Abstract
EPRs are the treatment of choice following resection of tumours. These have been used for 39 years in our institution. There has been concern regarding the long term survival of endoprosthesis; this study investigates the fate of the reconstruction.
Methods
Between 1966 and 1995, 3716 patients were seen with a suspected neoplasm and 776 patients underwent EPRs. Patients receiving growing endoprostheses were excluded from the study as they invariably require revision, leaving 667 replacements. Insufficient data was available in 6 cases, leaving 661 patients in the study group. Information was reviewed concerning the diagnosis, survival of implant and patient, subsequent surgery, complications and functional outcome. Kaplan-Meier survival analysis was used for implant survival with end points defined as revision for mechanical failure (aseptic loosening, implant fracture, instability, avascular necrosis, periprosthetic fracture, pain and stiffness) and revision for any cause (infection, local recurrence and mechanical failure).
Results
Mean age at diagnosis was 34 years. Overall patient survival was 52.7% at 10 years and 45.7% at 20 years. The mean follow-up for all patients was 9 years, and for those patients who survived their original disease, the mean follow-up was 15 years. 227 (34%) patients underwent revision surgery, 75 patients for infection (33%), 36 patients for locally recurrent disease (16%) and mechanical failure in 116 patients (51%). With revision for mechanical failure as the end-point, implant survival was 75% at 10 years and 52% at 20 years. With revision any cause as an end-point implant survival was 58% at 10 years and 38% at 20 years. Overall limb salvage was maintained in 91% of patients at 10 years from reconstruction and 79% at 20 years. There was a significant difference between survival of implant between implantation sites, with the proximal humeral implant survival being the best and tibial reconstructions being the worst. The MTSS functional score was available on 151 patients, with a mean score of 25/30 (83%) at last follow-up visit.
Discussion
Our results are comparable with other series of long term follow-up and justifies their continued use.
