Do High Doses of Multiple Antibiotics Loaded into Bone Cement Spacers Improve the Success Rate in Staphylococcal Periprosthetic Joint Infection When Rifampicin Cannot Be Employed?

Abstract

Rifampicin is one of the mainstays in treating staphylococcal prosthetic joint infection (PJI). However, discontinuation due to intolerance, drug interactions, and adverse events is common. Two-stage revision surgery remains the gold standard, with the number of revision arthroplasties steadily increasing. This study aims to evaluate the effectiveness and safety of a novel two-stage revision protocol for staphylococcal prosthetic joint infection (PJI) utilizing bone cement spacers loaded with multiple high doses of antibiotics. Additionally, it seeks to analyze outcomes in patients ineligible for rifampicin treatment. A retrospective review of 43 cases of staphylococcal hip and knee prosthetic joint infections (PJIs) from 2012 to 2020 was conducted. In all instances, a commercial cement containing 1 g of gentamicin and 1 g of clindamycin, augmented with 4 g of vancomycin and 2 g of ceftazidime, was employed to cast a spacer manually after thorough surgical debridement. We report an eradication rate of 82%, with no significant differences observed (p = 0.673) between patients treated with (84%, n = 19) and without rifampicin (79%, n = 24). There were no disparities in positive culture rates (7%), spacer replacement (18%), or survival analysis (p = 0.514) after an average follow-up of 68 months (range 10–147) in the absence of systemic toxicity and surgical complications superimposable to those previously reported. In conclusion, two-stage revision with local high doses of ceftazidime, vancomycin, gentamicin, and clindamycin demonstrates high effectiveness in treating staphylococcal PJIs. Notably, systemic rifampicin does not influence the outcomes. This protocol, with multiple high doses of antibiotics loaded into the bone cement spacer, is presented as a viable and safe alternative for patients unsuitable for rifampicin treatment.