Periprosthetic joint infection after oncologic mega-implant reconstruction using a 24-hour cefazolin prophylaxis: a monocentric retrospective cohort study

Periprosthetic joint infection (PJI) is a major complication after oncologic reconstruction with mega-implants. The optimal duration and regimen of antibiotic prophylaxis remain debated, with wide variability in clinical practice. This study aimed to report infection rates and associated risk factors in patients undergoing oncologic mega-implant reconstruction using a standard 24-hour cefazolin-based prophylaxis, similar to conventional arthroplasty. We retrospectively reviewed 107 oncologic mega-implants implanted between 2015 and 2023 at a tertiary referral centre. All patients received a 24-hour cefazolin-based prophylaxis. PJI was defined according to the 2011 Musculoskeletal Infection Society criteria. Patient-, surgical-, tumour-, and peri-hospital- related variables were collected. Infection-free implant survival was assessed using Kaplan-Meier analysis, and univariate analyses identified factors associated with infection. Deep infection occurred in 22 of 107 mega-implants (20.6%), mostly within the first two postoperative years. Infection rates varied by anatomical site, with pelvic reconstructions showing the highest incidence (38.7%), compared with lower-limb (15.2%) and upper-limb (10.0%) reconstructions. Pelvic location (p=0.003), postoperative wound dehiscence (Henderson type 1B; p<0.001), and tumour extension into surrounding soft tissues (p=0.03) were significantly associated with infection. Operating time and hospital stay were longer in infected cases but strongly collinear with pelvic reconstruction. In this cohort, infection rates observed after oncologic mega-implant reconstruction in patients treated with a 24-hour cefazolin-based prophylaxis fell within the range reported in the existing literature. Pelvic reconstructions and compromised soft tissues were associated with higher infection risk, suggesting that a uniform prophylactic strategy may not be appropriate for all anatomical locations, particularly the pelvis.