4.8 Surface replacement of the shoulder

The only part of any prosthesis that is of use to the patient is the new, shiny joint surface. The patient is unaware of whether this surface is fixed by a stem or cement or whether it is a pure surface replacement.

Development of the Copeland surface replacement arthroplasty (CSRA) for arthritis began in 1979, with the first clinical use of the prosthesis occurring 7 years later, in 1986. In 1993, hydroxyapatite coating was introduced to allow biologic fixation with bony ingrowth. Simple instruments were designed to allow anatomic placement of the humeral head based on identifying the centre of the sphere. Once this point has been identified, the prosthesis can be positioned to replicate the original anatomic bearing surface, including version, offset, and angulation.

The basic concept and design of surface replacement was to remove as little of the bone stock as possible and not to broach the humeral shaft. Initial fixation was by a cementless press fit single peg of taper fluted design using a hydroxyapatite coating for osseointegration (Figure 4.8.1). The head was made of cobalt-chrome, and the glenoid component of ultra-high-molecular-weight polyethylene (UHMWP). The radius of curvature of the head and glenoid were designed to conform with each other. Eight head sizes replicate the original anatomic bearing surface, including version, offset, and angulation. The surface replacement prosthesis has demonstrated clinical results at least equal to those of conventional stemmed prostheses.

Zippel in Germany implanted two surface replacements that were fixed by a transosseous screw but no follow-up is recorded for these cases. Steffee and Moore in the United States were implanting a small hip-resurfacing prosthesis into the shoulder and, in Sweden, in greater numbers, a surface replacement SCAN (Scandinavian) cup was being used as a cemented surface replacement. It was based on the same surface replacement philosophy that the CRSA was designed on.

The most frequent shoulder arthroplasty performed is a stemmed cemented total shoulder replacement. However, there are certain disadvantages to this type of design. The most common component to fail is the glenoid component with a reported incidence of radiolucent lines on radiographs as high as 30% and a significant rate of late loosening of the glenoid component over time. Glenoid failure may be associated with significant bone loss and cause major problems when revision surgery is undertaken, sometimes making the revision ‘impossible’.

Similar problems can occur with stemmed humeral components, albeit less frequently. The bone loss, however, can be very marked, and revision from a stemmed humeral implant can be extremely difficult, often requiring splitting the humerus to remove the stem and cement. This may result in perforation of the shaft and fractures. Fractures may also occur at the stem tip due to stress risers in the area (Figure 4.8.2), and are difficult to treat, especially in this elderly and infirm group of patients. In cementless stems, reaming of the bone can result in a fracture.

Deformities in the humeral shaft can occur with previous fractures and may pose a challenge getting a stemmed prosthesis past the deformed area. Similarly, it is not possible to use a stemmed implant in the presence of other metalware such as a nail or screw in the medullary cavity of the humerus. Although screws and nails can be removed, the stem of an elbow replacement coming up the shaft may preclude a stemmed shoulder replacement altogether.

An additional problem with stemmed fixation is that of ‘pilot error’. It has been shown that approximately 30% of the unsatisfactory results after shoulder replacements are due to component malposition and that, if version is incorrect by more than 15 degrees, this can lead to a painful shoulder. Incorrect placement can and does occur, and removal and repositioning of a stem is a major reconstruction problem. One always has to bear in mind what would happen if the joint should fail and thus plan a backup or revision procedure.

There is no evidence to indicate why the humeral stem should extend half the length of the humerus, sometimes with the cement tracking down to the elbow. On the other hand, one needs only to resurface the damaged surface of the humeral head to achieve comparable results and function to the stemmed prosthesis.

The commonest indication for surface replacement is arthritis, either degenerative or inflammatory. Other indications are mild avascular necrosis, instability arthropathy, post-trauma arthritis, post-infective arthritis, and arthritis secondary to glenoid dysplasia or epiphysis dysplasia. This prosthesis could be considered for Seebauer type I cuff tear arthropathy.

The best results are achieved in cases of osteoarthritis where the rotator cuff is intact. Conversely, suboptimal results are seen in individuals with cuff tear arthropathy and also in instances of post-traumatic arthritis. The surface replacement arthroplasty can be used in circumstances of moderate erosion of the humeral head, by bone grafting the defect. If there is more than 60% contact between the undersurface of the trial prosthesis and humeral head, after it has been milled, then it would be suitable for surface replacement. In other words, up to 40% of the humeral head surface area may be replaced by bone graft.

The contraindications for surface replacement arthroplasty include active infection, bone loss of the humeral head exceeding 40% of the surface, and acute fractures. Cases with an irreparable cuff tear would be suitable for the extended articular surface (EAS) replacement or a reverse shoulder arthroplasty.

Approximately 92% of our cases requiring shoulder arthroplasty receive a surface replacement. It is our opinion that surface replacement should be the standard prosthesis of choice for all cases, unless specifically contraindicated. The question now is not when to use a surface replacement, but what are the limited residual indications for a stemmed implant?

When deciding to do a hemiarthroplasty or a total arthroplasty, one must balance the possibility of longer-term glenoid wear and loosening in a total replacement with the possibility of late glenoid erosion in a hemiarthroplasty, necessitating conversion to a total replacement.

Our surface replacements had a 1% revision rate over 4 years. A survivorship analysis of the two groups in our series showed 97.1% survival of the hemireplacements and 81.7% survival of the total replacements. (Log rank test p = 0.0028). The total shoulders functioned slightly better, but revision rates were higher in the total shoulder group (18.3%) as compared to the hemiarthroplasty group (2.1%) (p <0.0001) (Figure 4.8.3).

We perform and recommend hemiarthroplasty now as a general rule, with the glenoid being replaced only infrequently, in view of the long-term results in our series.

The detailed surgical technique is outside the scope of this chapter, but the reader is referred to its detailed description in the article by Copeland and colleagues for further information.

Good clinical early results have been obtained with cups developed by Steffee and Moore (1984), and by Jónsson et al. (1986). Zippel in Germany implanted two surface replacements that were fixed by a transosseous screw but no follow-up is recorded for these cases. Rydholm and Sjögen from Sweden reported the results of the ‘SCAN cup’ in 1993 and 2003. Rydholm performed 84 SCAN cups, a hemispherical cemented cup, in 70 patients, and 72 cups in 59 patients were followed for 4.2 years (range, 1.5–9.9 years). The clinical results obtained showed 94% of the patients being pleased regarding pain relief and 82% reporting improved shoulder mobility. Shoulder function was significantly improved. Radiographs were analysed regarding the position of the cup, proximal migration of the humerus, and glenoid attrition during the follow-up period. Any change of the distance between the superior margin of the cup and the greater tuberosity and/or change of inclination of the prosthesis were regarded as signs of prosthetic loosening. With that definition, 25% of the cups were found to be loose at follow-up. Prosthetic loosening, however, had no bearing on the clinical result. Progressive proximal migration of the humerus in 38% of the shoulders and central attrition of the glenoid in 22% of the shoulders did not show any relationship to gain of mobility, pain relief, or functional ability. Of note, no central fixation peg was used for this cup. Long-term follow-up at 13 years included 54 cups in 46 patients (13 patients deceased, no revisions). Six cups had been revised 10 years (range, 5–16 years) after the index operation (four– for persistent pain, one for stiffness, and one for prosthetic loosening). Pain at rest on a 100-mm visual analogue scale was 15mm (range, 0–62mm) and pain on motion was 32mm (range, 0–85mm). Twenty-six (50%) could comb their hair (compared with 56% at first follow-up), 32 (62%) could wash their opposite axilla (90% at first follow-up), and 31 (60%) could reach behind (77% at first follow-up).

Alund et al. reported on 40 shoulder surface replacements for rheumatoid disease using the SCAN prosthesis. They reported one revision to total shoulder replacement, and 39 shoulders were followed up for a mean of 4.4 years (0.9–6.5 years). The median Constant score was 30 (15–79), mean proximal migration of the humerus 5.5mm (standard deviation [SD] 5.2mm) and mean glenoid erosion 2.6mm (SD 1.7mm). Proximal migration and glenoid erosion did not correlate with shoulder function or pain. Radiographic signs of loosening (changes in cup inclination combined with changes in cup distance above the greater tuberosity) occurred in one-quarter of the shoulders. At follow-up, 26 patients were satisfied with the procedure, despite poor shoulder function and radiographic deterioration.

Fink prospectively evaluated 45 Durom cups in 39 patients (30 women, 9 men) with rheumatoid disease. The average follow-up was 45.1 ± 11.6 months with a minimum of 36 months. Fifteen shoulders had an intact cuff (group A), 18 had a partial tearing or a repaired rotator cuff (group B), and 12 shoulders a massive cuff tear (group C). In group A rheumatic shoulders, the Constant score increased from 21.5 ± 9.6 points preoperatively to 66.1 ± 9.8 points at 36 months postoperatively; in shoulders of group B, from 19.6 ± 9.7 points preoperatively to 64.9 ± 9.6 points at 36 months postoperatively; and in shoulders of group C, from17.5 ± 8.7 points to 56.9 ± 9.8 points at the latest follow-up examination. All shoulders were pain-free at the latest examination. No complications, component loosening or changes of cup position were observed.

From 1993, the entire non-articular surface (implant–bone interface) of the glenoid and humeral components has been hydroxyapatite coated. The initial mechanical press-fit is thus followed later by a biological fix with bony ingrowth due to the hydroxyapatite coating. This is the current Mark III design. Between September 1993 and August 2002, 209 shoulders underwent surface replacement arthroplasty at our unit using the Mark III prosthesis with hydroxyapatite coating. Clinical and radiological outcome was assessed at an average duration of follow-up of 4.4 years. No evidence of radiolucency was seen in any humeral implant. Thomas et al. reported a 6.3% incidence of lucencies in their series using the Mark III implant. Asymptomatic non-progressive lucency of less than 2mm was seen in seven of the 29 glenoid components inserted, which did not require further treatment.

Six shoulders (2.8%) required revision surgery (one malposition of glenoid, two instability, and three painful arthroplasties). Using the Kaplan–Meier analysis, the probability that the implant would survive to the start of the tenth year after surgery was estimated to be 96.4%. The results of Mark III CSRA are comparable to conventional stemmed prostheses. There was no difference between hemiarthroplasty and total shoulder arthroplasty in terms of functional outcome. No hemiarthroplasty has been revised for component loosening.

Table 4.8.1 summarizes the results of surface replacement prostheses published thus far.

Revision surgery is greatly simplified having originally implanted a cementless surface replacement. At the time of revision of a surface replacement arthroplasty, the only bone lost is the bone that would have been removed had a stemmed prosthesis been used at the first operation. There is no need to remove a cemented stemmed prosthesis, which can be associated with loss of bone stock, perforation, and fracture of the humeral shaft. The preservation of bone facilitates revision to a stemmed prosthesis or to glenohumeral arthrodesis.

Surface replacement of the shoulder has been proven to be at least as successful as stemmed implants in the treatment of shoulder arthritis. The hydroxyapatite coating has been a major advance in reducing lucent lines and loosening. The bone-preserving nature of the implant allows it to be used in a most situations, including cases of deformity. If complications do occur, then they can be more easily treated, and the results of surface hemiarthroplasty appear to be comparable with stemmed hemiarthroplasty. The geometry and mechanics of the shoulder joint are now much better understood. It is no longer justifiable to continue with intramedullary (either cementless or cemented) fixation in a straightforward arthritic problem.

Future prostheses for the shoulder are likely to be of the bone-preserving nature. As materials improve, prosthesis wear will hopefully become less of a problem. Modern technology allows for more accurate preoperative planning. Computer assistance during surgery could translate this planning to a practical solution to optimize implant position and soft tissue balancing, which ultimately with improved materials should increase longevity of the prosthesis and improve function after shoulder arthroplasty. The next challenge facing us will probably be that of regenerating the surface!