Hemiarthroplasty versus total shoulder prosthesis: Results of cemented glenoid components

Hemiarthroplasty versus total shoulder prosthesis: Results of cemented glenoid components

Hemiarthroplasty versus total shoulder prosthesis: Results of cemented glenoid components M. Pfahler, MD, F. Jena, L. Neyton, F. Sirveaux, MD, and D. ...

205KB Sizes 13 Downloads 96 Views

Hemiarthroplasty versus total shoulder prosthesis: Results of cemented glenoid components M. Pfahler, MD, F. Jena, L. Neyton, F. Sirveaux, MD, and D. Molé, MD, Munich, Germany, and Nancy Cedex, France

In this retrospective study, we compared the results of 705 total shoulder arthroplasties (TSAs) with 469 hemiarthroplasties (HSAs), all having been performed with the Aequalis shoulder prosthesis. Each group, both TSA and HSA, was comparable by age (mean, 63.9 years) and sex (853 men and 321 women). Each group had comparable dominance and preoperative Constant scores (mean, 29 points). The length of follow-up averaged 43 months (range, 24 –110 months) in both groups. The postoperative functional outcome and subjective assessment demonstrated the superiority of TSA over HSA independent of age or rotator cuff status (Constant score, 65.7 vs 56.3 points). The analysis of the radiographs showed a disturbing 68% of cases with radiolucent lines developing around the glenoid component and their subsequent progression with time. We saw an adverse effect on functional outcome by the presence of these radiolucent lines. This review would suggest that TSA is superior to HSA in most cases with chronic pathologic entities. HSA remains a satisfactory solution in specific cases. In the future, we need to optimize the designs of the glenoid implant and develop a better implantation technique to avoid the problems associated with glenoid replacement. (J Shoulder Elbow Surg 2006; 15:154-163.)

T

he use of a glenoid implant in chronic pathologic conditions of the shoulder continues to be controversial. The surgeon’s fundamental decision of whether to perform a total shoulder arthroplasty (TSA) or hemiarthroplasty (HSA) remains a key question. In most From the Center for Shoulder and Elbow Surgery, Airport Clinic M, Munich, Germany, and the Department of Orthopaedics, Ludwig-Maximilians-University Munich, Klinikum Grosshadern, Munich, Germany, and Clinique de Traumatologie et d’Orthopédie, Nancy Cedex, France. Reprint requests: M. Pfahler, MD, Center for Shoulder and Elbow Surgery, Airport Clinic M, Munich, Terminalstraße Mitte 18, D-85356 Munich, Germany ([email protected]). Copyright © 2006 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2006/$32.00 doi:10.1016/j.jse.2005.07.007

154

studies, the results of TSA are reported to be better than those of HSA.2,4,6 However, all authors agree on the difficulties associated with implanting the glenoid component6 and the risk of its loosening over time.2,9 Because of this risk, some authors have abandoned the glenoid implant, some reserve it only for the older and less active patient,1,8 and others use noncemented glenoid implants, which have not proven their superiority to date.10 The purpose of this study was to compare the results of 705 Aequalis TSAs (Aequalis Prosthetic System, Tornier Company, St. Ismier Cedex, France) with a cemented glenoid implant with the results of 469 Aequalis HSAs. This review compares the functional outcome of these two types of shoulder arthroplasty as measured by the Constant score. We performed a thorough radiographic analysis of TSAs looking at the anchorage of the glenoid component and the presence or absence of its loosening, as well as its significance with regard to clinical outcome. Although this represents a retrospective multicenter study, it has the advantage of evaluating a large number of cases in which the same type of implant was used. We felt that analysis of the results should provide useful information and some conclusions that would be helpful in clinical practice. MATERIALS AND METHODS The material analyzed in this study came from the International Aequalis Multicenter Group, whose retrospective results were presented in Nice, France, in 2001. The group was composed of 57 surgical teams from 9 European countries and the Republic of South Africa. Our analysis evaluated 469 Aequalis HSAs and 705 Aequalis TSAs with a cemented, all- ployethylene glenoid. The TSA cases were isolated from 1072 TSAs. We excluded 351 cases in which a metal-backed glenoid component was used; these will be the subject of a separate analytic study. Sixteen cases were excluded because a non-Aequalis glenoid implant was used. In 585 cases (83%), a keeled glenoid implant with a flat back was used, in 62 cases (9%), a pegged glenoid implant with a flat back was used; and in 58 cases (8%), a keeled glenoid with a convex back was used. These three types of polyethylene implants were analyzed as one group (Figure 1). The comparison of the polyethylene glenoid implants was described in a different study. The mean patient age in the whole series was 63.9 years (range, 15–90 years). There were 853 women

Pfahler et al

J Shoulder Elbow Surg Volume 15, Number 2

155

and TSA (Table II), as well as its role in the appearance of radiographic loosening signs at the glenoid implant. Secondary glenoid erosions and osteoarthritis in HSAs were documented but not specifically evaluated within this series.

Statistics Figure 1 Types of glenoid implants used; with pegs (A); keel, flat back (B); and keel, curved back (C).

(72.7%) and 321 men (27.3%). The dominant arm was affected in 723 cases (62.3%). Age, sex, and dominant arm were comparable in both groups (TSA and HSA) (Table I). All patients were re-evaluated, with a mean follow-up of 43 months (range, 24 –110 months). Preoperative and postoperative radiographs were evaluated. The mean preoperative Constant score was 28.5 points (37.8% adjusted Constant score) in the HSA group and 28.6 points (38.4%) in the TSA group; it was 28.6 points (38.2%) in the total series. The preoperative functional conditions were comparable in both groups, with the exception of a difference of mobility in active anterior elevation (80.5° in HSA group versus 87° in TSA group, P ⫽ .004) and external rotation (17.5° in HSA group versus 12° in TSA group, P ⬍ .001). To evaluate the importance of the rotator cuff, its status was documented clinically, radiographically, and intraoperatively. We analyzed the influence of the integrity of the cuff (intact, partial tear, or complete tear). Overall, the condition of the rotator cuff was better in the TSA group. In this group, we noted partial tears in 9.3%, complete tears in 7.4%, and 83.3% without tears of the supraspinatus tendon. In the HSA group, we noted partial tears in 32%, complete tears in 8.5%, and 59.5% without tears (P ⬍ .001). Fatty infiltration of the rotator cuff was evaluated by computed tomography according to the classification of Goutallier et al.5 Fatty infiltration stages 3 and 4 of the infraspinatus and subscapularis muscles were greater in the HSA group (infraspinatus, 24.2%, and subscapsularis, 14.4%) versus the TSA group (infraspinatus, 7.4%, and subscapularis, 6.1%). With our study, we wanted to analyze the functional outcome of the two groups. We looked for a correlation between functional outcome and any of the parameters of age, sex, dominance, preoperative clinical conditions, or condition of the rotator cuff. In addition, we wanted to analyze the radiographic signs of glenoid loosening in the TSA group by use of the radiolucent line (RLL) score.6 Using this score, we judged the presence of radiolucent lines in 6 zones surrounding the glenoid implant. In each zone, we measured the size of the radiolucent line. The radiolucent lines were graded as follows: no radiolucent line, 0 points; less than 1 mm, 1 point; less than 2 mm, 2 points; and 2 mm or greater, 3 points. We summed the points in each zone and obtained a range score of 0 to 18 points (Figure 2). We looked for predisposing factors for the presence of radiographic loosening, along with its possible functional consequence. We were also interested in determining what role the preoperative pathology played in the comparison of HSA

We used the Fisher exact test for qualitative variables with two modalities and the ␹2 test for qualitative variables with more than 2 modalities. P ⬍ .05 indicated a statistically significant relationship between the variables and the groups. We utilized the Student t test (2-sided) to compare the means of the two groups. P ⬍ .05 indicated a significant difference between the means of the variables studied in the two groups. The Levene test was used to determine whether the variances were equal (P ⬎ .05) or unequal (P ⬍ .05).

RESULTS Complications and revisions

There were 57 immediate complications: 18 (3.8%) in the HSA group and 39 (5.5%) in the TSA group (P ⫽ .2). There were 9 immediate surgical revisions: (0.8%) 5 (1.1%) in the HSA group and 4 (0.6%) in the TSA group (P ⫽ .49). There were 135 late complications: 55 (11.7%) in the HSA group and 80 (11.3%) in the TSA group (P ⫽ .85). There were 97 secondary surgical revisions (8.3%): 54 (11.5%) in the HSA group and 43 (6.1%) in the TSA group (P ⫽ .001). In the immediate complication group, there were 7 humerus fractures in the HSA group and 13 in the TSA group. There were 2 temporary neurologic deficits in the HSA group and 11 in the TSA group. There were 4 dislocations in the HSA group and 6 in the TSA group (Table III). In the late complication group, there were 13 complications related to humerus fractures (4 in the HSA group and 9 in the TSA group), as well as 15 infections (7 in the HSA group and 8 in the TSA group), 23 cuff deficiencies (7 in the HSA group and 16 in the TSA group), and 28 instabilities (12 in the HSA group and 16 in the TSA group). There were 20 prosthesis failures, 1 in the HSA group and 19 in the TSA group. There were 18 cases of secondary glenoid erosion in the HSA group (3.8%) (Table IV). The primary reason for immediate revision was fracture (5), infection (2), and hematoma (2). Secondary revisions in the TSA group as a result of glenoid loosening were seen in only 9 cases (1.3%). Functional outcome

The functional outcome, evaluated by use of the Constant score, was significantly better in all categories for the TSA group than for the HSA group (P ⬍ .001) (Table V). The subjective assessment of the outcome indicated

156

Pfahler et al

J Shoulder Elbow Surg March/April 2006

Table I Distribution of age, sex, and dominant side in total series and within groups HSA and TSA

HSA (n ⫽ 469) TSA (n ⫽ 705) Total series (n ⫽ 1174) Significance

Age (y) (range)

Sex

Dominant side

63.2 (16–88) 64.3 (15–90) 63.9 (15–90) P ⫽ NS

330 (70.4%) women and 139 (29.6%) men 523 (74.2%) women and 182 (25.8%) men 853 (72.7%) women and 321 (27.3%) men P ⫽ NS

299 (64%) 424 (61%) 723 (62.3%) P ⫽ NS

NS, Nonsignificant.

Table II Distribution of etiologies

Fracture sequelae Rheumatoid arthritis Primary osteoarthritis Eccentric arthrosis Osteonecrosis Arthrosis after anterior instability Other etiologies Total

HSA

TSA

114 (66%) 49 (31.5%) 102 (19.5%) 72 (85.5%) 58 (76.5%)

59 (34%) 107 (68.5%) 418 (80.5%) 12 (14.5%) 18 (23.5%)

25 (47%) 49 (44%) 469

28 (53%) 63 (56%) 705

Total series 173 156 520 84 76 53 112 1174

Table III Immediate complications

Figure 2 RLL score (0 –18 points). Evaluation of 6 zones. Lucent lines are measured in millimeters (0 –3).

significant superiority of TSA over HSA as well. Of the TSA patients, 92.4% were satisfied or very satisfied versus 84.6% of the HSA patients (P ⬍ .001) (Table VI). The superiority of TSA over HSA was not influenced by the age of the patient at the time of surgery, the condition of the rotator cuff, or the follow-up. We compared patients aged 65 years or younger with patients aged greater than 65 years. In the younger age group, we noted a Constant score of 61 points after HSA versus a score of 66.6 points after TSA (P ⫽ .001). In the older age group, we noted a Constant score of 51.6 points after HSA versus a

Complication

HSA

TSA

Humerus fracture Infection Hematoma Subscapularis lesion Temporary neurologic deficit Irreversible neurologic deficit Anterior dislocation Posterior dislocation Total

7 0 3 1 2 1 3 1 18

13 3 2 1 11 3 4 2 39

score of 65 points after TSA (P ⫽ .001). This would indicate that glenoid replacement represented a functional gain of 9% in younger patients and a 20% gain in older patients compared with HSA (Table VII). We evaluated patients who did or did not have a rotator cuff defect. We noted a Constant score of 60.1 points after HSA and 66.3 points after TSA in cases with an intact rotator cuff. In those patients with a rotator cuff defect, the Constant score was 49.7 points after HSA and 56.7 points after TSA. In this review, we noted a significant lowering of the Constant score in those cases with a partial or complete tear of the supraspinatus tendon (P ⫽ .001), independent of a glenoid replacement. However, the functional outcome was better when the glenoid component was replaced. We noted a gain of 10% in the Constant score in cases without a tear and 14% in cases with a tear (P ⫽ .12) (Table VIII).

Pfahler et al

J Shoulder Elbow Surg Volume 15, Number 2

Table IV Late complications Complication

HSA

TSA

Humerus fracture Infection Subscapularis lesion Supraspinatus lesion Anterior instability Posterior instability Superior migration Glenoid erosion Prosthetic failure Total

4 7 2 5 7 3 8 18 1 55

9 8 5 11 9 7 12 0 19 80

Table V Postoperative Constant score with categories

Pain Mean SD Minimum Maximum Mobility Mean SD Minimum Maximum Activity Mean SD Minimum Maximum Power Mean SD Minimum Maximum Constant score Mean SD Minimum Maximum Adjusted score Mean SD Minimum Maximum

HSA

TSA

Total series

11.4 3.46 0 15

12.5 3.04 0 15

12.1 3.25 0 15

P ⬍ 0.001

24 9.35 0 40

29.3 8.46 2 40

27.2 9.18 0 40

P ⬍ 0.001

13.9 4.7 0 38

16.2 4.3 2 38

15.3 4.6 0 38

P ⬍ 0.001

6.8 5.26 0 25

7.8 5.23 0 25

7.4 5.26 0 25

P ⬍ 0.001

56.3 18.57 4 95

65.7 16.92 13 98

62 18.18 4 98

P ⬍ 0.001

74.3% 23.39 6 127

88.3% 23.51 19 140

82.8% 24.43 6 140

P ⬍ 0.001

Significance

157

Table VI Postoperative subjective assessment assigned by patient to 1 of 4 categories

HSA TSA Total series

Very satisfied

Satisfied

Disappointed

Not satisfied

47.3% 66.5% 59%

37.3% 25.9% 30.4%

13.1% 7.4% 9.6%

2.3% 0.2% 1%

graphs revealed radiolucent lines in 479 cases (67.9%). There were no radiolucent lines in 226 cases (32.1%). We were able to compare several serial radiographs in 129 cases. Evaluation of the serial group demonstrated stability in 64 cases and increasing radiolucency in 65 cases. In 601 cases, the quality of the radiographs allowed us to use the RLL score to evaluate the radiographic findings with regard to glenoid loosening. We defined 3 categories of findings. In category 1, there were 371 cases (61.7%) with no or minimal radiolucent lines (RLL score, 0 –5 points); in category 2, there were 162 cases (26.9%) with intermediate radiolucent lines (RLL score, 6 –11 points); and in category 3, there were 68 cases (11.3%) with severe radiolucent lines (RLL score ⱖ12 points). The mean RLL score was 4.7 points in all cases (Figure 3). Those cases with an RLL score greater than 12 were interpreted as being consistent with a loose glenoid. Correlation of clinical and radiographic parameters

There was not a significant effect on functional outcome throughout the follow-up period in the total series. However, the Constant score in the TSA group had a slight decrease with long-term follow-up, whereas the HSA group remained stable (Table IX). Radiographic results

We analyzed radiographic findings in the TSA group, specifically noting radiolucent lines around the cemented glenoid implants. Analysis of the radio-

We found no predetermining factors in the appearance of radiolucent lines or glenoid loosening (RLL score ⱖ12 points) associated with the epidemiologic data (ie, age, sex, or dominant limb). There was no correlation with the presence of fatty infiltration of the rotator cuff, the condition of the rotator cuff, or the preoperative Constant score with the occurrence of radiolucent lines in the postoperative radiographic follow-up. The type of implant, the glenoid size, the diameter of the head, and the humeral stem were without influence. However, the appearance of glenoid loosening (RLL score ⱖ12 points) seemed to be more frequent in cases with an inclination of 135° or 140° (13.9% or 19.7%) than in cases with an inclination of 125° or 130° (9.4% or 8.2%) (P ⫽ .05). The specific implant evaluated yielded the opportunity of using a fixed inclination of 125°, 130°, 135°, or 140°. The risk of glenoid loosening did not depend on cuff repair at the time of implantation. In cases with a rotator cuff repair, 10.5% had glenoid loosening, whereas 11.6% had it in those cases without a repair. We did note glenoid loosening in 14.4% of cases after a release with capsulotomy but only 6.7% of cases without any release (P ⫽ .02).

158

Pfahler et al

J Shoulder Elbow Surg March/April 2006

Table VII Postoperative Constant score in relation to patients aged 65 years or less and those aged greater than 65 years Patients aged <65 y Constant score

Adjusted score

228 (48.6%)

61 18.58 4 95

75% 22.23 6 122

241 (51.4%)

51.6 17.4 5 95

73.7% 24.51 7 127

323 (45.8%)

66.6 16.4 17 98 P ⫽ .001

84.1% 22.01 20 140

382 (54.2%)

65 17.34 13 95 P ⫽ .001

91.8% 24.20 19 138

n HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Significance

Patients aged >65 y n

Constant score

Adjusted score

Table VIII Postoperative Constant score in relation to rotator cuff status Intact rotator cuff n HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Significance

287 (68%)

566 (90.7%)

Rotator cuff deficiency

Constant score

Adjusted score

60.1 17.83 5 93

77.8% 22.45 7 122

66.3 16.42 13 96 P ⫽ .001

88.7% 22.95 19 138

n

135 (32%)

58 (9.3%)

Constant score

Adjusted score

49.7 15.64 18 86

69% 20.79 26 111

56.8 21.18 14 86 P ⫽ .001

79.9% 29.56 20 123

Table IX Postoperative Constant score in relation to follow-up <3 y

HSA TSA Total series

3–5 y

>5 y

n

Constant score

n

Constant score

n

Constant score

177 205 382

55 67.3 61.6

160 204 364

55.1 65.1 60.7

80 229 309

57.3 63.7 62.1

Postoperative immobilization had no influence on the RLL score. However, patients in rehabilitation centers demonstrated a significantly increased risk of the occurrence of radiolucent lines and glenoid loosening. Fifteen percent of cases with glenoid loosening occurred in patients in rehabilitation centers, whereas only four percent of patients not treated in a rehabilitation center had glenoid loosening (P ⬍ .001) (Table X). Follow-up. Long-term follow-up is crucial in the evaluation of the appearance of radiolucent lines and

its correlation with glenoid loosening. We evaluated all of the cases in three separate groups based on the length of time of follow-up. The first group, with a follow-up time of less than 3 years, had an RLL score of 2.6. The second group, with a follow-up time of 3 to 5 years, had an RLL score of 4.7. The third group, with a follow-up time of more than 5 years, had an RLL score of 6.4 (P ⬍ .001). Effect on function with the appearance of radiolucent lines and glenoid loosening. The appearance of radiolucent lines around the glenoid component did

Pfahler et al

J Shoulder Elbow Surg Volume 15, Number 2

159

Figure 3 Radiologic example of the progress of radiolucent lines on the left side in a woman aged 72 years.

Table X RLL score in relation to immobilization and rehabilitation Postoperative immobilization Yes

No

Rehabilitation center Yes

No

n 222 (32.3%) 466 (67.7%) 459 (67.5%) 221 (32.5%) RLL score of 0–5 64% 65% 59% 76% RLL score of 6–11 25% 23% 26% 20% RLL score ⱖ12 11% 12% 15% 4% Significance P ⫽ NS P ⬍ .001 NS, Nonsignificant.

not significantly influence the functional result of TSA unless the RLL score was greater than 12. Even with the development of radiolucent lines, the function of the TSA group was still better than in the HSA group (P ⬍ .001) (Table XI). The increase of radiolucent lines did not influence the results of TSA. The Constant score was 63 in

patients with increasing radiolucent lines and 65.5 in patients in whom the radiolucent lines were stable over time (P ⫽ .39). However, there was a significantly lower Constant score in patients with an RLL score greater than 12 and evidence of glenoid loosening (Table XII). Functional outcome in different pathologic entities

The comparison of functional outcome in TSA and HSA in the treatment of the more important pathologic entities demonstrated the superiority of TSA in most of these entities. TSA demonstrated better results than HSA in shoulders with primary osteoarthritis, rheumatoid arthritis, and sequelae of fractures. TSA also seemed to be better in shoulders with osteoarthritis after anterior instability. HSA yielded better results in those shoulders with osteonecrosis and eccentric osteoarthritis. The rate of glenoid loosening did not seem to be influenced by the pathologic entity. However, we found a higher risk of loosening in patients with rheumatoid arthritis and osteoarthritis after anterior instability. These latter pathologic entities demon-

160

Pfahler et al

J Shoulder Elbow Surg March/April 2006

Table XI Influence of radiolucent lines on Constant score in comparison with HSA Constant score

Adjusted score

395

56.3 18.57 4 95

74.3% 23.39 6 127

169

66.7 17.55 14 98

90.7% 24.57 20 140

65.4 16.67 13 96

87.3% 23.05 19 130

62 18.18 4 98

82.8% 24.43 6 140

n HSA Mean SD Minimum Maximum TSA without radiolucent lines Mean SD Minimum Maximum TSA with radiolucent lines Mean SD Minimum Maximum Total series Mean SD Minimum Maximum

434

998

Table XII Influence of RLL score on Constant score in comparison with HSA Constant score

Adjusted score

395

56.3 18.57 4 95

74.3% 23.39 6 127

349

68.1 16.0 14 98

91.4% 22.54 20 140

150

66.6 15.3 20 91

88.9% 21.31 29 124

62

54.7 17.25 15 84

73.9% 23.95 22 112

956

62.1 18.03 4 98

82.8% 24.21 6 140

n HSA Mean SD Minimum Maximum TSA: RLL score of 0–5 Mean SD Minimum Maximum TSA: RLL score of 6–11 Mean SD Minimum Maximum TSA: RLL score ⱖ12 Mean SD Minimum Maximum Total series Mean SD Minimum Maximum

strated the highest rates of the appearance of radiolucent lines and glenoid loosening (Tables XIII and XIV). DISCUSSION The material in this article comes from a multicenter study of TSA versus HSA in the treatment of shoulder pathology. This type of study is affected by some fairly typical disadvantages when one comes to analysis of its findings. We were able to evaluate the results of 1174 cases of shoulder arthroplasty using a consistent implant in all of the cases. We were comfortably able to draw some clear and important conclusions from our analysis. In general, the comparison of TSA and HSA for the treatment of chronic shoulder pathologic entities incontestably favors TSA with the use of a glenoid component. The two groups (469 HSAs and 705 TSAs) were comparable in age, sex, and preoperative Constant score. The rate of secondary complications was relatively high (11.5%); however, it was comparable in both groups. In contrast to other study results,2,4 the rate of revision (8.3%) was not principally because of the glenoid component. The rate of revision was significantly higher in the HSA group than in the TSA group. The functional and subjective results demonstrate the superiority of TSA over HSA in the treatment of chronic pathologic entities of the shoulder. TSA predominance is no longer questionable8; it has been

confirmed by each of the parameters evaluated by the Constant score (pain, activity, mobility, strength), independent of age or rotator cuff status. The proposition that HSA should be used in younger and more active patients1,8 or in patients with a rotator cuff tear2,3,9 can no longer be justified. One area of doubt remains, and it is the effect of time. The Constant score after TSA was significantly superior to HSA at the follow-up at 3 years and again at 5 years, but there was a slight decrease after 5 years in the TSA group, whereas the results in the HSA group were stable or demonstrated a slight improvement. Long-term follow-up studies should be able to evaluate this parameter further. Although we see the established superior results of TSA, the results of HSA are satisfying, with a distinct improvement of the Constant score (from 28.5%– 37.8% to 56.3%–74.3%). This would indicate the role of HSA for specific local or technical conditions. Analysis of the radiologic evaluation of the glenoid components revealed very important information, especially with regard to the analysis of the radiolucent lines. The reported occurrence rate of radiolucent lines varies between 30%7 and 92%.6 We were concerned about the rate (67.9%) in our series. It seems that radiolucent lines progress with time. We observed a significant increase in the RLL score over

Pfahler et al

J Shoulder Elbow Surg Volume 15, Number 2

Table XIII Comparison of HSA and TSA in different etiologies Etiology

Group

Constant score

HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

55.4 20.56 4 90

46.8 13.91 21 73 43.4 12.63 24 60 46.5 13.74 21 73

69.3 16.45 32 95 61.5 13.15 38 91 67.5 16.0 32 95 P ⫽ NS 56.5 15.04 34 90 64.1 21.71 13 89 60.3 18.82 13 90

Other etiologies HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

70 14.58 17 98

P ⫽ NS

Significance P ⫽ NS

HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

P ⫽ .001

63.9 16.04 26 93

68.9 15.03 17 98

Constant score

Arthrosis after anterior instability

57.6 17.39 15 85

Eccentric arthrosis HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

P ⫽ .03

51.8 16.49 15 81

55.7 17.26 15 85

Group

HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

65.4 16.6 26 90

Primary osteoarthritis HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

Etiology Osteonecrosis

49.8 20.49 4 88

Rheumatoid arthritis HSA Mean SD Minimum Maximum TSA Mean SD Minimum Maximum Total Mean SD Minimum Maximum

Table XIII Comparison of HSA and TSA in different etiologies (continued)

Significance P ⬍ .001

Fracture sequelae

161

P ⫽ NS

59 14.57 33 89 55 12.94 38 68 56.9 18.36 33 89

NS, Nonsignificant.

time (2.6 with follow-up ⬍3 years, 4.7 with follow-up between 3 and 5 years, and 6.4 with follow-up ⬎5 years). To our knowledge, this progression had not been documented in the literature before this study. The appearance of radiolucent lines may reflect a decrease in functional outcome with time. This fact is disputed by other investigators.2,6 Our results demonstrate that function remains stable as long as the RLL

162

Pfahler et al

J Shoulder Elbow Surg March/April 2006

Table XIV RLL score in different etiologies

Etiology Fracture sequelae Rheumatoid arthritis Primary osteoarthritis Eccentric arthrosis Osteonecrosis Arthrosis after anterior instability Other etiologies

RLL score of 0–5

RLL score of 6–11

RLL score >12

78% 61.7% 65% 83.3% 77.8%

10.1% 24.3% 26% 8.4% 22.2%

11.9% 14% 11% 8.3% 0%

50% 71.5%

28.6% 19%

21.4% 9.5%

score is less than 12. However, we see a significant decrease in function when the score exceeds 12 (continuous radiolucent line ⱖ2 mm). This confirms our impression that such severe radiolucent lines correspond with the clinical impression of loosening. This finding of progression of radiolucent lines over time and its corresponding clinical finding of decreased function represent our principal concern regarding the use of a cemented glenoid component; yet, in this study, the Constant score in the TSA group with a loose glenoid (54.7) remains comparable to that of HSA (56.3). The question remains: Will that potential effect, in time (ie, ⬎ 5 years) lead us to sacrifice the excellent short- and mid-term results of TSA for a primary hemiarthroplasty with a better long-term result? The development of an ideal glenoid component along with an optimal implantation technique to avoid the problem of loosening should be the main goal of further research. We were unable to discover any predictive factors for the appearance of radiolucent lines at the bone-cement interface of the glenoid implant. The absence of predictive factors has been confirmed by other authors.2,6,9 However, it seems that there may be a relationship of an increased inclination angle of the prosthesis, an extensive capsular release, and the use of an aggressive physiotherapy program in a rehabilitation center to the development of radiolucent lines. With the above reservations, we are of the opinion that the implantation technique is the most important method in establishing a stable glenoid component. This opinion was also given by Brems,2 who reported his findings of radiolucent lines (69%) in the immediate postoperative period. Franklin et al9 noted the adverse effect of rotator cuff deficiency in TSA. We found that the status of the rotator cuff had no influence on the appearance of radiolucent lines. Their frequency of occurrence was independent of the cuff being intact, torn, repaired, or not repaired. The decision to use a glenoid component in shoulder arthroplasty should be based on the preoperative

evaluation of the pathologic entity. Certainly, TSA has had a proven superiority in chronic pathologic conditions, such as primary osteoarthritis, rheumatoid arthritis, and fracture sequelae. On the other hand, it seems preferable to perform HSA in the treatment of osteonecrosis and instability arthritis, as well as those conditions that are a result of neurologic or inflammatory damage. Eccentric osteoarthritis and cuff arthropathy represent special entities. Even though there are only a minor number of cases, the results of TSA and HSA were comparable in these groups. The rate of occurrence of radiolucent lines was no higher than the rate with the more common pathologic chronic entities. This is in disagreement with the so-called rockinghorse phenomenon, described by Franklin et al.3 In conclusion, our study confirms the superiority of TSA over HSA in most chronic pathologic entities, demonstrated by short- and mid-term follow-up. We have been aware of the problems surrounding the glenoid component, and we have analyzed the radiographic findings regarding the rate and occurrence of radiolucent lines and their evolution, as well as clinical implications. Currently, they are not responsible for an elevated rate of revision.2 In the future, we need to maintain a longer follow-up period. We also need to research and develop a glenoid component with stable and reliable fixation along with an improved implantation technique. This study was only made possible by and with the collaboration of the International Multicenter Aequalis Study Group: P. Adeleine, D. Adrian, P. Ahrens, J. Armengol-Barallat, R. Aswad, C. Avidor, A.-M. Bartels, M. Basso, K. Benchik El Fegoun, B. Berghs, E. Bernard, F. Bertrand, P. Boileau, A. Boulahia, D. Boutens, U. Brunner, C. Bures, B. Burger, B. Candau, D. Chauveaux, J.-S. Coste, J. Dayez, J. De Beer, L. De Wilde, F. Deprey, P.-L. Docquier, F. Duparc, T. B. Edward, R. El- Abiad, R. J. H. Emery, G. Fama, L. Favard, J.-P. Franceschi, C. Garreau de Lubresse, D. Gazielly, Ch. Gerber, A. Godenèche, F. Gohlke, W. Harrop-Griffiths, R. Hertel, L. Hubert, D. Huguet, B. Jost, F. Kelberine, J.-F. Kempf, Y. Kerjean, S.G. Krishnan, L. Lafosse, M. Laurent, S. Lautman, J.-C. Le Huec, O. Léger, O. Lehmann, E. Lesprit, Ch. Levigne, J. Loehr, M. Loew, J.-P. Marchaland, J. Matsoukis, D. Molé, J.-M. Mora, C. Nerot, L. Neyton, L. Nové- Josserand, R. W. Nyffeler, D. Oudet, F. Pain, M. Pfahler, S. Reig, J.-C. Schaeffer, F. Schild, R. J. Sinnerton, F. Sirveaux, P. Staccini, W. Tabib, C. Torrens, Ch. Trojani, Ph. Valenti, Ch. Vandermaren, J. Vanhaecke, G. Versier, G. Walch, and W. J. Willems. REFERENCES

1. Bell SN, Gschwend N. Clinical experience with total arthroplasty and hemiarthroplasty of the shoulder using the Neer prosthesis. Int Orthop 1986;10:217-22. 2. Brems J. The glenoid component in total shoulder arthroplasty. J Shoulder Elbow Surg 1993;2:47-54. 3. Franklin JL, Barrett WP, Jackins SE, Matsen FA. Glenlid loosening in total shoulder arthroplasty. J Arthroplasty 1988;3:39-46.

J Shoulder Elbow Surg Volume 15, Number 2

4. Gartsman GM, Roddey TS, Hammerman SM. Shoulder artrhoplasty with or without resurfacing of the glenoid in patients who have osteoarthritis. J Bone Joint Surg Am 2000;82:26-34. 5. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fatty muscle degeneration in cuff ruptures. Pre- and postoperative evaluation by CT scan. Clin Orthop 1994;304:78-83. 6. Molé 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. Berlin: Springer Verlag; 1999. p. 163-71.

Pfahler et al

163

7. Neer CS. Replacement arthroplasty for glenohumeral osteoarthritis. J Bone Joint Surg Am 1974;56:1-13. 8. Rodosky MW, Bigliani LU. Indications for glenoid resurfacing in shoulder arthroplasty. J Shoulder Elbow Surg 1996;5:231-48. 9. Torchia ME, Cofield RH, Settergren CR. Total shoulder arthroplasty with the Neer prosthesis: long-term results. J Shoulder Elbow Surg 1997;6:495-505. 10. Wallace AL, Phillips RL, MacDougal GA, Walsh WR, Sonnabend DH. Resurfacing of the glenoid in total shoulder arthroplasty. J Bone Joint Surg Am 1999;81:510-18.