Etiology and results of tumor endoprosthesis revision surgery in 64 patients.

December 1980 to December 1995, 278 patients underwent primary custom endoprosthesis replacements for neoplastic disease at the University of California, Los Angeles and have been followed up for a minimum of 2 years or to death. The endoprosthesis reconstruction failed in 64 patients, including 10 additional patients referred for revision of their replacements. Failure was defined as the complete removal of a prosthesis for any reason. The causes of failure were aseptic loosening (44%), fatigue fracture (16%), local recurrence (14%), infection (13%), and failure of the expansion mechanism (6%). Forty-eight of 64 failed endoprostheses were managed by endoprosthesis reconstruction with most being revised using the same type of prosthesis. Nine of these patients with failed replacements experienced a second failure and four went on to require an amputation. Aseptic loosening and mechanical failure accounted for most of the failures and they were revised successfully. Sixty percent of the infected cases were salvaged satisfactorily by endoprosthetic revision, whereas, 89% of the local recurrences resulted in amputation. Based on endoprosthesis survival the 7-year failure rates were 31% and 34% for primary and revision reconstructions, respectively. The functional results for the patients with endoprosthesis revisions either were better, unchanged, or on average only slightly lower than results of patients with a surviving endoprosthesis.