Aim of Prospective Study To assess the long-term survival of Souter-Strathclyde replacements and determine the causes of failure.

Materials and Methods: Between 1977–97, 445 Souter-Strathclyde replacements were undertaken by the presenting author in 321 patients, ranging in age from 25–81 years (mean 59). The Larsen X-ray grading at the time of surgery was Gd III-2%, Gd IV-43%, Gd V-55%. The standard humeral component was used in 76% and the long-stemmed model in 24% of cases, the standard ulna in 91%, and the long-stemmed in 9%.

Clinical and radiological assessments were carried out before surgery and at 6 months and one year after surgery and annually thereafter until the death of the patient or the revision of one or both components, the mean follow-up being 8.9 years (S.D. 5.6).

The survival statistics are based on Kaplan-Meier survival curves and Cox regression analysis applied to 3 Groups with differing end-points: 1) Revision of one or both components of the prosthesis 2) Ditto or the development of a complication, seriously threatening the survival of the arthroplasty and 3) As in (2) or definite evidence of progressive radiological loosening.

Results Very satisfactory pain relief was achieved in 90% of cases. Mean flexion improved from 129° to 144°, pronation from 42° to 59°, and supination from 43° to 62°. Once achieved, these results were well maintained. Mean extension was slightly reduced after surgery (47° to 52°)

For Group 1: the survivorship at 5, 10, 15 & 20 years was 95, 89, 82, & 81% respectively; for Group 2: 93, 87, 76 & 69%; and for Group 3: 91, 84, 72, & 68%. In Group 1, indications for revision or removal of one or both components were persistent dislocation (5), fracture (5), aseptic loosening complicated in some cases by fracture (20) and infection (14). In Group 2 the threatening complications included instability (2), un-united fracture of the ulna (4), infection (7), and clinical loosening (2)

Conclusions This method of elbow replacement offers very successful and durable pain relief and restoration of function in the adult rheumatoid patient. Survivorship in the first decade after surgery is very acceptable. Later, due to increasing disability and reliance on crutches, resulting from multi-joint involvement, more problems are likely to arise from aseptic loosening, fracture and infection.

Aim of study To determine the radiographic changes which can be regarded as indicative of probable eventual aseptic clinical loosening.

Methods 150 TERs performed in 121 adult rheumatoid patients (95 female) with a mean age of 59 years (34–81) were followed up for a mean of 8 years. All X-rays until the final review or revision of each TER were independently reviewed by the principal author. The cement/bone interface around each component was divided into Zones and any radiolucencies were graded within each Zone. This data was then analysed to determine in which zones and at which grading of severity, radiolucencies are of importance in predicting aseptic loosening.

Results Humeral Components Radiolucencies occurred quite frequently in Zones 1& 5. Where they became active and progressive, the usual pattern was for them to extend into Zones 2 and 4 and eventually into Zone 3. The incidence of lucencies of all grades of severity in Zones 1& 5 with standard humeral components was 55 and 53% respectively, in Zones 2& 4 29 and 33%, and Zone 3 18%. The significance of the radiolucencies also varied markedly in the different Zones. In Zones 1& 5, only 22% went on eventually to complete lines > Grade 3 severity in all Zones, 14% eventually requiring revision, while in Zones 2& 4, the corresponding figures were 36% with 22% requiring revision, and in Zone 3 60%, with 22% revision.

The results with the long-stemmed implants showed a very similar trend, the important Zones being 2B, 4B, and 3.

Ulnar Components Lucencies in Zones 1& 2 and in 3& 8 occurred in 90 & 73% of elbows respectively but are probably of little significance as only 15& 10% were later associated with the development of complete lines in other zones. The Zones of significance appear to be 5& 7, and especially Zone 6. Although lucencies were found in these Zones in only 35, 43 and 28% respectively, 49, 40 and 51% of these went on to form complete lucencies in all Zones, the eventual revision rate being 51, 21, and 32%.

Conclusions Many TERs demonstrate areas of lucency on follow-up radiographs but we would argue that it is only of importance in specific locations (humeral zone 3 for the standard implant and humeral Zones 2B, 4B and 3 for the long-stemmed implant and ulnar Zones 5& 7 and especially 6) and when it is of Grade 3 severity or more. Such cases require to be monitored very regularly and carefully so as to carry out revision at the optimal time should this eventually be required.