Abstract
Background: Osseointegrated implants for patients with transfemoral amputations (TFAs) are a novel treatment under development, and prospective long-term evidence is lacking. The objectives were to determine patient -reported outcomes (PROs) and complications after ten years compared to before treatment and to compare the first five-year period with the later five-year period with regard to the outcomes. Methods: In a nonrandomized, prospective cohort study, patients with TFAs treated between 1999 and 2007 with the Osseointegrated Prosthesis for the Rehabilitation of Amputees (OPRA) system (n = 51) (28 men/23 women; mean age at amputation: 32 years old; mean age at treatment: 44 years old in a single university hospital were followed for ten years. PROs included the Questionnaire for Persons with a Transfemoral Amputation (Q-TFA, four scores 0-100) and the Short Form 36 Health Survey (SF-36, ten scores 0-100) and were answered before treatment and until the ten-year follow-up after treatment. Analyses of differences in PRO scores were conducted using Wilcoxon's signed rank test. The implant survival and revision-free rates with respect to adverse events (implant revision, mechanical complications, and deep infections) were presented as Kaplan-Meier graphs with 95% confidence intervals (CIs). The incidences of events per ten and five person-years were calculated. Spear -man's correlation analysis was used for analyses of associations between adverse events. Results: PROs showed statistically significant mean improvements between baseline and the ten-year follow-up with regard to all Q-TFA scores: the prosthetic use score (+36), prosthetic mobility score (+18), problem score (-28) and global score (+38) (all p < 0.001), and the SF-36 physical functioning score (+26, p < 0.001) and physical component score (+6, p < 0.01). No PROs showed a statistically significant deterioration. Over the ten years, 12 patients were lost (one lost to follow-up, one dropped out of the study, two died, and eight had implants removed (four before five years and four between five and ten years). At ten years, the revision-free survival rates were 83% (CI: 69%-91%), 65% (CI: 49%-77%) and 17% (CI: 7%-29%) for implant revision, deep infection and mechanical complications, respectively. Mechanical complications, 3.9 per 10 person-years (CI: 2.2-5.1) consti-tuted the most common serious adverse event and were more common during the last five years than during the first five years (p < 0.001). No significant difference in the incidence of deep infections was observed between the earlier and the later five-year periods: 0.3 per 5 person-years (CI: 0.1-0.5) vs. 0.3 per person-years (CI: 0.1-0.5) (p = 0.740). Correlation analyses between the earlier and later five years revealed a positive association between deep infections and implant removal (0.57, p < 0.001) and between mechanical complications and adverse events (0.65, p < 0.001). Conclusion: Improved PROs were demonstrated ten years after the introduction of a novel principle for bone anchorage of amputation prostheses. Nevertheless, an increasing rate of mechanical complications is of concern.