Observations on retrieved glenoid components

The Journal of Arthroplasty Vol. 18 No. 3 2003

Observations on Retrieved Glenoid Components Ralph Hertel, MD, and F. T. Ballmer, MD

Abstract: To obtain more information on the pattern of damage of prosthetic glenoid components, we analyzed 7 retrieved glenoid components. The consecutive series included 2 standard polyethylene components and 5 highly crystalline polyethylene glenoids (Hylamer; DePuy Dupont Orthopaedics, Warsaw, IN) retrieved 3 to 12 years after implantation. At revision, 4 of 5 Hylamer components were fractured. Common wear patterns were i) deformation and crumbling of the rim, particularly at the inferior hemicircumference, probably caused by direct contact of the humerus with the prosthetic component; ii) roughening (abrasion and scratching) of the adjacent articulating surface; and iii) concentric and congruous wear centered posteriorly. Available glenoid components may cover an excessive sector of the head. This can result in mechanical restriction of glenohumeral motion and abutment of the humerus against the glenoid rim. Abutment may cause major shear forces and therefore cause glenoid loosening. The value of articular surface mismatch is questionable because retrieved glenoids were worn to a conforming joint. Key words: shoulder, arthroplasty, glenoid component, complication, radial mismatch, glenohumeral impingement. © 2003 Elsevier Inc. All rights reserved.

Despite encouraging early reports, glenoid loosening has been recognized as a major concern in total shoulder replacement [1]. Reasons for glenoid failure have been associated with the indication for glenoid replacement, the surgical technique, and particular features of the implant. The most critical implant-related features are size, conformity of the articular surface, and type of fixation in the scapular bone [2]. Previous studies on retrieved glenoid components have focused on mechanical damage of the articulating surface and have not considered the glenohumeral relationship [3,4].

We examined a consecutive series of retrieved glenoid components, including the humeral component, to provide some clinical arguments concerning glenoid implant design.

Material and Methods Seven glenoid components were retrieved because of symptomatic loosening at an average time of implantation of 5.5 years (3.1 to 12.2 years). Patients included 4 women and 3 men. At revision, the mean patient age was 56 years (34 to 80 years). The diagnoses were osteoarthritis in 5 and rheumatoid arthritis in 2 patients. The rotator cuff was intact in all patients. Five glenoid components were made of highly crystalline polyethylene (Hylamer; DePuy Dupont Orthopaedics, Warsaw, IN) and 2 were made of standard, ultra-high molecular weight polyethylene (1 Neer-3m [Smith and Nephew Inc, Memphis, TN] and 1 Aequalis-Tornier [Tornier, Montbonnot, France] component). The

From the Department of Orthopaedic Surgery, Inselspital, University of Berne, Switzerland. Submitted September 14, 2001; accepted September 5, 2002. No benefits or funds were received in support of this study. Reprint requests: Ralph Hertel, MD, Department of Orthopedic Surgery, Inselspital, University of Berne, 3010 Berne, Switzerland. © 2003 Elsevier Inc. All rights reserved. 0883-5403/03/1803-0019$30.00/0 doi:10.1054/arth.2003.50048

361

362 The Journal of Arthroplasty Vol. 18 No. 3 April 2003 macrostructured surface as caused by direct bony contact), and scratching. Deformation included rim deformation, rim crumpling, and articulating surface deformation. Deterioration included squamous delamination, pitting, and decoloration. The articulating surfaces were checked for embedded bodies (either metal or polymethylmethacrylate particles). The degree of articular congruency resulting from the fine abrasion process was checked against the retrieved prosthetic head.

Results and Observations Fig. 1. A glenoid retrieved 4.3 years after implantation (patient No. 2) is shown. The inferior rim is deformed, and the articular surface shows a secondary facet whose radius matches the radius of the head. The center of the new facet is shifted dorsally and inferiorly. Deformation of the rim was probably caused by abutment, with the calcar (medial neck) region against the glenoid.

Hylamer-Global shoulder components and the Aequalis component featured an original radial mismatch of 4 mm. The Neer component featured conforming articular surfaces. At the time of revision all patients suffered from increasing mechanical pain localized in the dorsal aspect of the scapula. Serial radiographs showed and increasing width of radiolucent lines around the pegs or the keel, indicating loosening of the glenoid component. The relationship between the glenoid and the prosthetic head as well as between the glenoid and the humerus was assessed intraoperatively by the first author (R.H). The main focus was on possible impingement of the humerus against the glenoid. Examination of the retrieved components under stereoscopic magnification included the evaluation of the articular surface and the assessment of rim lesions and of conformity of the articular facet. To provide optimal visualization of characteristic structural changes, representative areas were additionally examined using scanning electron microscopy (SEM) (Philips XL 30 FEG, Philips, Amsterdam, the Netherlands). Description of surface alterations was derived from the terminology of Hood, Wright, and Burstein [5]. It included the following major categories: abrasion, deformation, deterioration, and inclusion of debris. Abrasion was further subdivided as fine abrasion (synonymous to polishing and burnishing), rough abrasion (macroscopically irregular,

All glenoid components (Table 1) were grossly loose and could be easily extracted. Four of 5 Hylamer glenoids were fractured. The fractures ran through the narrowest point of the glenoid and were predominantly vertically oriented in 3 of 4 cases (Figs 1, 2). The glenoid rim was deformed and obviously mechanically damaged (crumpled) to a varying degree. Deformation was most prominent at the inferior aspect (Fig. 1) and crumpling was predominantly evident at the anterior rim (Fig. 3). The articulating surface close to the deformed rim showed a varying degree of roughening (rough abrasion and scratching) (Fig. 4). These areas corresponded to regions of direct contact of the humerus with the glenoid as observed intraoperatively. For example, the roughening shown in Fig. 3 was due to direct scraping of the calcar against the glenoid. Anterior rim lesions were most probably due to direct contact of the lesser tuberosity with the side-

Fig. 2. A glenoid retrieved 5.2 years after implantation (patient No. 4) is shown. (A) The new conforming articular facet is centered superiorly. (B) The original articular surface showing a radial mismatch of 4 mm is roughened and obviously functionally excluded.

Glenoid Wear • Hertel and Ballmer

363

Discussion

Fig. 3. A glenoid retrieved 5.1 years after implantation (patient No. 1) is shown. (A) The new conforming articular facet is centered superiorly. (B) The original articular surface is roughened and shows grinding tracks induced by protruding spurs based on the neck of the humerus. Mechanical lesions (distortion and crumpling) of the anterior (arrows) and posterior rim are clearly visible. The superior defect is an artifact induced by extraction.

wall of the implants. The central articulating surface appeared matte, as if sandblasted, in all Hylamer specimens. The radius of the new concavity was smaller than the original radius of the 6 glenoids that had an initial radial mismatch. The newly formed concavity perfectly matched the radius of the prosthetic head (Fig. 1) in all specimens. In the 2 standard polyethylene components, surface delamination and pitting occurred (Fig. 4). This made precise determination of the new radius impossible. In all specimens, the new articulating facet was centered slightly posterior to the center of the original glenoid component (Fig. 1). In 4 cases, the center was shifted posteriorly and superiorly (Fig. 1), and in 3 cases the center was shifted posteriorly and inferiorly (Figs. 2, 3). Delamination and discoloration were most prominent in the standard polyethylene components. Inclusions were not seen.

Observations of retrieved glenoids have a potential impact on prosthetic design and implantation. Rim roughening and deformation have been noted by previous investigators [3,4]. These authors believed this to be a result of cold flow. Further insight could be obtained by analyzing the articular pairing in situ at the time of revision. At in vivo inspection, it was evident that rim deformation was caused by direct abutment of the humeral metaphysis (neck region and tuberosities) against the prosthetic device. Abutment can potentially lead to large shear moments on the glenoid component. The magnitude of these moments depends on the angular velocity of the arm, its weight, and the time allowed for energy dissipation. Approximate calculations of the resulting impulse showed that abutment might lead to large shear forces acting on the glenoid. These forces could be among the primary causes of loosening. The relative sizes of the humeral head and the glenoid determine the possible range of motion. The possible range of motion depends on the magnitude of the 3-dimensional sector of glenoid covering the humeral head and on the ratio of the head radius versus the head height (ie, the sector of the sphere the head comprises). Examination of 5 commercially available prosthetic systems [6] revealed an abutment-free arc of glenohumeral abduction ranging between 16° and 97° (intermediate sizes allowed approximately 65°). The arc of glenohumeral rotation ranged between 32° and 120° (intermediate sizes approximated 60°) [6]. For most pairings, the range of possible motion was clearly insufficient for normal joint function. Exercising the joint beyond the mechanically imposed limit will lead to glenohumeral impingement and might induce loosening. Articular conformity is a highly debated issue. Some radial mismatch has been shown to allow a more physiologic motion and to decrease rim loads imposed by the prosthetic head, therefore reducing the so-called rocking horse phenomenon [7,8]. The rocking horse phenomenon has been considered the primary cause of glenoid loosening [7,8]. The loosening moments may be initiated by eccentric load acting on a segment of the glenoid rim, occurring when the head tends to sublux in a given direction. Nevertheless, clinical evidence of improved outcomes for mismatched pairings is still missing. In fact, our observations of retrieved glenoid components, on the contrary, favor a con-

364 The Journal of Arthroplasty Vol. 18 No. 3 April 2003

Fig. 4. Scanning electron microscopy of representative surface alterations is shown. (A) Transition between the congruent articulating surface and the incongruent original articulating surface is shown. The latter can be recognized by the machine-structured surface. A clear transition line (TL) between congruent and incongruent articulating surfaces can be observed (patient No. 2). (B) The transition between the congruent articulating surface and rough abraded surface (patient No. 3) is shown. In this case, the incongruent segment is not visible. Rough abrasion was due to direct contact with humeral bone. (C) The transition between the congruent articulating surface and rough abraded surface (patient No. 6) is shown. Rough abrasion was due to direct contact with humeral bone. (D) Minor stages of rim deformation and rim crumpling due to glenohumeral impingement (patient No. 1) is shown. (E) Polyethylene wear: fine abrasion and scratching of the articulating surface is seen (patient No. 1). (F) Polyethylene degradation in the form of pitting (patient No.1) is seen. (G) Polyethylene degradation reveals squamous delamination (patient No. 1).

forming design. Nature reshaped nonconforming glenoids to a conforming articular facet, reducing the stresses on the polyethylene imposed by the mismatch. It is possible that the initial contract geometry leads to stresses that exceeded the yield stress of polyethylene. These observations were particularly evident for Hylamer glenoids; standard polyethylene tended to deform and to delaminate. These observations confirm those of other researchers [2– 4].

Weldon and Scarlat et al. [9] also noted alteration of the surface geometry of the polyethylene components in vivo. The authors assessed the change in intrinsic stability of glenoid components as indicated by the balance stability angle and concluded that glenohumeral stability decreases over time. At first glance, this finding contrasts our observation of increasing articular conformity. However, the explanation is that we were looking at different aspects of the same problem. Weldon and Scarlat et

Glenoid Wear • Hertel and Ballmer

365

Table 1. Synopsis of Retrieved Glenoids

Gender

Age (y)

Time to Revision on (y)

1

M

59

5.1

2

F

34

4.3

3

F

53

4.8

4

M

59

5.2

5

F

80

4.1

6

F

62

12.2

7

M

49

3.1

Patient No.

Type of Glenoid

Type of Head

Radial Mismatch (mm)

Global 48 pegged Hylamer Global 44 pegged Hylamer Global 44 keeled Hylamer Global 52 pegged Hylamer Aequalis 43/ 46M standard polyethylene Neer keeled standard Polyethylene Global 52 pegged Hylamer

Aequalis

4

Intact

From 1 to 10

Yes, dorsocranially

Aequalis

4

Horizontal fracture

From 4 to 9

Yes, dorsocaudally

Global

4

Vertical fracture

Circumferential

Yes, dorsocaudally

Global

4

3 Part fracture

No lesion

Yes, dorsocranially

Aequalis

3

Intact

From 4 to 10

Yes, dorsocranially

Neer

0

Intact

From 4 to 8

Yes, dorsocranially

Aequalis

4

Vertical fracture

From 6 to 9

Yes, dorsocaudally

al. [9] did not have matching heads available to check for secondary congruency. Their methodology basically assessed the depth of the glenoid concavity, which is biased by any associated rim deformity. This may result in a more shallow joint concavity, even though the articulating part of the joint surface perfectly matches the humeral head. Hylamer has virtually no cold flow properties, but can be ground by the prosthetic head (like a grindstone on sandstone). To our knowledge, this material is no longer used [10]. Hylamer did allow more insight into in vivo wear conditions. Based on our in vivo observations, we can assume that full articular conformity would be preferable. The advantages would be the reduction of contact stresses and therefore less wear; guided motion centered over the glenoid (not posteroinferiorly), less eccentric load on the glenoid, and better load distribution at the subprosthetic bone interface [11]. Our observations were recently confirmed by experimental data on glenohumeral kinematics [12]. Kelkar and Wang et al. [12] concluded that the proximity of the origin of the helical axes to the geometric center of the humeral head articular surface confirmed that glenohumeral elevation is mainly concentric. Glenohumeral translations have probably been overestimated [13], and the extreme ranges of motion may not apply to the artificial joint environment. For

Fragmentation of the Glenoid

Glenoid Rim Lesion (x to y o’clock)

Secondary Congruence

practical purposes, the following scenario seems realistic: at rest there may be a slight caudal subluxation of the humeral head; in the presence of a functional cuff this is immediately corrected when active elevation is initiated; during elevation, the center of rotation remains fairly constant and at the end of the range of motion, the head tends to translate inferiorly. Concavity compression seems to be the mechanism responsible for dynamic centering of the head on the glenoid socket [14,15]. There is sufficient evidence that the radial mismatch between the bony articulating surfaces [16,17] is compensated for by varied cartilage thickness, including functional thickness under compression [18 –20]. In vivo dynamic magnetic resonance examinations are a clear proof for functional articular congruity.

Conclusions Available glenoid components may cover an excessive amount of the head. This may result in mechanical restriction of glenohumeral motion and abutment of the humerus against the glenoid rim. Abutment may cause major shear forces and therefore be a cause of glenoid loosening. The value of articular surface mismatch is questionable because retrieved glenoids were “naturally” worn to a con-

366 The Journal of Arthroplasty Vol. 18 No. 3 April 2003 forming joint. The relative dimensions of glenoid components and the concept of radial mismatch may require revision.

Acknowledgments The authors express their gratitude to Mr. R. Babl, Institute of Anatomy, University of Berne, for his support in obtaining the scanning electron microscopy analysis.

9.

10.

11.

References 12. 1. Torchia ME, Cofield RH, Settergren CR: Total shoulder arthroplasty with the Neer prosthesis: long-term results. J Shoulder Elbow Surg 6:495, 1997 2. Wirth MA, Seltzer DG, Senes HR. An analysis of failed humeral head and total shoulder arthroplasty. The 10th Open Meeting of American Shoulder and Elbow Surgeons, New Orleans, February, March 1994 3. Gunther SB, Graham J, Norris TR, et al: Retrieved glenoid components: a classification system for surface damage analysis. J Arthroplasty 17:95, 2002 4. Scarlat MM, Matsen FA, 3rd: Observations on retrieved polyethylene glenoid components. J Arthroplasty 16:795, 2001 5. Hood R, Wright T, Burstein A: Retrieval analysis of total knee prostheses: a method and its application to 48 condylar prostheses. J Biomed Mater Res 17:829, 1983 6. Schoell F, Ballmer FT, Hertel R: Gleno-humeral impingement: a likely cause of glenoid loosening. In Monteiro J, Cartucho A, Amaral L (eds): 14th Congress of the European Society for Surgery of the Shoulder and the Elbow, Lisbon, September 2000 7. Harryman DT, Sidles JA, Harris SL, et al: The effect of articular conformity and the size of the humeral head component on laxity and motion after glenohumeral arthroplasty: a study in cadavera. J Bone Joint Surg Am 77:555, 1995 8. Severt R, Thomas BJ, Tsenter MJ, et al: The influence of conformity and constraint on translational forces

13.

14.

15.

16.

17.

18.

19.

20.

and frictional torque in total shoulder arthroplasty. Clin Orthop 292:151, 1993 Weldon EJ 3rd, Scarlat MM, Lee SB, Matsen FA 3rd: Intrinsic stability of unused and retrieved polyethylene glenoid components. J Shoulder Elbow Surg 10:474, 2001 Chmell MJ, Poss R, Thomas WH, Sledge CB: Early failure of Hylamer acetabular inserts due to eccentric wear. J Arthroplasty 11:351, 1996 Mansat P, Couteau P, Darmana R, et al: Mechanical analysis of a scapula implanted with a glenoid prostheis using the finite element method. In Monteiro J, Cartucho A, Amaral L (eds): 14th Congress of the European Society for Surgery of the Shoulder and the Elbow, Lisbon, September 2000 Kelkar R, Wang VM, Flatow EL, et al: Glenohumeral mechanics: a study of articular geometry, contact, and kinematics. J Shoulder Elbow Surg 10:73, 2001 Poppen NK, Walker PS: Normal and abnormal motion of the shoulder. J Bone Joint Surg Am 58:195, 1976 Howell SM, Galinat BJ: The glenoid-labral socket: a constrained articular surface. Clin Orthop 243:122, 1989 Lippit SP, Vanderhooft JE, Harris SL, et al: Glenohumeral stability form concavity-compression: a quantitative analysis. J Shoulder Elbow Surg 2:27, 1993 Iannotti JP, Gabriel JP, Schneck SL, et al: The normal glenohumeral relationships: an anatomical study of one hundred and forty shoulders. J Bone Joint Surg Am 74:491, 1992 McPherson EJ, Friedman RJ, An YH, et al: Anthropometric study of normal glenohumeral relationships. J Shoulder Elbow Surg 6:105, 1997 Boileau P, Walch G: Normal and pathological anatomy of the glenoid: effects on the design, preparation and fixation of the glenoid component. p. 127. In Walch G, Boileau P (eds): Shoulder arthroplasty. Springer, Berlin, 1999 Soslowsky LJ, Flatow EL, Bigliani LU, et al: Quantitation of in situ contact areas at the glenohumeral joint: a biomechanical study. J Orthop Res 10:524, 1992 Soslowsky LJ, Flatow EL, Bigliani LU, Mow VC: Articular geometry of the glenohumeral joint. Clin Orthop 285:181, 1992