Scholarly article on topic 'Metal-backed glenoid implant with polyethylene insert is not a viable long-term therapeutic option'

Metal-backed glenoid implant with polyethylene insert is not a viable long-term therapeutic option Academic research paper on "Clinical medicine"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"Total shoulder arthroplasty" / "uncemented glenoid" / "glenoid survival" / revision}

Abstract of research paper on Clinical medicine, author of scientific article — Pascal Boileau, Grégory Moineau, Nicolas Morin-Salvo, Cyrille Avidor, Arnaud Godenèche, et al.

Background The aims of this study were to determine the survival of anatomic total shoulder arthroplasty with uncemented metal-backed (MB) glenoid components with a polyethylene (PE) insert in primary osteoarthritis, to assess the reasons for revision surgery, and to identify patients and diagnostic factors that influence failure rates. Methods Between 1994 and 1999, 165 patients (mean age, 68 years) with primary osteoarthritis were treated with anatomic total shoulder arthroplasty using an uncemented MB/PE glenoid component. Outcomes were assessed both clinically and radiologically with a minimum of 2 years of follow-up. A prosthetic survival curve was constructed with the end point defined as either partial or complete revision, using 100% confidence intervals. Results Survival rate free of revision was 46% (100% confidence interval, 32%-54%) at 12 years. At a mean follow-up of 8.5 years (range, 2-16 years), revision was required in 61 patients (37%); 80% of shoulders undergoing revision (49 of 61) had evidence of PE wear. Glenoid loosening (because of osteolysis secondary to wear debris), soft tissue deficiency, and prosthetic instability were the most common modes of failure. Younger patients and biconcave glenoids (with posterior humeral subluxation) have a negative effect on implant survival. Proximal humerus osteolysis was significantly more frequent in shoulders with PE wear. Exchange of the PE insert (with conservation of the MB tray) was possible in only 3% of the revised shoulders. Conclusion Uncemented MB glenoid resurfacing is not a viable long-term therapeutic option because of accelerated PE wear leading to early revision surgery. Conservation of the MB tray with reinsertion of a new PE insert is rarely possible because of glenoid bone loss, implant loosening, soft tissue deficiency, and prosthetic instability. Younger patients and biconcave glenoids have a negative effect on implant survival.

Academic research paper on topic "Metal-backed glenoid implant with polyethylene insert is not a viable long-term therapeutic option"

J Shoulder Elbow Surg (2015) ■, 1-10

Journal of Shoulder and Elbow Surgery

www.elsevier.com/locate/ymse

Metal-backed glenoid implant with polyethylene insert is not a viable long-term therapeutic option

Pascal Boileau, MDa*, Grégory Moineau, MDa, Nicolas Morin-Salvo, MDa, Cyrille Avidor, MDa, Arnaud Godeneche, MDb, Christophe Lévigne, MDc, Mohamed Baba, MDa, Gilles Walch, MDb

aHopital de l'Archet 2, Université Nice Sophia Antipolis, Nice, France bCentre Medical Santy, Lyon, France cClinique du Parc, Lyon, France

Background: The aims of this study were to determine the survival of anatomic total shoulder arthroplasty with uncemented metal-backed (MB) glenoid components with a polyethylene (PE) insert in primary oste-oarthritis, to assess the reasons for revision surgery, and to identify patients and diagnostic factors that influence failure rates.

Methods: Between 1994 and 1999, 165 patients (mean age, 68 years) with primary osteoarthritis were treated with anatomic total shoulder arthroplasty using an uncemented MB/PE glenoid component. Outcomes were assessed both clinically and radiologically with a minimum of 2 years of follow-up. A prosthetic survival curve was constructed with the end point defined as either partial or complete revision, using 100% confidence intervals.

Results: Survival rate free of revision was 46% (100% confidence interval, 32%-54%) at 12 years. At a mean follow-up of 8.5 years (range, 2-16 years), revision was required in 61 patients (37%); 80% of shoulders undergoing revision (49 of 61) had evidence of PE wear. Glenoid loosening (because of osteolysis secondary to wear debris), soft tissue deficiency, and prosthetic instability were the most common modes of failure. Younger patients and biconcave glenoids (with posterior humeral subluxation) have a negative effect on implant survival. Proximal humerus osteolysis was significantly more frequent in shoulders with PE wear. Exchange of the PE insert (with conservation of the MB tray) was possible in only 3% of the revised shoulders.

Conclusion: Uncemented MB glenoid resurfacing is not a viable long-term therapeutic option because of accelerated PE wear leading to early revision surgery. Conservation of the MB tray with reinsertion of a new PE insert is rarely possible because of glenoid bone loss, implant loosening, soft tissue deficiency, and prosthetic instability. Younger patients and biconcave glenoids have a negative effect on implant survival.

Institutional Review Board approval was not required for this study. It is a *Reprint requests: Pascal Boileau, MD, Hopital de l'Archet 2, Uni-

retrospective study of patients whose surgery followed validated tech- versite Nice Sophia Antipolis, 151, route de St Antoine de Ginestiere,

niques. No unnecessary invasive examination was performed. All patients F-06200 Nice, France.

consented to participate in this review and agreed that their anonymized E-mail address: boileau.p@chu-nice.fr (P. Boileau).

data could be used for scientific purposes.

1058-2746/© 2015 The Authors. Published by Elsevier Inc. on behalf of Journal of Shoulder and Elbow Surgery Board of Trustees. This is an open access

article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

http://dx.doi.org/10.1016/j.jse.2015.02.012

ARTICLE IN PRESS

2 P. Boileau et al.

Level of evidence: Level IV, Case Series, Treatment Study.

© 2015 The Authors. Published by Elsevier Inc. on behalf of Journal of Shoulder and Elbow Surgery Board of Trustees. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).

Keywords: Total shoulder arthroplasty; uncemented glenoid; glenoid survival; revision

Total shoulder arthroplasty (TSA) has proved to be a beneficial treatment for patients with primary glenohumeral osteoarthritis (OA) presenting with shoulder stiffness and pain. However, long-term fixation of the glenoid component remains an unsolved problem, and loosening of cemented polyethylene (PE) glenoid components represents an important cause of failure in TSA.17,22,24 The rate of glenoid lucent lines after cemented all-PE implants in anatomic TSA has been reported to be up to 90%. To improve glenoid fixation and to reduce glenoid lucent lines, uncemented fixation with porous coated or tissue-ingrowth components has been developed with the aim of achieving more stable fixation to the bone and a corresponding in-

1 7-9 15 17 23

crease in implant survival. Despite the body of

proof indicating that uncemented glenoid implants are at risk for failure and revision, uncemented metal-backed (MB) glenoid components are still commonly used in

TSA 1,5-9,13-15,19,20,23,25-27

Moreover, with the recent success of reverse prostheses, there is currently renewed interest in the development of "universal" uncemented glenoid MB implants. These implants could be used for both anatomic (with a PE insert) and reverse (with a metallic sphere) shoulder arthroplasty. The MB glenoid tray would allow the surgeon to more easily convert failed TSA into reverse shoulder arthroplasty (RSA) without having to revise the glenoid component. In theory, exchange of the PE insert for a metallic sphere (with conservation of the MB tray) would make the revision procedure easier and faster.

In 2015, at least 2 questions remain to be answered: (1) Is an uncemented MB glenoid component with a PE insert an acceptable option for TSA in the long term? (2) Does the uncemented MB glenoid allow an easier revision surgery in case of failure with conservation of the MB tray and simple exchange of the worn PE? To try to answer these questions, we conducted a continuous retrospective multicenter cohort study of patients, with a single etiology (primary OA), treated with the same unconstrained anatomic TSA with uncemented bone-ingrowth MB glenoid components. The aims of this study were to determine the survival rate of this type of implant, to assess the reasons for revision surgery, and to identify patients and diagnostic factors that influence failure rates. We hypothesized that (1) glenoid resurfacing with an MB implant with PE insert would not be acceptable in the long term because of accelerated PE wear and glenoid loosening and (2) the MB tray could not be conserved in case of revision because of glenoid bone loss and implant loosening.

Materials and methods Study design

The inclusion criteria were as follows: (1) TSA for the treatment of primary OA implanted by a senior shoulder surgeon; (2) gle-noid resurfacing with use of the same uncemented MB glenoid component; and (3) minimum clinical and radiologic follow-up of 2 years. The exclusion criteria included the following: (1) TSA for all others causes; (2) glenoid resurfacing with use of a cemented full-PE glenoid component; and (3) follow-up of <2 years. Between 1994 and 1999, 178 TSAs with the same uncemented MB glenoid component were implanted for the treatment of primary OA in 5 orthopedic centers; 13 patients died or were lost to follow-up, leaving 165 TSAs in 158 patients (115 women and 43 men). At the time of shoulder arthroplasty, the mean age of the patients was 68 years (range, 35-89 years). Seven patients had surgery on both sides. The dominant arm was operated on in 105 cases (64%). The presence of primary OA was confirmed in all patients on true anteroposterior (AP), axillary, and lateral radiographs. Preoperative computed tomography (CT) arthrography was performed in 150 patients (90%) and preoperative magnetic resonance imaging in 5 patients (3%) to evaluate glenoid bone stock, presence of a rotator cuff tear, and fatty infiltration of the rotator cuff muscles. According to the Walch classification,29 the glenoid morphology was identified as type A in 72 cases (44%), type B in 75 cases (45%), and type C in 8 cases (5%). According to the Goutallier classification,16 the global fatty degeneration index evaluated on CT arthrography was <1 for 45 shoulders (27%) and >1 in 105 patients (64%).

Operative technique and implants

Surgeons used the same surgical technique and the same implant in the different centers. A deltopectoral approach was used in all patients. The subscapularis tendon and anterior capsule were divided at the medial edge of the lesser tuberosity in 154 cases (93%) (i.e., simple tenotomy); the tendon and capsule were detached with a fleck osteotomy of the lesser tuberosity in 11 cases (7%). The long head of the biceps was tenodesed in 86 cases (52%). There were 15 partial (9%) and 11 full-thickness (7%) tears of the supraspinatus. The humeral implant was cemented in all except 2 cases. The same anatomic, unconstrained, modular, and adaptable humeral component was used in all patients (Aequalis shoulder prosthesis; Tornier Inc., Houston, TX, USA).2'3 The same uncemented, bone-ingrowth MB glenoid implant was used in all cases (Aequalis MB glenoid prosthesis; Tornier). The thickness of the glenoid component was 7 mm (3 mm for the metal tray and 4 mm for the PE insert). Two expansion screws (with 3 petals for each) achieved the initial

ARTICLE IN PRESS

MB glenoid is not viable long term 3

4 nil !,r zone RLL Thickness points

< 1 mm 1-2 mm >2 mm

l X 3

2 X 3

3 X 3

4 X 3

5 X 3

total 15/15

Figure 1 Analysis of radiolucent lines (RLLs). Humeral radiolucent lines were analyzed radiographically in 8 zones (A). Glenoid RLLs were quantified in 5 zones, taking into account their thickness (B). The maximum glenoid RLL score was 15 points. A glenoid component with a score >10 points was considered to be loose. (Mole classification)22

stability of the MB component. The MB component was hy-droxyapatite coated and had a porous surface to allow bone ingrowth. Cancellous bone grafting (autologous bone graft harvested from the cut humeral head) was performed under the tray in 20 cases of type A glenoid, and a structural bone graft (iliac crest bone graft) was performed in 2 cases of type B2 glenoid.

Outcomes assessment

Patients were observed at 6 weeks, 6 months, 12 months, and then yearly. At each visit, patients were evaluated clinically and radio-logically. The minimum follow-up was 2 years, with a mean of 8.5 years (range, 2-16 years). Preoperative and postoperative assessment included the measurement of active range of motion (anterior elevation, external rotation, internal rotation), the Constant score, and the normalized Constant score calculated as a percentage of the normal reference value adjusted for age and gender.11 At the last follow-up, the subjective shoulder value was evaluated.

Radiographs were assessed for radiolucent lines (RLLs), and a glenoid radiolucency score, based on the method described by Mole et al,22 was calculated for each patient. A numeric value was applied to each of the 5 zones surrounding the glenoid implant according to the thickness of the RLL (Fig. 1). RLLs were scored on a scale of 0 to 15. A score >10 points indicates definite radiologic loosening of the MB implant. Any migration or shift of the component or any metallic breakage received by convention a score of 15 points, even if no RLLs were present, and was considered definite loosening. Furthermore, wear of the PE insert was assessed in the AP radiographs and on CT scans. A narrowing of the joint line (between glenoid and humeral components) was considered wear. Humeral radiolucency around the humeral component was evaluated in 8 zones (Fig. 1). A postoperative CT scan was performed in 76 patients, mainly in case of complications or before revision surgery.

Four independent observers not involved in the surgical procedure (G.M., N.M.-S., C.A., A.G.) performed all radiographic and CT scan analyses. The postoperative radiographic assessment at last follow-up or before revision surgery included the following: glenoid RLLs to calculate the RLL score22; glenoid component migration or tilt (superior, inferior, medialization, anterior, or posterior); petal or tray fractures; PE wear as no, partial, or complete (metal-metal contact); dislocation or subluxation

Figure 2 Example of complete PE wear with metal-metal contact and glenoid loosening. Note the superior humeral subluxation (indicating cuff insufficiency), the glenoid bone loss with tilt of the glenoid implant, and the osteolysis of the medial calcar of the humerus (PE and metal granuloma).

(anterior or posterior); superior humeral migration; and humeral lucent lines in 8 zones.

Definitions

Glenoid loosening (Fig. 2) was considered to be present if there was either glenoid component migration or shift, a petal or tray fracture, or a glenoid RLL score >10 points.2

PE wear was considered to be present if the thickness between the humeral head prosthesis and the tray of the glenoid implant

Figure 3 Example of accelerated wear of the PE insert without glenoid loosening. (A) Immediate postoperative AP radiograph demonstrates thick space between prosthetic head and MB glenoid. (B) Radiograph taken 5 years postoperatively shows disappearance of the joint line with metal-metal contact. (C) Retrieved glenoid implant demonstrates complete (central) wear of the PE, leading to metal-metal contact. Conservation of the MB tray with reinsertion of a new PE insert was not possible.

decreased between the postoperative and latest imaging on AP or axillary views. PE wear was said to be complete if metal-metal contact was observed on either radiographs or CT scans (Fig. 3) and partial if a space between the humeral head prosthesis and tray component was still visible.

PE dissociation was diagnosed on radiographs, on CT scan images, or intraoperatively during revision or reoperation procedures (Fig. 3).

A subscapularis tear or insufficiency was diagnosed on clinical examination at review in cases of a positive lift-off test result or intraoperatively during revision or reoperation procedures.

A supraspinatus tear or insufficiency was diagnosed on the AP radiographic view if there was superior migration of the humeral head or intraoperatively during revision or reoperation procedures.

Anterior or posterior instability was diagnosed on CT scans if there was subluxation or dislocation of the humeral component.

A revision was defined as surgery with partial or total exchange or removal of the prosthesis.

A reoperation was defined as surgery for a soft tissue problem without exchange of the prosthesis.

Statistical analysis

The distribution of data was analyzed with the D'Agostino-Pearson test. The preoperative and postoperative data were analyzed for statistically significant differences between means; paired data were compared by a paired t test, and unpaired results were compared with the Mann-Whitney U test. The c2 test and Fisher exact test for small numbers were used to compare categorical data. The significance level was set at P < .05. Kaplan-Meier analysis was used to estimate implant survival and the removal or revision of the glenoid component for any reason. A prosthetic survival curve was constructed with the end point defined as revision of TSA and using 100% confidence intervals (CIs; maximal CI 100% if all patients deceased or lost to follow-up were considered as not revised, and minimal CI 100% if all patients deceased or lost to follow-up were considered as revised).

Figure 4 Survival curve of TSA with uncemented MB implant in primary OA. The end point was partial or total revision of the shoulder prosthesis component. Max CI 100% if all patients deceased or lost to follow-up were considered as not revised. Min CI 100% if all patients deceased or lost to follow-up were considered as revised.

Results

Implant survival

The rate of prosthetic survival (end point = revision) was 60% (100% CI, 44%-71%) and 46% (100% CI, 32%-54%) at 10 and 12 years, respectively (Fig. 4). There is a severe drop of the curve after the fourth year.

Complications

The complications that were often associated are summarized in Table I. The most frequent complication was, by far, wear of the PE insert observed in 85 cases (51%) either on follow-up radiographs or during revision surgery (Figs. 3, 5, and 6). This complication was rarely found in isolation but rather in association with glenoid

component loosening, soft tissue problems, and prosthetic instability (Fig. 7).

Revision surgery

At a mean follow-up of 8.5 years (range, 2-16 years), 66 patients required further surgery (40%). A revision was required in 61 patients (37%), and reoperation was required in 5 patients (3%). Of the 61 shoulders, 49 (80%) undergoing revision had evidence of PE wear. Although the reasons to revise the TSA were multiple, the primary reasons are listed in Table II.

The types of revision procedure are summarized in Table III. The anatomic TSAs were converted to RSA in case of glenoid bone loss, soft tissue deficiency, or prosthetic instability that precluded reinsertion of the anatomic TSA. Complete exchange of TSA was performed in case of difficulty in exposing the glenoid (despite removal of the prosthetic head) or proximal humeral bone loss. In all other cases, the humeral component was kept in place and the MB glenoid component was removed. In case of severe glenoid bone loss, an iliac crest bone graft was performed, whereas in elderly and fragile patients, the glenoid component was simply removed. Finally, removal of the worn PE insert (with conservation of the MB tray) with reinsertion of a new PE insert was possible only in 2 cases.

Glenoid implant failures

The 57 revised glenoid implant failures have been studied in detail (Table IV). As already mentioned, glenoid implant failures were often associated with other complications. Thirty-three patients with complete wear of the PE insert needed further surgery. In these patients, PE wear was seen without glenoid loosening. The mean interval between primary surgery and revision for PE wear was 78 months (range, 6-152 months). Twenty-two patients had complete loosening

of the glenoid component and needed further surgery. In 16 patients, glenoid loosening was seen after 2 years and was associated with severe PE wear and glenoid bone loss. Six patients had expansion screw problems (fractured or not open) on immediate postoperative radiographs with absence of primary fixation and rapid glenoid loosening (in the first 2 years). The mean interval between primary surgery and revision for glenoid loosening was 86 months (17-141).

Prognostic factors

Two prognostic factors were found to be associated with the occurrence of complications or revision surgery: young age and biconcave glenoid with posterior humeral subluxation.

Patients with complications were younger at the time of surgery (66.7 years [35-83] vs. 70.9 years [43-89]) than patients without complications (P = .03). Revised patients were younger (65.3 years [35-80] vs. 69.7 years [43-89]) than those not revised (P = .005).

Recurrent posterior subluxation at last follow-up was significantly more frequent in the type B glenoid subgroup (16 of 75) than in the type A subgroup (4 of 72) (P = .007). This was associated with asymmetric posterior wear of the PE insert (Fig. 5). Revision arthroplasty was performed earlier in the type B glenoid subgroup than in type A (P = .05) (Table V).

RLLs on the humeral side were diagnosed in 65 cases (39%) and were more frequent in zones 1, 7, and 8 (Fig. 1). Osteolysis of the proximal humerus and RLLs in zones 7 and 8 were significantly more frequent in shoulders with glenoid loosening and complete PE wear (P < .05).

No correlation was found between the occurrence of complications or revision and the following parameters: gender, dominant or nondominant side, supraspinatus tears, rotator cuff muscle fatty infiltration, postoperative glenoid or humeral component positioning, and humeral head size.

Clinical outcomes

The functional results are summarized in Table VI. They were evaluated in the 104 shoulders that still retained their implant at the last follow-up. The mean follow-up was 102 months (range, 24-191 months). There were statistically significant improvements in all measures of active range of motion and in all parameters of the Constant score. The mean postoperative subjective shoulder value was 78% (range, 40%-100%). Of the patients who had their TSA in place, 76% (79 of 104) were still satisfied or very satisfied with their result.

Discussion

This study confirms our 2 hypotheses: uncemented MB glenoid resurfacing with a PE insert is not a viable

Table I Postoperative complications

Postoperative complication * No. (%)

Wear of PE insert 85 (51)

Dissociation of PE insert 2(4)

Glenoid loosening 32 (19)

Soft tissue insufficiency 36 (21)

Subscapularis rupture 18 (11)

Supraspinatus tear 18 (11)

Prosthetic instability 31 (19)

Posterior 25 (15)

Anterior 6(4)

Deep infection 2 (1)

Shoulder stiffness 2 (1)

Postoperative humeral fracture 1 (<1)

* Patients could have more than one associated complication. Not all complications have been revised.

Figure 5 Example of recurrence of posterior humeral subluxation with asymmetric (posterior) wear of PE insert. (A) Preoperative CT scan shows biconcave (B2) glenoid with severe posterior subluxation of the humerus. (B) Postoperative CT scan shows recurrence of posterior humeral subluxation (despite the correct orientation of the glenoid implant) and asymmetric (posterior) wear of PE insert and metal tray. Wear and flattening of the prosthetic head is also visible.

Figure 6 Accelerated wear of the PE without glenoid loosening. At 12 years of follow-up, there is complete wear of the PE insert (with metal-metal contact), but the MB glenoid implant is still well fixed in the bone.

Figure 7 Associated soft tissue complications with accelerated wear or dissociation of the PE. (A) Subscapularis rupture after traumatic anterior dislocation. (B) Supraspinatus tear with progressive upward migration and decreased acromiohumeral space.

long-term therapeutic option, and conservation of the MB tray with reinsertion of a new PE insert is rarely possible at the time of the revision. Our long-term study confirms our previous findings and demonstrates that (1) MB glenoid implants with a PE insert have a low and unacceptable rate

of survival; (2) PE wear is the main complication found in 80% of the revised shoulders; (3) accelerated wear of the PE insert is associated with glenoid loosening (because of wear products and osteolysis), prosthetic instability, and soft tissue deficiency; (4) younger patients and those with

Table II Primary reasons for revision surgery after TSA

Primary reason for revision surgery No. of cases (%)

Complete PE wear 33 (20)

(with metal-metal contact)

Glenoid loosening (16/22 with PE wear) 22 (13)

Prosthetic instability 7(4)

Recurrent posterior humeral 5

subluxation with PE wear

(all with PE wear)

Anterior dislocation 2

PE insert dissociation 2(1)

Subscapularis tear 1 (0.6)

Infection 1 (0.6)

Total 66 (40)

Table III Type of revision surgery procedures

Type of revision surgery No. of cases/165 (%)

Conversion of TSA to RSA 18 (11)

Exchange TSA for a new TSA 15 (7)

(with cemented PE glenoid)

Glenoid removal and PE glenoid 15 (7)

reimplantation

Glenoid removal and iliac crest 7 (4)

bone graft

Glenoid removal alone 4 (2)

New PE inserted on MB 2 (2)

Total 61 (37)

Table IV Revision of failed glenoid implant and associated

complications found

Glenoid implant No. Associated complications*

failures (revised)

Complete PE wear 33 Instability (12)

(without glenoid o Posterior (10)

loosening) o Anterior (2)

Cuff tears (8)

Subscapularis rupture (7)

PE insert dissociation (4)

Deep infection (1)

Humeral loosening (1)

Isolated PE insert 2

dissociation

MB glenoid loosening 22 PE wear (16)

(with or without Instability (5)

PE wear) o Posterior (4)

o Anterior (1)

Rotator cuff tear (2)

Subscapularis rupture (2)

Stiffness (1)

* Patients could have more than one associated complication.

preoperative posterior humeral subluxation and biconcave glenoid (type B glenoid) have an increased risk of failure and a higher risk of revision; (5) osteolysis of the proximal humerus is significantly more frequent in shoulders with PE wear and glenoid loosening; and (6) exchange of the PE insert with conservation of the MB tray is possible in only 3% of the revised shoulders. The high rate of complications and low survival rate of MB implants with a PE insert found in this study have led us to definitely abandon the use of this type of implant for anatomic TSA. The fact that accelerated wear of the PE insert is associated with severe glenoid loosening (because of wear products and osteol-ysis) and exchange of implants calls into question the development and use of universal uncemented glenoid MB implants.

The main limitations of our study are inherent to a retrospective cohort study without any control group. However, our therapeutic study has several notable strengths. First, the number of cases is relatively large, and all patients have been reviewed for clinical and radiologic examination with a minimum follow-up of 2 years. Second, the follow-up is long enough to evaluate long-term clinical and radiologic outcomes (mean, 8.5 years; range, 216 years). Third, the group of patients is homogeneous with a single diagnosis (primary OA), the same surgical

technique, and a single implant used. Fourth, although our study is retrospective, the patients have been observed prospectively, and the dropout rate was extremely low: only 8% of patients lost to follow-up in the study period.

There is a paucity of literature that addresses the results of uncemented glenoid MB components in TSA.2,7,9,19,20,25,27 Most reports are of small series with

5 6 18 19 25 27 31

short-term or midterm follow-up ^^^ (see online supplementary Table; available on the journal's website at www.jshoulderelbow.org). Our series, which is large and with long follow-up, reinforces the findings of local and national shoulder joint registries.8,13 Fox et al13 recently reported the results of the Mayo Clinic shoulder joint registry. They found that the survival rate free of revision for 254 Cofield MB glenoid implants (Smith & Nephew, Memphis, TN, USA) was 67% at 15 years. They found, like us, that glenoid loosening and failures were largely related to wear products and osteolysis. Similarly, data retrieved from the New Zealand National Joint Registry have shown that even with a short follow-up (mean follow-up of 3.5 years), the revision rate for uncemented glenoids (SMR, Lima LTO, Udine, Italy) was 3 times that of cemented glenoid implants.8

Our study highlights the fact that accelerated wear of the PE insert is the main limiting factor for the use of MB implants in anatomic TSA, leading to glenoid loosening. Aseptic glenoid loosening is the result of macrophage and giant cell osteolytic activity induced by PE and metallic wear debris.4,7-9 Factors associated with accelerated wear of the plastic insert in MB glenoid are well known today1,2: (1) insufficient thickness of the PE insert (4 mm with our

Table V Influence of glenoid type on complication, prosthetic posterior instability, and revision rates

Glenoid type A n = 72 Glenoid type B n = 75 P value (test)

Complications 42 (58%) 54 (72%) .15 (c2)

Revisions 22 (30%) 33 (44%) .13 (c2)

Prosthetic posterior instability 4 (5.5%) 16 (21%) .007 (FET)

Delay of revision (months) 92 (31-141) 73 (17-152) .05 (t test)

FET, Fisher exact test.

Table VI Functional results in the 104 patients with TSA still in place at last follow-up (102 months [24-191])

Preoperative Postoperative P value (test)

Active mobility Active anterior elevation (°) Active external rotation (°) Constant score Pain/15 points Activities/20 points Range of motion/40 points Strength/25 points Absolute Constant score/100 points Adjusted Constant score (%) 88 ± 26 (30 to 170) 5 ± 15 (-30 to 50) 4.2 ± 1.8 (0 to 8) 7.2 ± 2.7 (0 to 14) 13.6 ± 6.5 (0 to 28) 2.7 ± 3.1 (0 to 14) 27.6 ± 10.2 (6 to 52) 38 ± 24 (9 to 73) 140 ± 27 (70 to 180) 35 ± 19 (-10 to 70) 12.5 ± 3.1 (4 to 15) 15.9 ± 4.5 (3 to 20) 29.4 ± 9 (4 to 40) 7.1 ± 4.8 (0 to 25) 64.9 ± 17.8 (17 to 96) 94 ± 25.8 (20 to 136) <.0001 (MWT) <.0001 (MWT) <.0001 (MWT) <.0001 (MWT) <.0001 (MWT) <.0001 (MWT) <.0001 (MWT) <.0001 (MWT)

MWT, Mann-Whitney U test.

implant); (2) overtensioning of the soft tissues and increased load on the PE insert because of excessive global thickness of the glenoid component (7 mm); (3) increased rigidity resulting from the MB tray, which is responsible for increased stresses in the PE insert; and (4) recurrent posterior instability or soft tissue deficiency that results in eccentric loading and asymmetric wear of the PE surface.

Only 6 patients had early glenoid loosening (occurring in the first 2 years) because of expansion screw problems (fractured or not open). In all other patients, glenoid loosening occurred (after 2 years) and was associated with PE wear and severe glenoid bone osteolysis. These results suggest that the weak link in the chain is not the MB fixation to the bone but rather the rapid and unexpected wear of the PE insert leading to glenoid bone osteolysis. Complete wear of the PE insert generates plastic and metal wear debris that has at least 3 consequences: (1) biologic: wear debris is responsible for a granuloma and severe glenoid bone loss; (2) mechanical: bone loss leads to glenoid component shift and loosening; and finally (3) clinical: mobility of the glenoid implant (and synovitis) is responsible for pain and progressive loss of motion, whereas in some patients we observed that metal-metal contact was responsible for "noisy" shoulders.

In accordance with previous studies,8,13 our data show that younger age has a negative effect on implant survival. This is not surprising and has been shown in hip and knee but also in shoulder prostheses: younger patients have a higher level of

activity, which increases the risk of accelerated PE wear and aseptic glenoid loosening.10,26 Thus, contrary to common belief, the uncemented MB component with PE insert offers no advantage for improved survival in young patients who have longer life expectancy. Furthermore, the hope that the MB tray could be left in place and only the worn PE would be exchanged is not realistic; this was possible in only 2 patients in our series. In all other revised shoulders, associated bone loss or soft tissue deficiencies forced the surgeon to remove the glenoid component and to bone graft.

According to Walch et al,30 biconcave glenoids with posterior humeral subluxation (B1 and B2 glenoids) are found in about half of arthritic patients who undergo preop-erative CT scans before TSA. Another interesting and important finding of the present study is that such patients with biconcave glenoid and posterior humeral subluxation had a higher risk of glenoid loosening and earlier revision (Fig. 5). We have observed that after anatomic TSA, posterior humeral subluxation may recur with time and cause asymmetric accelerated PE wear, resulting in metal-metal contact and glenoid loosening.2 These results are important and should lead surgeons (1) to routinely ask for CT scans (or magnetic resonance images) before deciding on a treatment option in patients with primary OA and (2) to discuss alternative treatment options in case of severe posterior subluxation with B2 glenoid. Recently, RSA has been proposed in case of humeral subluxation associated with B2 glenoid, and

8 12 21 2728

early good clinical outcomes have been reported. ^^

The low survival rate of MB implants with PE insert found in this study (only 46% free of revision at 12 years) is a concern. As shown in Figure 4, there is a severe drop of the survival curve after the fourth year. In other words, complications start to be severe enough to require revision surgery about 4 years after primary prosthesis implantation. The catastrophic failure rate of MB glenoids with PE insert found in this study should lead surgeons who still use this kind of implant to trace their patients and call them back for regular radiographic controls. Interestingly, we found that PE wear was associated with the development of osteolysis of the proximal part of the humerus. This is an important and indirect radiologic sign that should alert surgeons and lead them to propose revision surgery before more severe humeral and glenoid osteolysis occurs.

Gilles Walch has received royalties from Tornier.

The other authors, their immediate families, and any research foundation 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.

Supplementary data

Supplementary data related to this article can be found online at 10.1016/j.jse.2015.02.012.

References

Conclusion

With the recent success of reverse prostheses, there is currently renewed interest in the development of universal uncemented glenoid MB implants that could be used for both anatomic shoulder arthroplasty and RSA. These MB glenoid implants would allow surgeons to more easily convert failed TSA into RSA without having to revise the glenoid component. In theory, exchange of the PE insert for a metallic sphere (with conservation of the MB tray) would make the revision procedure easier and faster. As mentioned before, conservation of the MB tray with reinsertion of a new PE insert (after removal of the worn PE insert) was possible in only 3% (2 of 61) of the revised shoulders. Thus, the belief that a modular uncemented glenoid system would be easily converted from TSA to RSA is called into question by our studies. On the basis of our results, we can expect that such universal MB glenoid implants with PE insert (i.e., the anatomic version) will lead invariably to high rates of complication and revision.

Acknowledgment

The authors wish to thank Luc Favard, MD (Centre Hospitalier Universitaire de Tours, France), and Francois Sirvaux, MD, and Daniel Mole, MD (Clinique de Traumatologie de Nancy, France), for providing some cases included in this series.

Disclaimer

Pascal Boileau has received royalties from Tornier.

Christophe Levigne has received royalties from Tornier.

1. Boileau P, Avidor C, Krishnan SG, Walch G, Kempf J-F, Molé D. Cemented polyethylene versus uncemented metal-backed glenoid components in total shoulder arthroplasty: a prospective, double-blind, randomized study. J Shoulder Elbow Surg 2002;11:351-9. http://dx. doi.org/10.1067/mse.2002.125807

2. Boileau P, Avidor C, Walch G. Prothèse totale d'épaule avec un implant glenoidien metallique non-scelle: etude prospective de 354 implants suivis è plus de 2 ans. In: Boileau P, Walch G, Molé D, editors. 2000 shoulder prostheses: two to ten year follow-up. Montpellier: Sauramps meédical; 2001. p. 489-507.

3. Boileau P, Walch G. The three-dimensional geometry of the proximal humerus. Implications for surgical technique and prosthetic design. J Bone Joint Surg Br 1997;79:857-65.

4. Boileau P, Watkinson DJ, Hatzidakis AM, Balg F. Grammont reverse prosthesis: design, rationale, and biomechanics. J Shoulder Elbow Surg 2005;14(Suppl S):147S-61S. http://dx.doi.org/10.1016/j.jse.2004. 10.006

5. Budge MD, Nolan EM, Heisey MH, Baker K, Wiater JM. Results of total shoulder arthroplasty with a monoblock porous tantalum glenoid component: a prospective minimum 2-year follow-up study. J Shoulder Elbow Surg 2013;22:535-41. http://dx.doi.org/10.1016/jjse.2012. 06.001

6. Castagna A, Randelli M, Garofalo R, Maradei L, Giardella A, Borroni M. Mid-term results of a metal-backed glenoid component in total shoulder replacement. J Bone Joint Surg Br 2010;92:1410-5. http://dx.doi.org/10.1302/0301-620X.92B10.23578

7. Cheung EV, Sperling JW, Cofield RH. Polyethylene insert exchange for wear after total shoulder arthroplasty. J Shoulder Elbow Surg 2007; 16:574-8. http://dx.doi.org/10.1016/jjse.2006.12.009

8. Clitherow HDS, Frampton CMA, Astley TM. Effect of glenoid cementation on total shoulder arthroplasty for degenerative arthritis of the shoulder: a review of the New Zealand National Joint Registry. J Shoulder Elbow Surg 2014;23:775-81. http://dx.doi.org/10.1016/jjse. 2013.08.022

9. Cofield RH, Daly PJ. Total shoulder arthroplasty with a tissue-ingrowth glenoid component. J Shoulder Elbow Surg 1992;1:77-85.

10. Collin P, Tay AK, Melis B, Boileau P, Walch G. A ten-year radiologic comparison of two-all polyethylene glenoid component designs: a prospective trial. J Shoulder Elbow Surg 2011;20:1217-23. http://dx. doi.org/10.1016/j.jse.2011.06.012

11. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop Relat Res 1987;214:160-4.

12. Denard PJ, Walch G. Current concepts in the surgical management of primary glenohumeral arthritis with a biconcave glenoid. J Shoulder Elbow Surg 2013;22:1589-98. http://dx.doi.org/10.1016/jjse.2013.06.017

13. Fox TJ, Cil A, Sperling JW, Sanchez-Sotelo J, Schleck CD, Cofield RH. Survival of the glenoid component in shoulder arthroplasty. J Shoulder Elbow Surg 2009;18:859-63. http://dx.doi.org/10. 1016/j.jse.2008.11.020

14. Fucentese SF, Costouros JG, Kuhnel S-P, Gerber C. Total shoulder arthroplasty with an uncemented soft-metal-backed glenoid component. J Shoulder Elbow Surg 2010;19:624-31. http://dx.doi.org/10. 1016/j.jse.2009.12.021

15. Gerber C, Costouros JG, Sukthankar A, Fucentese SF. Static posterior humeral head subluxation and total shoulder arthroplasty. J Shoulder Elbow Surg 2009;18:505-10. http://dx.doi.org/10.1016/j.jse.2009.03.003

16. Goutallier D, Postel J-M, Gleyze P, Leguilloux P, Van Driessche S. Influence of cuff muscle fatty degeneration on anatomic and functional outcomes after simple suture of full-thickness tears. J Shoulder Elbow Surg 2003;12:550-4. http://dx.doi.org/10.1016/S1058-2746(03)00211-8

17. Kempf JF, Walch G, Lacaze F. Results of shoulder arthroplasty in primary gleno-humeral osteoarthritis. In: Walch G, Boileau P, editors. Shoulder arthroplasty. New York: Springer; 1999.

18. Lawrence TM, Ahmadi S, Sperling JW, Cofield RH. Fixation and durability of a bone-ingrowth component for glenoid bone loss. J Shoulder Elbow Surg 2012;21:1764-9. http://dx.doi.org/10.1016/]. jse.2011.12.013

19. Martin SD, Zurakowski D, Thornhill TS. Uncemented glenoid component in total shoulder arthroplasty. Survivorship and outcomes. J Bone Joint Surg Am 2005;87:1284-92. http://dx.doi.org/10.2106/JBJS.C00947

20. McElwain JP, English E. The early results of porous-coated total shoulder arthroplasty. Clin Orthop Relat Res 1987;218:217-24.

21. Mizuno N, Denard PJ, Raiss P, Walch G. Reverse total shoulder arthroplasty for primary glenohumeral osteoarthritis in patients with a biconcave glenoid. J Bone Joint Surg Am 2013;95:1297-304. http://dx. doi.org/10.2106/JBJS.L.00820

22. Mole D, Roche O, Riand N, Levigne C, Walch G. Cemented glenoid component: results in osteoarthritis and rheumatoid arthritis. In: Walch G, Boileau P, editors. Shoulder arthroplasty. New York: Springer; 1999. p. 163-71.

23. Montoya F, Magosch P, Scheiderer B, Lichtenberg S, Melean P, Habermeyer P. Midterm results of a total shoulder prosthesis fixed with a cementless glenoid component. J Shoulder Elbow Surg 2013; 22:628-35. http://dx.doi.org/10.1016/j.jse.2012.07.005

24. Norris TR, Iannotti JP. Functional outcome after shoulder arthro-plasty for primary osteoarthritis: a multicenter study. J Shoulder Elbow Surg 2002;11:130-5. http://dx.doi.org/10.1067/mse.2002. 121146

25. Rosenberg N, Neumann L, Modi A, Mersich IJ, Wallace AW. Improvements in survival of the uncemented Nottingham Total Shoulder prosthesis: a prospective comparative study. BMC Musculoskelet Disord 2007;8:76. http://dx.doi.org/10.1186/1471-2474-8-76

26. 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;13:604-13. http://dx.doi.org/10.1016/jjse.2004.03.013

27. Taunton MJ, Mcintosh AL, Sperling JW, Cofield RH. Total shoulder arthroplasty with a metal-backed, bone-ingrowth glenoid component. Medium to long-term results. J Bone Joint Surg Am 2008;90:2180-8. http://dx.doi.org/10.2106/JBJS.G00966

28. Torchia ME, Cofield RH, Settergren CR. Total shoulder arthroplasty with the Neer prosthesis: long-term results. J Shoulder Elbow Surg 1997;6:495-505.

29. Walch G, Badet R, Boulahia A, Khoury A. Morphologic study of the glenoid in primary glenohumeral osteoarthritis. J Arthroplasty 1999; 14:756-60.

30. Walch G, Moraga C, Young A, Castellanos-Rosas J. Results of anatomic nonconstrained prosthesis in primary osteoarthritis with biconcave glenoid. J Shoulder Elbow Surg 2012;21:1526-33. http://dx. doi.org/10.1016/j.jse.2011.11.030

31. Wallace AL, Phillips RL, MacDougal GA, Walsh WR, Sonnabend DH. Resurfacing of the glenoid in total shoulder arthroplasty. A comparison, at a mean of five years, of prostheses inserted with and without cement. J Bone Joint Surg Am 1999;81: 510-8.