Abstract

Shoulder resurfacing is an attractive concept be-cause it preserves, rather than removes, the humer-al head. With a less invasive surgery there is the promise of better function as well as a less difficult revision if it is later needed. We asked if shoulder resurfacing procedures improved function and had satisfactory implant survival over the long term. 61 patients (74 shoulders) with shoulder resurfac-ing procedures were followed for a minimum of 20 years. Of these there were 41 total resurfacing procedures and 33 hemiarthroplasties. Seven addi-tional patients were lost to follow-up. The average age at the time of surgery was 58 years. 84% of pa-tients were followed until their death. The humer-al component we used consisted of a cup for the surface replacement with a short central peg. The humeral component was placed either with or with-out cement. The glenoid was resurfaced with a cemented polyethylene or polyurethane component. 95% of patients were satisfied with their result. There were no periprosthetic fractures, disloca-tions, or infections. Two humeral components were revised to stemmed prostheses-one for loosen-ing and one for unexplained pain. One humeral

resurfacing was revised from a cementless to a cemented resurfacing prosthesis. Twelve cemented polyethylene glenoid prostheses had radiolucen-cies but only three had symptoms requiring revision surgery. Three of four polyurethane glenoid pros-theses showed severe wear but none were loose or required revision surgery. Shoulder resurfacing is a successful procedure for most patients over the long term.

Introduction

Most patients with end stage arthritis of the shoul-der are treated by total shoulder replacement. 5,16,24 This involves removal of the humeral head and in-sertion of a stem into the medullary space of the humerus. The results are usually quite satisfac-tory, however, there remain some concerns. Stem supported devices can cause a stress riser or stress shielding effect.17,27 This can lead to periprosthetic fracture or prosthetic loosening.10 Another concern is “pilot error” with insertion of a stemmed device. If the version of a shoulder device is incorrect by just 15° a painful shoulder will be the result.7

Shoulder resurfacing is attractive because it pre-serves the humeral head. The retained humeral head continues to provide proprioceptive feedback. The resurfaced shoulder is stable and capable of an excellent range of motion. Since the humeral head is retained it is easier to assure correct version, offset and inclination of the prosthesis during surgery. With a limited amount of implanted material, infection if it occurs can be dealt with more easily. Because the hu-merus is not decapitated, shoulder resurfacing is also a less invasive surgery. There is less pain and blood loss and an easier rehabilitation from resurfacing

Shoulder Resurfacing

MINIMUM 20 YEAR FOLLOW-UP

James W. Pritchett MD

The author certifies that he has no commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest with the submitted article.

The author certifies that his institution has approved the human protocol for this investigation and the investigation was conducted in conformity with ethical principles of research and that informed consent was obtained.

Correspondence to: James W. Pritchett MD 901 Boren Ave. Seattle, WA 98104 206 779 2590 Fax 206 726 6166 email: Bonerecon@aol.com

ORIGINAL ARTICLE

1

as compared to total shoulder replacement.6,11,12

The first total shoulder resurfacing procedure was performed by Dr. Charles O. Townley in 1958 us-ing a metal humeral component and a polyurethane glenoid (Fig 1). Other investigators later performed shoulder resurfacing procedures using small hip re-surfacing prostheses.8,23 Zippel was the first to pub-lish a paper describing the use of a metallic surface replacement of the humeral head.28 With the advent of stem supported total shoulder prostheses shoul-der resurfacing was abandoned with the exception of a few surgeons.

Young patients, in particular, want to keep their bone, enjoy an active lifestyle and don’t want to worry about having a failed intramedullary stem supported shoulder prosthesis. Recently cemented and cementless hemi and total shoulder resurfac-ing prostheses have again become popular. The re-sults of these procedures at midterm follow-up have been favorable.6,12,12,21 It has been suggested, but not proven, that humeral resurfacing prostheses can loosen and fail more commonly over time than stem supported prostheses because they have less fixa-tion area.2,8,9 Also, there is fear that humeral neck fractures, or humeral head collapse may occur in a similar fashion as occurs with hip resurfacing.19,25

Our questions are: What are the functional results from shoulder resurfacing surgery? What is the

long term survival of shoulder resurfacing prosthe-ses? Finally, what are the complications of shoulder resurfacing procedures over the long term?

Materials and Methods

61 patients (74 shoulders) who underwent shoul-der resurfacing surgery using the Total Articular Replacement Arthroplasty (TARA) prosthesis from 1958 to 1989 were studied (Fig 2). None of the patients had prior implant arthroplasty, however, 18 patients had undergone prior surgery to treat a fracture, dislocation, impingement syndrome, or a rotator cuff tear. The diagnosis was osteoarthritis in 37 shoulders (50%), posttraumatic arthritis in 20 shoulders (27%), inflammatory arthritis in 12 shoul-ders (16 %), osteonecrosis in five shoulders (7%)

The indication for surgery in all patients was severe pain associated with limitations in function. All of patients had been treated extensively but unsuccess-fully with nonoperative methods. All patients were unwilling to continue to accept their symptoms and opted for surgical treatment even when advised there was no certainty of outcome.

Exclusion criteria were prior infection of the shoulder, insufficient bone stock, and neurologic injuries. There were 41 total resurfacing procedures and 33 hemiar-throplasties. Total arthroplasties were performed when there was significant glenoid erosion present.

2

Fig. 1 Total shoulder resurfacing with a polyurethane glenoid prosthesis.

Fig. 2 Total Articular Replacement Arthroplasty with a polyethylene glenoid prosthesis.

3

The patient population consisted of 32 women and 29 men. The mean age was 58 years (range, 32 to 71 years). All patients were followed until their death or a minimum of 20 years. 84% of patients had died by the time of final follow-up. The mean age at the time of death was 81 years (range, 59 to 92 years) and the mean follow-up was 28 years (range 20 to 41) (Table 1). Institutional Review Board Approval was obtained for this study.

Surgery was performed by two surgeons JWP and COT. A deltopectoral approach was used for all procedures. The subscapularis tendon was opened and the shoulder was dislocated anteriorly by external rotation. The labrum was debrided as nec-essary and any contractures were released. The humeral head was measured and prepared with a mill-ing device to accept the humeral cup. The humeral prosthesis with the best head coverage was chosen and was placed in anatomic version. An anatomic repair of the subscapularis was performed.

A standard sling was used for 6 weeks postopera-tively. Home exercises were started on the first postoperative day and consisted of passive circum-duction and pendulums as well as active range of motion exercises such as saws (back and forth mo-tion of the arm in the coronal plan with a flexed elbow). External rotation was allowed to within

30° of that obtained during the surgery, after the subscapularis repair. Patients participated in for-mal physical therapy or a therapist directed home program for an additional 6 weeks. No limitations were placed on patients’ activity after 3 months.Three of the earliest humeral components were placed without cement using a bolt through the lateral humeral cortex (Fig 3). Thirty additionalhumeral components were placed without a boltand without cement and 37 humeral com-ponents were cemented in place using polymethylmethacrylate (Simplex, Howmedica, Rutherford, NJ.) Four humeral components were cemented in place using polyurethane cement (Ostamer, William S. Merrell Co., Cincinnati, OH). The glenoid component, when used, was always ce-mented in place. Polyurethane cement was used for the polyurethane glenoid components but the poly-ethylene glenoid prostheses were cemented in place using polymethymethacrylate. The polyethylene glenoid prostheses became available in 1971 andhad a central keel (Howmedica, Rutherford NJ, Depuy Orthopaedics,Inc., Warsaw, IN) (Fig 2).

The polyurethane polymer was prepared by mixing the prepolymer with resin and a catalyst at the time of surgery and shaping it in situ or on the back table to the humeral prosthesis. When used the polyurethane acted as both the anchoring cement

Fig. 3 Early hemiresurfacing revised to a total resurfacing prosthesis because of humeral component loosening and glenoid wear.

Table 1. Survivorship among original patients treated shoulder joint resurfacing.

Years Since Surgery Number (%) Mean Age at death of follow- up (years; range)

Patients who had died 51(84) 81 (59-92) Less than 5 years 3 5-9 years 4 10-19 years 11 20-30 years 17 Longer than 30 years 16Patients alive at follow-up 10(16) 75 (54-94)Survival periods 20-30 years 8 30-40 years 1 40 years 1

4

and the glenoid bearing surface. The humeral prostheses were all made of cobalt chromium (Depuy Orthopaedics, Inc., Warsaw IN; Howmedica, Rutherford, NJ).

Patients were followed prospectively and asked to return at 1 year, 2 years, 5 years and every 5 years thereafter. The functional assessments were made at 2 years. Patients were asked about their activity and specifically whether they were limited. Any lim-itations were specific to that particular patient’s life goals. Clinical examinations to final follow-up were available for 25 patients (41 %). Patients were asked about the need for additional surgery on their shoul-der after their resurfacing procedure. If surgery had been performed, the patient was asked to provide information about that procedure. Information about the patient’s shoulder function was obtained from the family for patients who had died.

Postoperative radiographs were assessed for the presence or absence of lucent lines and their width in relation to time. Definite loosening was defined as a change in the position of the component and probable loosening was defined as an unchanged po-sition but progressive radiolucent lines involving all parts of a component. Because radiographs were not standardized (they were made by different technicians using different techniques on different

equipment over many years) no specific radiographic measurements were made.

The subjective result after surgery was rated by the pa-tient as “very satisfied”, “satisfied,” “somewhat satis-fied”, “somewhat disappointed,” or “very disappointed.”

Results

Most patients experienced little or no pain and only two experienced severe pain. Of the patients assessed for postsurgical activity, 39% participated in strenuous athletics or work and only two were dissatisfied with their function. Ninety-two percent of patients were not limited in their activities (Table 2). 31 patients (36 shoulders) were very satisfied with the result of their shoulder resurfacing procedure. 22 patients (27 shoulders) were satisfied four pa-tients (7 shoulders) were somewhat satisfied and four patients (4 shoulders) patients were somewhat or very disappointed. We were unable to determine a difference in satisfaction or function based on whether the patient had a hemi or total resurfacing procedure.

96% (71 of 74) of the humeral prostheses remained in place without symptoms or need for revision at

Table 2. Functional results of shoulder joint resurfacing

Pain Number (%) Assessed 2 years after procedures

No pain 63 (85) Slight pain 8 (11) Moderate pain 1 (2) Severe pain 1 (2)Function: postsurgical activity Highly active 29 (39) Strenuous sports or job Active and no limitations necessary 38 (51) Moderately active 5 (7) Inactive 2 (3)Patient satisfaction Satisfied with outcome 70 (95) Dissatisfied with outcome 4 (5) 2 patients were dissatisfied because of limited motion or weakness

2 patients were dissatisfied because of pain

5

the time of death or final follow-up. Of the 28 gle-noid prostheses available for final review, twelve polyethylene glenoid significant radiolucencies but only four had symptoms (Fig 4). Three of four polyurethane components showed wear through but none loosened or required revision. All four of these patients continued to report a satisfactory outcome. After the polyurethane wore away these shoulders seemed to function as a hemiarthroplasty.

There were no infections, subluxations, or peripros-thetic fractures. There were two temporary nerve palsies. One involved the entire brachial plexus and one involved the axillary nerve only.

There were seven revision surgeries. One of the ear-liest cementless hemiresurfacing components was revised for loosening and glenoid wear (Fig 3). It was revised to a cemented total resurfacing prosthe-sis. Another hemiarthroplasty was revised to a stem supported hemiarthroplasty for pain. This implant was not loose at the time of revision surgery. One resurfacing humeral prosthesis was placed on an insufficient humeral head and this prosthesis loos-ened, requiring revision to a stemmed prosthesis (Fig. 5). One hemiarthroplasty was revised to a total resurfacing prosthesis by the addition of a glenoid. Two loose polyethylene glenoid prostheses were

removed and one was revised.

Discussion

Shoulder resurfacing in our patients achieved its goal of reducing pain, and increasing function. The satisfaction results in this study are as good as or better than those reported using a conventional stemmed humeral prosthesis. 5,15,16,24 The survival of the resurfacing humeral prosthesis is very satis-factory but we found like other authors using total shoulder replacement prostheses that the glenoid component remains vulnerable to wear and loosen-ing over time.

We saw few complications in our patients with shoul-der resurfacing. Subluxation and dislocation did not occur and we attribute this to correct sizing and version of the humeral component since the implant was placed on the natural humeral head. Instability is a reported complication with total shoulder replacement.16,18,20,24 Also, there were no periprosthetic fractures. The lack of stress shielding with resurfacing is possibly the reason. There were two instances of humeral prosthesis loosening requiring revision. One involved a first generation cementless prosthesis attached with a bolt through the lateral humeral cortex. This design was

Fig. 4 Asymptomatic glenoid loosening. Fig. 5 Revision of a failed humeral resurfacing to a stemmed prosthesis because of an insufficient head following trauma.

6

discontinued after three cases. The other humeral prosthesis that loosened was in a case where the hu-meral head was deficient. There were no infections.

There are limitations of this study. The case num-bers are small but this is consistent with most other reports of shoulder prostheses. Nevertheless, small case numbers suggest caution in interpreting the in-cidence of uncommon complications such as peri-prosthetic fracture and infection. This study inves-tigates a prosthesis that was in evolution as it was being used. Both cemented and cementless humeral fixation was used but we are unable to draw conclu-sions on this variable. There were no autopsy re-trievals or pathologic specimens.

Other limitations are that the two surgeons involved performed many of the clinical and radiographic analyses. Complete follow-up data (including ra-diographs) were not available on all of the patients so questionnaires were relied upon for some of the information. The primary outcome, prosthesis sur-vival, is known for all but seven patients who were lost to follow-up. Because this is a single patient series, there is no comparison to stemmed humeral devices.

This is primarily a report on prosthesis survival and functional improvement after shoulder resurfacing. Data on range of motion, specific radiographic mea-surements, and strength are not presented because of lack of standardization. We did not use specific outcome measurement instruments as none of the commonly used scales were available during most of the time our patients were initially treated.3,4,14 We recognize that, although most patients experi-enced substantial improvement and performed at their individual desired level after surgery, there is no way to standardize this data. We did not find any reason to limit the activity of a patient based on the findings in this study.

The indications for shoulder resurfacing are the same as those for the use of any shoulder prosthesis. Shoulder resurfacing can be used in the treatment of

osteoarthritis, rheumatoid arthritis, post dislocation arthritis, and osteonecrosis.6,11,12 There are limita-tions to resurfacing. If there is not enough humeral head remaining then a stemmed device is needed (Fig 5). We find that 70% of the humeral head re-maining is enough.

One of the advantages of resurfacing is that no bone stock is lost. If there were an infection or instability, the joint is amenable to arthrodesis. This is not the case after stemmed shoulder prostheses.Total shoul-der resurfacing is more technically demanding than total shoulder replacement as access for preparation of the glenoid is more difficult without removing the humeral head.

Shoulder resurfacing may be more successful than hip resurfacing.19,25 There are several differences between shoulder and hip resurfacing. The applied loads are much less on the shoulder. The shoulder joint has a much reduced offset compared to the hip. Shoulder prostheses have a greater radius of curvature. The shoulder joint is much less con-strained compared to the hip. Soft tissues may be more critical to function for the shoulder as com-pared to the hip.

Contemporary shoulder resurfacing began again in 1986 with a cementless hydroxyapatite prosthesis. The revision rate for this humeral prosthesis is 1-2%. When glenoid resurfacing is performed loosening occurs in up to 10% of patients. 95% of patient re-port a satisfactory functional outcome.6,11,12,21

The favorable results of shoulder resurfacing are no doubt partially due to the ability to obtain correct version, inclination, offset and size. These factors are easily judged and relate directly to the anatomy. For most surgeons, the skills of shoulder replace-ment are learned from fracture care of the occa-sional shoulder replacement procedure. In fracture and replacement cases the humeral head has been removed and version is estimated either indirectly or from guides during surgery.

7

Most shoulder prostheses are placed by surgeons who do only a few procedures a year. 30 percent of unsatisfactory results after shoulder replacement are due to component malposition.7

Resurfacing arthroplasty is a different procedure than shoulder replacement. With stemmed implants the position of the prosthetic humeral head, to a sig-nificant degree, is determined by the relationship of the stem to the humeral shaft.27 Some of the newer modular total shoulder prostheses attempt to more closely duplicate natural shoulder anatomy but long term results on these implants are not yet available. For resurfacing, the prosthetic head is placed on the natural humeral head (Fig. 6). In patients with severe arthritis and glenoid erosion the humeral head may have become medialized and correcting this fully with a stem supported total shoulder replacement may result in enough lateralization that stiffness and pain occurs postoperatively. With resurfacing there is usually less lateralization (Fig 7).

The long-term debate over whether a total or hemi-arthroplasty procedure is best is not solved with our

work. In stemmed total shoulder replacement the rate of conversion from a hemiarthroplasty to a to-tal shoulder replacement is up to 12%. The conver-sion rate may be less with resurfacing. We attribute at least some of the success of our hemiresurfacing procedures to releasing soft tissue contractures and restoring the normal humeral contours. This likely improves the kinematics of shoulder function. Since the version and spacing of the joint are improved less subsequent glenoid pain and wear may result. This may suggest performing a hemiarthroplasty in many cases.5, 6,11,12

Total shoulder resurfacing is a valid procedure and like total shoulder replacement will result in excel-lent pain relief and function. The problem of gle-noid loosening remains. The difficulties with total shoulder resurfacing are primarily on the glenoid side. Well performed humeral resurfacing rarely fails over time. This is true with or without the use of cement.

Polyurethane as a bearing surface may significantly improve the future results of glenoid resurfacing. Compared to polyethylene, polyurethane has lower frictional properties. Polyurethane surfaces are eas-ier to lubricate because they are hydrophilic while polyethylene is hydrophobic. Polyurethane wear debris does not cause osteolysis and bone loss. The modulus of elasticity of polyurethane can be made similar to cartilage, unlike polyethylene which is

Fig. 6 Axillary view of humeral resurfacing prosthesis showing anatomic positioning.

Fig. 7 20 year result of resurfacing prosthesis performed for post dislocation arthritis. The center of rotation suggests correct lateralization.

8

about 70 times stiffer. Polyurethane has been se-lected for several orthopedic applications such as the artificial cervical disk. The polyurethane used in our early cases was very crude compared to what is available today.1,13,22,26

The options available for glenoid fixation are im-proving. For cementless applications we are cur-rently using trabecular metal backed polyethylene glenoid prostheses as we continue to work on our polyurethane bearing surface.

Shoulder resurfacing is an attractive option for treating shoulder arthritis. It works well, is bone conserving and avoids the concerns of a stem sup-ported prosthesis. There are more salvage options if there is a failure.

References

1. Carbone A, Howie DW, McGee M, Field J, Pearcy M, Smith N, Jones E. Aging performance of a compliant bearing acetabular prosthesis in an ovine hip arthroplasty model. J Arthroplasty. 2006;21:899-906.2. Egund N, Jonsson E, Lidgren L, Kelly I, Petterson H. Computed Tomography of Humeral Cup Arthroplasties. Acta Radiol 1987;28S71-73.3. Constant CR. Age related recovery of shoulder function after injury (thesis) Cork Ireland: University College;1986.4. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop Relat Res. 1987; 214:160-164.5. Gartsman GM, Roddey TS, Hammerman SM. Shoulder arthroplasty with or without resurfacing of the glenoid in patients who have osteoarthritis. J Bone Joint Surg Am 2000; 82:26-34.6. Furst M, Fink B, Ruther W. The DUROM cup humeral surface replacement in patients with rheumatoid arthritis. J Bone Joint Surg Am. 2007;89:1756-62.7. Hasan SS, Leith JM, Campbell B, Kapil R, Smith KL, Matsen FA 3rd. Characteristics of unsatisfactory shoulder arthroplasties. J Shoulder Elbow Surg. 2002;11:431-41.8. Johnson E, Egund N, Kelly I, Rydholm U, Lidgren L. Cup arthroplasty of the rheumatoid shoulder. Acta Orthop Scan. 1986;57:542-6.9. Jonsson E, Lidgren L, Mjoberg B, Rydholm U, Selvic. Humeral cup fixation in rheumatoid shoulders. Acta Orthop Scand 1990;61:116-11710. Kumar S, Sperling JW. Haidukewych GH, Cofield RH. Periprosthetic humeral fractures after shoulder arthroplasty. J Bone Joint Surg Am 2004;86:680-911. Levy O, Copeland SA. Cementless surface replacement arthroplasty (Copeland CSRA) for osteoarthritis of the shoulder. J Shoulder Elbow Surg. 2004;86:512-8.

12. Levy O, Funk L, Sforza G, Copeland SA. Copeland surface replacement arthroplasty of the shoulder in rheumatoid arthritis. J Bone Joint Surg Am. 2004;86:512-8.13. Mandarino MP, Salvatore JE. Polyurethane polymer-its use in fractured and diseased bones. Am J Surg. 1959;97:442-449.14. Murray MP, Gore DR, Gardner GM, Molinger LA. Shoulder motion and muscle strength of normal men and women in two age groups. Clin Orthop Relat Res. 1985;192:268-273.15. Neer CS. Articular replacement for the humeral head. J Bone Joint Surg Am. 1955;37:215-228.16. Neer CS, Watson KC, Stanton FJ, Recent experience in total shoulder replacement. J Bone Joint Surg Am. 1982;64:319-337.17. Orr TE, Carter DR. Stress analysis of joint arthroplasty in the proximal humerus. J Orthop Res. 1985; 3:360-371.18. Pritchett JW. Inferior subluxation of the humeral head after trauma or surgery. J Shoulder Elbow Surg. 1997;6:356-9.19. Pritchett JW. Curved-stem hip resurfacing-minimum 20-year followup. Clin Orthop Relat Res. 2008;466:1177-1185.20. Pritchett JW, Clark JM. Prosthetic replacement for chronic unreduced dislocations of the shoulder. Clin Orthop Relat Res. 1987;216:89-93.21. Rydholm U. Surface replacement of the humeral head in rheumatoid shoulder. J Shoulder Elbow Surg. 1993;2:286-95.22. Scholes SC. Compliant layer acetabular cups: Friction testing of a range of materials and designs for a new generation of prostheses that mimics the natural joint. Proc Inst Mech Eng 2006; 220:583-596.23. Steffee AD, Moore RW. Hemi-resurfacing arthroplasty of the shoulder. Contemp Orthop 1984;9:51-59.24. Torchia ME, Cofield RH, Settergren CR. Total shoulder arthroplasty with the Neer Prosthesis: long-term results. J Shoulder Elbow Surg. 1997;6:495-505.25. Townley CO. Hemi and total articular replacement arthroplasty of the hip with a fixed femoral cup. Orthop Clin North Am. 1982;13:869-894.26. Treharne RW, Greene AH. The case for the use of polycarbonate- urethane in orthopaedic implants. Med-Tech, Spring 2008;1:18-22.27. Walker PS. Some bioengineering considerations of prosthetic replacement for the glenohumeral joint. Symposium on the total joint replacement of the upper extremity. ed. AE Inglish, AAOS. 1982;25-32. CV Mosby. St Louis.28. Zippel J.(Dislocation-proof shoulder prosthesis model BME). Z Orthop Ihre Grenzgeb. 1975;113:454-457. German