Purpose: The incidence of recurrent infections in patients following one or two stage revision for infected megaprostheses after resection of bone tumours was investigated. The difference between retaining at least one well fixed stem and a complete removal of the megaprosthesis during a two stage revision was also analysed.
Methods: 627 patients who experienced a replacement of a musculoskeletal tumour by megaprostheses were recorded. An infection occurred in 83 of 621 patients available for follow-up. 61 patients underwent one stage revision, and 16 patients two stage revision for the first revision surgery. In the entire study period, two stage revision was performed 32 times (first, second, and third revision).
Results: The cumulative incidence analysis showed a reinfection probability after one stage revision of 18% at one year, 30% at two years, 39% at five years, 46% at ten years, and 56% at 15 years. After two stage revision, a reinfection probability of 28% at two years, and 48% at five years was calculated. Cumulative incidence curves did not differ significantly (Gray's test; p = 0.51) between one and two stage revision (with and without complete removal of the stems). In two stage revision (n = 32), a statistically significant difference in infection rates between patients treated with complete removal of the megaprosthesis (n = 18) including anchorage stems and patients with at least one retained stem (n = 14) was shown (Fisher's exact test, p = 0.029).
Conclusion: Two stage revisions with complete removal of the megaprosthesis showed the best results among limb salvage procedures for the treatment of infected megaprosthesis.