Limb Salvage After Failed Infected Total Knee Arthroplasty: A Comparison of Cement-Based Artificial Arthrodesis and External Fixator Arthrodesis

Background: Persistent or recurrent periprosthetic joint infection after total knee arthroplasty is a devastating complication. When revision fails, knee arthrodesis remains an important limb-salvage option. Cement-based artificial arthrodesis may provide an alternative when extensive bone loss or soft-tissue compromise limits conventional fusion, but comparative evidence is scarce. Methods: This retrospective exploratory cohort included 36 patients whose index salvage procedures were performed between January 2014 and December 2023 after failed treatment of periprosthetic knee infection. Sixteen patients underwent external fixator arthrodesis, and 20 underwent cement-based artificial arthrodesis. Postoperative SF-12 and SF-36 scores, ambulation, limb shortening, extensor mechanism status, microbiological findings, reinfection, and complications requiring surgery were evaluated. Results: The groups differed at baseline: preoperative ESR and CRP were higher in the cement-based group, whereas extensor mechanism insufficiency was more frequent in the external fixator group (81.3% vs. 20%; p = 0.001). No statistically significant between-group differences were detected in most SF-12 and SF-36 domains; Energy/Fatigue was higher in the external fixator group (p = 0.011), an isolated exploratory finding. Mean postoperative shortening was numerically greater after external fixation (4.94 ± 3.66 vs. 3.15 ± 0.49 cm; p = 0.07). Complications requiring surgical intervention occurred in 31.2% and 25% of patients, respectively. Reinfection requiring additional surgery occurred in 0/16 external fixator patients and 1/20 cement-based patients. Conclusions: Cement-based artificial arthrodesis may represent a practical limb salvage option in selected patients with failed infected TKA, particularly when direct bone apposition is difficult. Because treatment allocation was non-random, baseline characteristics were imbalanced, and the cohort was small, so the findings should be interpreted as exploratory and not as evidence of equivalence or superiority.