Risk factors of uncontrolled periprosthetic knee joint infection after two-stage reimplantation

Abstract

Background: Periprosthetic infection after total knee arthroplasty is a challenging problem, and physicians should identify risk factors to decrease recurrence. However, risk factors for reinfection with two-stage reimplantation have not yet been well established. The purpose of this study was to assess treatment outcomes of subsequent two-stage knee reimplantation and identify risk factors for uncontrolled periprosthetic knee joint infections.

Methods: We retrospectively reviewed 70 knees diagnosed with a periprosthetic knee joint infection treated with two-stage reimplantation between September 2011 and October 2016 at our institution. Patients in the controlled infection group (group C) required no further medication or surgical treatment within 2 years after reimplantation. Patients in the uncontrolled infection group (group U) displayed symptoms of active infection after resection arthroplasty or were reinfected after two-stage reimplantation. We compared group C and group U, and analyzed potential risk factors for uncontrolled prosthetic joint infection (PJI).

Results: Of 70 knees included in this analysis, 53 (75.7%) were clinically deemed free from infection at the latest follow-up. The remaining 17 knees (24.3%) required additional surgical procedures after two-stage reimplantation. Demographics were not statistically significantly different between the two groups. Wound complications were statistically more frequent in group U (p = 0.030). Pre-reimplantation C-reactive protein (CRP) was statistically different between groups C and U (0.44 and 1.70, respectively, p = 0.025). Among the cultured microorganisms, fungus species were statistically more frequently detected in group U compared with group C (p = 0.031).

Conclusions: The reinfection rate of our two-stage reimplantation protocol was 24.3% in the included cases. Wound complications, higher pre-reimplantation CRP levels, and fungus species were statistically more common in group U compared with group C. Our findings will help in counseling patients and physicians to understand that additional caution may be required when treating PJI if the aforementioned risk factors are present.