (SD)	Min; Max	n	Mean (SD)	Min; Max
CMS	257	73 (11)	15.5; 96	99	72 (13)	19; 89
ASES	227	88 (12)	30; 100	102	88 (13)	33; 100
SST	226	9 (2)	4; 12	99	10 (2)	5; 12
SSV	259	85 (13)	10; 100	104	85 (13)	25; 100
VAS	253	1 (1)	0; 7	103	0 (1)	0; 6

Min, minimum; Max, maximum; PROMs, patient-reported outcome measures; CMS, Constant-Murley Score; ASES, American Shoulder and Elbow Surgeons; SST, Simple Shoulder Test; SSV, Subjective Shoulder Value; VAS, visual analog scale; SD, standard deviation.

Radiologic assessment

The distribution of osteolysis by humeral zone, radiographic view (frontal and axillary), and follow-up timepoint is summarized in Table III. At 5 years, frontal radiographs revealed 24 instances of osteolysis confined to zones 1, 2, and 5 among the 95 images assessed. Similarly, axillary radiographs identified 12 osteolytic lesions within the same zone across 71 available radiographs. However, at each visit, none of the patients demonstrated osteolysis affecting all predefined humeral zones.

Table III.

Localization of osteolysis according to the 5 zones defined in Figure 1, assessed on 2 radiographic views (frontal and axillary views)

	Year 1		Year 3		Year 5
Humeral zones	Frontal (n = 697)	Axillary (n = 340)	Frontal (n = 240)	Axillary (n = 129)	Frontal (n = 95)	Axillary (n = 71)
Zone 1	36	20	17	4	6	4
Zone 2	7	6	5	1	4	1
Zone 3	3	0	0	0	0	0
Zone 4	1	2	0	1	0	0
Zone 5	39	13	30	10	14	7

Complications and revisions

Over the 5-year follow-up period, the isainlay reverse shoulder prosthesis demonstrated a total complication rate of 10.2% and a revision rate of 5.6%.

A total of 85 complications were encountered during the study period. Of these, 37 (43.5%) did not undergo revision surgery and were managed conservatively. Within this group, bone fractures were the most common occurrence (n = 13; 15.3%), followed by neurovascular injuries (n = 10; 11.8%) and dislocations (n = 9; 10.6%). Other complications included post-operative pain (n = 3; 3.5%), consisting of one case of pectoralis-related pain, one transient early post-operative pain episode with full resolution, and one acromioclavicular joint pain treated successfully with infiltration, soft tissue lesions (n = 3; 3.5%), and deep infection (n = 1; 1.2%).

The remaining 48 complications (56.5%) required surgical revision, with indications and corresponding case numbers detailed in Table IV. Among the patients revised for loosening, some underlying causes were traumatic falls (n = 3), metallosis (n = 2), excessive native glenoid wear (n = 1), osteoporotic bone failure secondary to rheumatoid arthritis (n = 1), and inadequate prosthesis position combined with intense activity (n = 1). Regarding patients revised for instability, the reported etiologies included a traumatic fall (n = 1), ligamentous laxity (n = 1), instability occurring during extreme internal rotation in the setting of complete loss of external rotation (n = 1), post-operative anterior shoulder hematoma preceding dislocation in an obese patient (n = 1), and tuberosity osteolysis potentially affecting prosthetic stability (n = 1). For the remaining cases of aseptic loosening (n = 7) and instability (n = 8), specific underlying etiologies were not documented. Furthermore, 2 patients underwent revision for persistent post-operative pain reported during follow-up and considered clinically significant by the treating surgeon. In one case, persistent pain was observed despite the absence of clear radiographic abnormalities. The initial surgery was performed via a deltopectoral approach with subscapularis tenotomy. Radiographic evaluation revealed increased humeral retroversion (>20°), without evidence of osteolysis or implant loosening. Although no definitive structural cause was identified, component malposition may have contributed to the symptoms, leading to revision. The second patient underwent revision due to subacromial impingement between the greater tuberosity and the acromion. Finally, one patient required revision due to implant fracture involving a glenoid baseplate locking screw. The index surgery included a glenoid-sided bone graft that failed to incorporate, resulting in nonunion. The implant remained in situ for just over 3 years before screw breakage occurred, leading to pain. Revision surgery involved bone reconstruction. At the latest follow-up, the patient was asymptomatic with restored mobility, including active overhead elevation.

Table IV.

Indications for revision surgery and number of cases

Indication for revision	Number of cases (n)	Percentage among the revisions (%)	Percentage among the cohort (%)
Aseptic loosening	15	31.3	1.8
Dislocation/instability	13	27	1.5
Fracture following trauma	9	18.8	1.1
Infection	8	16.7	0.9
Pain	2	4.2	0.2
Implant breakage	1	2.1	0.1

Among the documented and analyzed cases, no clear implant-related mechanical failure was identified. The indications for revision were predominantly related to complications commonly reported in shoulder arthroplasty, which are likely multifactorial in origin.

Discussion

The main finding of this study is that the isainlay reverse shoulder prosthesis provided significant and sustained improvement in shoulder function and pain relief over time, with a low rate of complications and revisions at mid-term follow-up. Across all clinical scores, marked gains were observed as early as the 1-year follow-up, and these improvements were maintained up to 5 years. These findings suggest that the isainlay stem configuration achieves reliable clinical performance in primary rTSA, comparable to or better than that reported for other contemporary reverse shoulder systems.

Comparable outcomes have been reported by Linke et al,¹³ who evaluated 99 patients after rTSA with a mean follow-up of 4 years. The average ASES score is increasing from 38 to 86. While final scores were similar, our cohort demonstrated a more marked functional improvement, evidenced by a greater ASES gain (61 vs. 48) and a slightly higher final score (88). Similarly, our mean SSV of 85 at 5 years aligns closely with the 83 reported in their series.

Findings from Schnetzke et al¹⁵ support these observations. In a cohort of 24 patients, they reported an improvement in the CMS from 22 pre-operatively to 57 at 2-year follow-up. Our results, with a mean score rising from 28 pre-operatively to 72 at 5 years, appear highly favorable, both in terms of improvement magnitude and long-term durability.

Regarding pain assessment, our results are consistent with previously published data. Goldenberg et al reported a mean VAS pain score of 2.2 at 56 months in a cohort of patients aged less than 65 years, while Barret et al observed a mean score of 0.25 at 44 months in patients aged more than 70 years.^8,10 In comparison, the mean VAS pain score in our cohort was 0 at 60 months, suggesting satisfactory pain relief within the range of values reported in the literature.

The Minimal Clinically Important Difference (MCID) is widely used to ensure that observed improvements are genuinely perceived as beneficial by patients. A systematic review by Yendluri et al¹⁸ reported the most commonly used MCID values for 5 outcome scores, including the ASES, SST, and CMS. The widely accepted MCID thresholds are 15.4 points for the ASES, 1.8 points for the SST, and 10.3 points for the CMS. Based on these references, all 3 scores used in our study exceed their respective MCID thresholds, indicating that the improvements observed after arthroplasty are clinically meaningful from the patient's perspective.

Regarding complications, our study reports a reoperation rate of 5.6% and an overall complication rate of 10.2%, reflecting satisfactory medium-term implant tolerance. These rates are consistent with, and favorably compare to, relevant literature.^1,4,7 In particular, a systematic review by Chelli et al analyzing outcomes in active patients reported, at mid-term follow-up, a revision rate of 7%.⁵

These data confirm the safety of this procedure in the early post-operative period, with a low risk of immediate complications. Although our data cover a medium-term follow-up, the observed rates remain comparable, highlighting the stability and reliability of the isainlay rTSA.

This study has several methodological characteristics that support the reliability and external validity of its findings. The large cohort size allows for adequate statistical power and reduces the influence of outliers, providing a more representative picture of the population undergoing primary rTSA. The multicenter design, involving 18 hospitals across different regions and healthcare settings, also contributes to the robustness of the data. By including both academic and nonacademic institutions with varying surgical practices, the study reflects real-world clinical diversity and limits the bias associated with single-center investigations. This design may therefore offer a more accurate estimate of the functional outcomes achievable with the isainlay reverse prosthesis in routine practice, although the 33% loss to follow-up should be taken into account when interpreting the generalizability of these findings.

Nevertheless, this study has several limitations that should be acknowledged. First, its retrospective design inherently limits the analysis to data available in medical records, which may be incomplete or inconsistently reported across centers. Second, as this study evaluates a single inlay implant design, the findings cannot be generalized to all inlay reverse shoulder arthroplasty systems given potential variations in implant geometry, modularity, and biomechanical behavior between manufacturers. In addition, the multicenter nature of the study introduces heterogeneity in surgical techniques, which may act as a confounding factor. Different surgical approaches were used according to surgeon preference, with associated variations in patient positioning and exposure. These differences may have influenced implant positioning and the technical execution of the procedure, particularly with respect to humeral preparation in an inlay configuration. Post-operative rehabilitation protocols were not standardized and were left to the discretion of the treating surgeons, potentially contributing to variability in functional outcomes across centers. Finally, not all patients completed all follow-up visits, with an overall loss to follow-up of 33%, which may introduce selection bias and affect the robustness of both clinical and complication outcomes. Then, one of the authors is employed by the manufacturer of the implant evaluated in this study, which may represent a potential source of bias. Although no external funding was received, surgeons were compensated for their contribution to clinical follow-up in accordance with applicable regulations. This relationship should be considered when interpreting the results. Longer-term data are needed to confirm the durability of these results and to assess late complications or implant survival beyond the current follow-up period.

Conclusion

This multicenter study confirms the good clinical and radiological outcomes of the isainlay reverse shoulder prosthesis system used as a primary treatment. The significant and sustained improvements in clinical scores (CMS, ASES, SST, SSV, and VAS) demonstrate the functional effectiveness of this prosthesis in managing shoulder pathologies, particularly eccentric omarthrosis, primary osteoarthritis, and massive rotator cuff tear. This reverse arthroplasty system demonstrates reliable functional outcomes and a satisfactory safety profile. Continued follow-up to 10 years will be essential to confirm the long-term durability of these results.

Disclaimers:

Funding: No funding was disclosed by the authors.

Conflicts of interest: Flore A. Bordes reports being employed by Move-Up, the manufacturer of the device evaluated in this study.

Any additional authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.