 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1999 The Journal of Bone and Joint Surgery, Inc. Resurfacing of the Glenoid in Total Shoulder Arthroplasty. A Comparison, at a Mean of Five Years, of Prostheses Inserted with and without Cement*  , SYDNEY, AUSTRALIAInvestigation performed at the Prince of Wales Hospital, University of New South Wales, Sydney
Background: Clinically evident loosening of a glenoid component inserted with cement in total shoulder arthroplasty is infrequent, but radiographic changes that indicate loosening at the implant-bone interface are common and have been associated with functional limitation. We compared the results of total shoulder arthroplasties in which the glenoid implant had been inserted with cement with those of arthroplasties in which a bone-ingrowth glenoid implant had been inserted without cement. Methods: The results of eighty-six consecutive total shoulder arthroplasties, performed by the same surgeon, were retrospectively reviewed between four and seven years after the operation. Fifty-eight shoulders in which the primary glenoid implant was in situ were assessed with use of the Simple Shoulder Test and Short Form-36 questionnaires, clinical examination, and fluoroscopic imaging of the glenoid implant-bone interface. Thirty-two of the glenoid components had been fixed with cement and twenty-six, without cement. Results: Complications occurred in 16 percent (fourteen) of the eighty-six shoulders, and 9 percent (eight) of the shoulders needed a revision operation. None of the revisions were done because of loosening of the glenoid component. Five of the eight revisions involved implants that had been inserted without cement. Three of these implants were revised because of early instability and two, because the polyethylene component had separated from the metal tray of the glenoid implant. With the numbers available, we could not detect any significant differences between the groups with respect to pain, range of motion, function of the shoulder, or general health. Radiographic analysis demonstrated a high level of interobserver agreement (kappa = 0.89). Radiolucent lines were observed after 41 percent (thirteen) of the thirty-two arthroplasties performed with cement compared with 23 percent (six) of the twenty-six arthroplasties performed without cement. The proportion of implants classified as probably loose was approximately three times greater in the group in which cement had been used. Eccentric wear of the posterior rim of the metal tray and focal osteolysis under the metal tray were observed in the group in which the component had been inserted without cement; these findings may indicate a potential for progression of radiographic loosening with increased durations of follow-up. Conclusions: We concluded that, despite the higher rate of early complications, the intermediate-term outcomes of arthroplasties in which the glenoid implant is inserted without cement are comparable with those of arthroplasties with cementing of the glenoid component.
Total shoulder arthroplasty has been established as an effective procedure for the treatment of end-stage degenerative and inflammatory arthropathy of the glenohumeral joint, with consistently good relief of pain reported in most series17,19. Early fully constrained prostheses21 have been largely superseded by unconstrained designs that depend to a great extent on the restoration of soft-tissue balance for stability. Neer's original press-fit humeral component, designed to be inserted without cement, was first implanted more than forty-five years ago18. Although the arthroplastic technique was later modified to incorporate the use of methylmethacrylate cement, the results of both techniques (with and without cement) were very satisfactory. Neer et al. reported radiographic evidence of loosening of only 2 percent (one) of fifty components inserted without cement and less than 1 percent (one) of 144 humeral components inserted with cement19, and these findings have been confirmed by other investigators8,20. Fixation of the glenoid component with use of methacrylate cement has been less reliable, particularly in the presence of instability or a massive tear of the rotator cuff2,12,26. In some reports, both all-polyethylene and metal-backed components implanted with cement demonstrated radiolucent lines at the interface with the host bone in most shoulders1,7, but the importance of these findings is not yet certain. Some authors have regarded radiolucent lines as a consequence of the initial cementing technique, with no relation to clinical loosening19. Generally, revision because of clinical loosening has been infrequent5. However, even with modern cementing techniques, at least one-fifth of these radiographic lines appear to progress over time7,20. Other studies have shown that progression of radiographic changes was associated with pain and decreased functional capacity6,8. In a long-term study, Torchia et al. reported radiographic evidence of definite loosening of 44 percent (thirty-nine) of eighty-nine glenoid components that had been inserted with cement23; these findings led other investigators to question the value of glenoid resurfacing27. Concerns about these radiographic changes, the long-term efficacy of cement, and the potential need for revision of cemented components have prompted the development of implants with porous surfaces, which allow ingrowth of bone, in an effort to achieve durable fixation with direct apposition of viable bone-forming cells to the implant. Although experience with the use of these implants is limited, some investigators have suggested that they are associated with a lower prevalence of radiolucent lines at the implant-bone interface than are implants fixed with cement11,16,26. In order to test the hypothesis that fixation of the glenoid component without cement results in less evidence of loosening and a superior clinical outcome compared with fixation of the component with cement, we compared the results of the two types of fixation in a consecutive series of total shoulder arthroplasties, performed by the same surgeon. The minimum duration of follow-up was four years after implantation.
Selection of the Patients The glenoid component was fixed with cement in forty-three patients (forty-seven arthroplasties). At the time of the review (at a mean of seventy-one months [range, forty-six to ninety months] postoperatively), seven patients had died of unrelated causes, two were not able to return because of another medical condition, three had moved overseas, and three could not be located. Thus, twenty-eight patients (thirty-two arthroplasties) were available for clinical and radiographic review. The glenoid implants included twenty Neer-II all-polyethylene components, five Neer-II metal-backed polyethylene components (Kirschner Medical, Timonium, Maryland), four Cofield all-polyethylene components (Smith and Nephew Richards, Memphis, Tennessee), and three Global all-polyethylene components (DePuy, Warsaw, Indiana). The humeral implants included thirteen Neer-II stems (one of which was inserted with cement), sixteen Cofield stems (one of which was inserted with cement), and three Global stems (all of which were inserted without cement). A revision procedure was done because of humeral malposition, to retrieve a loose fragment of cement that was causing impingement, and to fix a postoperative fracture of the greater tuberosity in one patient each. As none of these procedures was performed because of a complication associated with the glenoid component, these patients were not excluded from the review. The glenoid component was inserted without cement in thirty-eight patients (thirty-nine arthroplasties). At the time of the review (at a mean of fifty-six months [range, forty-seven to sixty-four months] postoperatively), one patient had died perioperatively, one was unable to return because of poor general health, two had moved overseas, and four could not be located, leaving thirty patients (thirty-one arthroplasties) who were recruited into the study. All of these patients had primary fixation of a Cofield metal-backed glenoid component without cement. The Cofield humeral component was used in all arthroplasties, and it was inserted without cement in all but one shoulder, in which cement was necessary for additional fixation. Five patients were subsequently excluded from the review. Two of them had a revision operation because of separation of the polyethylene liner from the metal tray of the glenoid component at two and four years after implantation, and these procedures are described in another study24. The other three patients had a revision of both the humeral and the glenoid component because of postoperative instability, which was related to incorrect version of the glenoid component, within three months after the primary procedure. Therefore, twenty-five patients (twenty-six arthroplasties) in whom the primary glenoid component was in situ were available for clinical and radiographic review.
Assessment of the Outcome After we had received the completed questionnaires, the patients returned for a clinical examination, which was conducted by the same observer (A. L. W.). The active and passive ranges of motion of both the involved and the contralateral shoulder were assessed with use of a manual goniometer (the measurements were approximated to the nearest 5 degrees), with the patient seated. Active internal rotation was scored with use of a scale of anatomical landmarks that the patient could reach. According to this scale, 1 indicated no motion; 2, the greater trochanter; 3, the posterior iliac crest; 4, the natal cleft; 5, the sacrum; 6, between the fourth and fifth lumbar vertebrae; 7, between the first and third lumbar vertebrae; 8, between the ninth and twelfth thoracic vertebrae; 9, between the fifth and eighth thoracic vertebrae; and 10, between the first and fourth thoracic vertebrae. Cross-body adduction was quantified by measurement of the residual distance from the anterolateral tip of the acromion to the olecranon and was expressed in centimeters. Impingement pain, crepitus, and tenderness were noted. Finally, radiographic examination of the glenoid component was performed with use of a standardized fluoroscopic technique. A single anteroposterior radiograph of the glenoid was made with the patient standing, such that the x-ray beam was perpendicular to the plane of the implant-bone interface, which was determined with use of either the wire marker embedded in the polyethylene or the metal backing of the implant, as described by Kelleher et al.15. Each radiograph was examined by a radiologist (R. L. P.) and an orthopaedic shoulder surgeon (G. A. MacD.) who were blinded to the identity of the patient, the results of the clinical examination, and the responses to the questionnaires. The observers noted the type of glenoid component as well as other features, such as a fracture of a screw or the cement mantle, wear of the metal rim, bead-shedding, cement fragments, and localized osteolysis or osteopenia. Specifically, the location of radiolucent lines between the glenoid implant (or cement) and bone were classified into five zones (Fig. 1) and measured with use of a digital micrometer (Digimatic; Mitutoyo, Tokyo, Japan), which had a resolution of 0.01 millimeter. The two independent observers were requested to give an opinion about the stability of the glenoid component on the basis of specific radiographic criteria1,10,15 (Table I).
Statistical Analysis
The two groups were comparable with respect to the distribution of age, gender, and diagnosis. Most patients were women, and more than twice as many patients had osteoarthritis as had rheumatoid arthritis (Table II). Because we changed our choice of glenoid component from one inserted with cement to one inserted without cement, the mean duration of follow-up of the group that had a component inserted with cement was longer (seventy-one compared with fifty-six months).
Complications
Subjective Assessment of the Outcome
The subjective evaluations with use of the visual-analog scales, which ranged from a score of 1 point (minimum) to 10 points (maximum) for each category, revealed no differences between the groups. Because the data were positively skewed, the median values are given. The score for pain at rest was 1 point for the group that had the component inserted with cement and 2 points for the group that had the component inserted without cement. Both groups had a score of 2 points for pain at night, a score of 3 points for the effect of the condition of the shoulder on the ability to use the arm at work or for recreational activities, and a score of 2 points for the effect of the condition of the shoulder on the quality of life. Each variable of the general health questionnaire was scored and weighted as previously described25. The resulting eight variables were plotted against normative data for the population of the United States25 (Fig. 2). With the numbers available, we could not detect any differences between the study groups, but the values for both groups were different from the normative values for individuals in the general population who were more than fifty-five years old, particularly with respect to the physical function, role (physical), and bodily pain categories. For example, the mean score (and standard deviation) for physical function was 55.8 ± 26.5 points for the group that had the component inserted with cement and 48.2 ± 26.8 points for the group that had the component inserted without cement compared with 76.2 ± 26.3 points for the normal age-matched population. However, the variation in all samples was so great that these differences were not found to be significant. In general, the pattern of variable scores in all categories was more closely matched to that of existing normative data for patients who were more than fifty-five years old and had unspecified osteoarthritis as a comorbidity25.
Clinical Examination
Radiographic Findings Of the thirty-two components that had been inserted with cement, four (13 percent) were found to be probably loose, two (6 percent) were possibly loose, and two were the subject of disagreement between the observers. Both of the discrepancies were between the categories of stable and possibly loose, and one observer (the orthopaedic surgeon) considered both implants to be stable. Of the twenty-six prostheses that had been inserted without cement, one (4 percent) was definitely loose, one was probably loose, and four (15 percent) were possibly loose. There was no disagreement between the observers with regard to the stability of the components in this group. Furthermore, no relationship could be found between the observers' opinions regarding the radiographic findings and either the pattern of responses to the Simple Shoulder Test questionnaire or the active and passive ranges of motion in any of the planes examined. Although the agreement between the observers was satisfactory, the analysis of the radiolucent zones was performed only by the radiologist in order to control for any inherent observer bias of the orthopaedic surgeon toward a particular type of glenoid fixation. The distribution of radiolucent lines was not found to be localized to any specific zone or zones or to be different between the two groups. Radiolucent lines were observed in thirteen (41 percent) of the thirty-two shoulders that had the glenoid component fixed with cement. However, the lines were at least one millimeter wide in only eight shoulders (25 percent) and were complete in only two of the eight (Fig. 3). Five glenoid components (16 percent) were associated with a radiolucent line that was less than one millimeter wide, and only one of those lines was complete. Sequestrated cement fragments were observed in nine (28 percent) of the thirty-two shoulders.
Radiolucent lines were seen in six (23 percent) of the twenty-six shoulders in which the component had been inserted without cement (p = 0.1), and all of the lines were at least one millimeter wide. The radiolucent line around two components (8 percent) was complete. One component that had been inserted without cement was deemed definitely loose (Fig. 4); the metal tray was clearly tilted with respect to the glenoid bone, and the most proximal screw was fractured. Several other radiographic findings that provided grounds for concern about the fate of the implant were noted in this group. In one shoulder, eccentric wear of the posterior rim of the metal tray, which was associated with superior subluxation of the humeral component, was noted on an oblique radiograph. This pattern of wear previously has been associated with separation of the polyethylene liner11,24. In another shoulder, focal osteolysis was seen adjacent to the collar of a screw in the implant, which was otherwise deemed stable according to our criteria (Fig. 5). This finding may reflect a localized response to polyethylene wear debris at the implant-bone interface, which could potentially worsen with time. Osteolysis was also seen surrounding the screws and the central column of implants that were classified as probably or possibly loose, but we could not determine whether the radiolucent lines were due to particulate debris-mediated resorption or to gross mechanical loosening because of inadequate initial fixation. Fracture of a screw was seen in three (12 percent) of the shoulders, although the appearance of the implant-bone interface in two shoulders was consistent with that of a stable implant. It is likely that fatigue failure due to subtle micromotion accounted for these screw fractures, although the screws may have broken at the time of insertion.
Insertion, without cement, of a glenoid component with bone-ingrowth capability in total shoulder arthroplasty is not a new concept, but the technique has not yet gained wide acceptance. To our knowledge, we are the first to specifically and directly compare, with use of standardized, validated outcome measures and independent radiographic analysis, the intermediate-term results of total shoulder arthroplasties, performed by the same surgeon, with insertion of the glenoid component with and without cement. In addition to conventional clinical examination, we used both a joint-specific tool and a generic measure of general health as previously recommended for assessment of the outcome of an intervention3. In a previous comparison, with a duration of follow-up similar to that in the present study, Weiss et al. reviewed the results of arthroplasties with a Neer prosthesis inserted with cement and those of arthroplasties with an English-McNab prosthesis inserted without cement26. They found no difference between the two modes of fixation with respect to improvement in the range of motion and functional capacity. However, only nine of the forty-two arthroplasties were performed without cement and therefore only limited conclusions can be made with respect to the fate of such components. Kelleher et al. compared thirty-seven glenoid components inserted with cement and sixty-three glenoid implants inserted without cement in order to develop a standardized radiographic technique, but no clinical or functional analysis was performed15. In our study, the major variable was the method of fixation of the glenoid component. The results demonstrated that fixation without cement provides an outcome that is comparable with that of fixation with cement in terms of relief of pain, subjective functional capacity, range of motion, and effect on general health. These findings are consistent with those of previous studies of total shoulder arthroplasties performed with cement2,8,14,17 and without cement10,11. The overall rate of complications was also similar to those in previous reports9,11,27. The rate of revision operations because of complications was almost twice as high in the group that had the component inserted without cement than it was in the group that had fixation with cement, but only the two revisions due to separation of the polyethylene liner from the glenoid component could be directly attributed to the selection of the implant designed to be inserted without cement. However, it is possible that after the change from cement fixation to fixation without cement, our initial operative inexperience with the new prosthesis may have contributed at least in part to the increased problems with instability. Given that revision of a glenoid component because of overt clinical loosening is uncommon7, there has been considerable controversy in the literature with regard to radiolucent lines at the implant-bone interface. It has been assumed that, because some of these radiolucent lines are present immediately postoperatively, interposition of blood and debris may be responsible, and it also seems likely that radiolucent lines on later radiographs represent nonmineralized fibrous tissue5. The prevalence of progressive radiolucent lines has varied, but conventional plain radiographic imaging has been used on a serial basis in most studies8,14,19,20,22,23. In a recent cadaveric study, with assessments by four observers, large variations were noted in native glenoid version, but, more importantly, the position of the glenoid was found to be a critical factor in the determination of the width of a simulated radiolucent line13. These observations bring into question the effectiveness of plain radiographs for documentation of the presence or progression of interface changes in relation to actual loosening4. On the basis of these findings and those in the earlier report by Kelleher et al.15, we elected to use a one-plane fluoroscopic technique that allowed correct positioning of the patient for a true perpendicular view of the implant-bone interface. The overall prevalance of radiolucent lines associated with the implants that had been fixed with cement was higher than that for the implants that had been fixed without cement. However, the prevalence of radiolucent lines that were at least one millimeter wide was the same for both groups. The criteria used to determine radiographic stability resulted in very good reliability between observers. The relatively small discrepancy between opinions was probably related to the interpretation of radiolucent lines less than one millimeter wide in the group that had the prosthesis inserted with cement, since the cement boundary was less radiographically distinct than the edge of the metal tray in the group that had the prosthesis inserted without cement. Early reports of fixation of glenoid bone-ingrowth implants without cement noted a low prevalence of loosening11,16,26. In one report26 comparing total shoulder arthroplasties performed with and without cement, radiolucent lines of one to two millimeters in width were present adjacent to fifteen (36 percent) of forty-two cemented glenoid components. Only one component was obviously loose, as demonstrated by medial migration of the cement mass. A complete radiolucent line was seen surrounding only one of nine glenoid components fixed without cement. Longer-term experience has not been as consistent, although the results have still been favorable for implants inserted without cement compared with those for implants inserted with cement15. With use of standardized fluoroscopic imaging, radiolucent lines were seen adjacent to nine of eleven Neer-I implants and seventeen (65 percent) of twenty-six Neer-II implants inserted with cement, and such lines were seen adjacent to twenty-six (41 percent) of sixty-three Cofield implants inserted without cement; complete lines were seen around one, four, and two implants, respectively15. The duration of follow-up was not specified for each type of component. In a review of Neer-II glenoid components implanted with use of improved cementing techniques, radiolucent lines were seen around more than 50 percent of the glenoid component in two (6 percent) of thirty-five shoulders, and the prevalence of lines did not appear to be affected by the presence of metal backing20. Overall, these results suggest that the method of fixation may not be the primary determinant of the development of radiolucent lines. Other factors, such as the preparation of the glenoid surface, the magnitude of stresses transmitted across the prosthetic articulation relative to the glenoid bone support, or the access of polyethylene wear debris to the implant-bone interface, ultimately may be more important. At present, there is no clear information concerning the optimum thickness of polyethylene or other bearing surfaces, rates of particle generation, and biological responses to polyethylene to guide decision-making. Moreover, the clinical importance of radiolucent lines remains uncertain, given the similar clinical outcomes in our patients. The findings of the present study must be considered in the context of several serious limitations. First, the design was retrospective and consequently a proportion of the patients could not be adequately assessed, although the rate of attrition was similar for the two groups. The assumption was made that the patients in whom the primary implant remained in situ had outcomes similar to those included in the study cohort. Second, radiographic analysis was conducted at one point in time only and therefore progressive changes could not be determined. However, as none of the patients had immediate fluoroscopic imaging after the operation, comparison with previous radiographs was not possible and comparison with nonstandardized conventional radiographs would not have been valid. Third, although several designs of glenoid component were inserted with cement, the assessment was applied to the group as a whole. No attempt was made to stratify the results according to a particular feature, such as metal backing, which might have affected stresses at the implant-bone interface. Fourth, as with the radiographic analysis, the tools used for the measurement of functional outcome were not used in the early postoperative period and therefore changes for an individual patient over time could not be established. Finally, given the difference of approximately fifteen months in the mean duration of follow-up between the two groups, there is a possibility that additional problems may arise over time in the group that had the prosthesis inserted without cement, which may affect the comparison with the group that had the prosthesis inserted with cement. Despite these limitations, we concluded that there are grounds for cautious optimism regarding bone-ingrowth fixation, although longer-term comparative studies are necessary to define more precisely the role for this type of implant. On the basis of our clinical and radiographic findings, resurfacing of the glenoid with a component inserted without cement seems to be a reasonable alternative to fixation with cement, as long as it is understood that the problem of changes at the implant-bone interface and their long-term implications has not yet been solved and that new problems (such as separation of the modular parts of the component and breakage of screws) have been introduced. We agree with Cofield11 that the generation of polyethylene wear debris may well be the factor limiting the survival of contemporary designs of total shoulder implants, as has been the experience with other prosthetic joints.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |