Background Periprosthetic joint infections (PJI) caused by pathogens, for which no biofilm-active antibiotics are available, are often referred to as difficult-to-treat (DTT). However, it is unclear whether the outcome of DTT PJI is worse than those of non-DTT PJI. We evaluated the outcome of DTT and non-DTT PJI in a prospective cohort treated with a two-stage exchange according to a standardized algorithm. Methods Patients with hip and knee PJI from 2013 to 2015 were prospectively included and followed up for ≥ 2 years. DTT PJI was defined as growth of microorganism(s) resistant to all available biofilm-active antibiotics. The Kaplan–Meier survival analysis was used to compare the probability of infection-free survival between DTT and non-DTT PJI and the 95% confidence interval (95% CI) was calculated. Results Among 163 PJI, 30 (18.4%) were classified as DTT and 133 (81.6%) as non-DTT. At a mean follow-up of 33 months (range 24–48 months), the overall treatment success was 82.8%. The infection-free survival rate at 2 years was 80% (95% CI 61–90%) for DTT PJI and 84% (95% CI 76–89%) for non-DTT PJI (p = 0.61). The following mean values were longer in DTT PJI than in non-DTT PJI: hospital stay (45 vs. 28 days; p < 0.001), prosthesis-free interval (89 vs. 58 days; p < 0.001) and duration of antimicrobial treatment (151 vs. 117 days; p = 0.003). Conclusions The outcome of DTT and non-DTT PJI was similar (80–84%), however, at the cost of longer hospital stay, longer prosthesis-free interval and longer antimicrobial treatment. It remains unclear whether patients undergoing two-stage exchange with a long interval need biofilm-active antibiotics. Further studies need to evaluate the outcome in patients treated with biofilm-active antibiotics undergoing short vs. long interval.

Background: There are currently no recommendations to direct the optimal diagnosis and treatment of fungal osteoarticular infections, including prosthetic joint infections and osteomyelitis. Active agents (fluconazole; amphotericin B) are regularly applied per os or intravenously. Other drugs such as voriconazole are used less frequently, especially locally. Voriconazole is less toxic and has promising results. Local antifungal medication during primary surgical treatment has been investigated by implanting an impregnated PMMA cement spacer using intra-articular powder or by daily intra-articular lavage. The admixed dosages are rarely based on characteristic values and microbiological and mechanical data. The purpose of this in vitro study is to investigate the mechanical stability and efficacy of antifungal-admixed PMMA with admixed voriconazole at low and high concentrations. Methods: Mechanical properties (ISO 5833 and DIN 53435) as well as efficacy with inhibition zone tests with two Candida spp. were investigated. We tested three separate cement bodies at each measuring time (n = 3) Results: Mixing high dosages of voriconazole causes white specks on inhomogeneous cement surfaces. ISO compression, ISO bending, and DIN impact were significantly reduced, and ISO bending modulus increased. There was a high efficacy against C. albicans with low and high voriconazole concentrations. Against C. glabrata, a high concentration of voriconazole was significantly more efficient than a dose at a low concentration. Conclusions: Mixing voriconazole powder with PMMA (Polymethylmethacrylate) powder homogeneously is not easy because of the high amount of dry voriconazole in the powder formulation. Adding voriconazole (a powder for infusion solutions) has a high impact on its mechanical properties. Efficacy is already good at low concentrations.

Amphotericin B is used for local delivery from polymethylmethacrylate to treat fungal prosthetic joint infections. The optimal amphotericin B formulation and the influence of different poragens in the bone cements are unknown. To investigate the necessary amount of amphotericin B in the bone cement to prevent Candida biofilm several amphotericin B formulations were studied: non-liposomal and liposomal with or without poragen gentamicin. For the non-liposomal formulation, standard bile salt, the sodium deoxycholate, was used and additionally N-methyl-D-glucamine/palmitate was applied. The activity of the released amphotericin B was tested against C. albicans, C. glabrata, C. parapsilosis and C. krusei biofilms with application of the isothermal calorimeter and standard microbiological methods. Compressive strength was measured before and after antifungal elution from the cements. There is less aggregated N-methyl-D-glucamine/palmitate amphotericin B released but its antifungal activity is equivalent with the deoxycholate amphotericin B. The minimum quantity of antifungal preventing the Candida biofilm formation is 12.5 mg in gram of polymer powder for both non-liposomal formulations. The addition of gentamicin reduced the release of sodium deoxycholate amphotericin B. Gentamicin can be added to N-methyl-D-glucamine/palmitate amphotericin B in order to boost the antifungal release. When using liposomal amphotericin B more drug is released. All amphotericin B formulations were active against Candida biofilms. Although compressive strength slightly decreased, the obtained values were above the level of strength recommended for the implant fixation. The finding of this work might be beneficial for the treatment of the prosthetic joint infections caused by Candida spp.

Biofilm-active antibiotics are suggested to improve the outcome in periprosthetic joint infection (PJI). However, the type, dose and duration of antibiotic treatment is rarely specified and their impact on outcomes is unknown. In this prospective cohort study, the infection and functional outcome were compared in 131 patients with knee PJI treated with or without biofilm-active antibiotics. The infection and functional outcome were evaluated by the Kaplan–Meier survival method to estimate the probability of infection-free survival; comparison between subgroups was performed by log-rank test. The influence of variables on the survival probability was analysed using univariate and multivariate Cox proportional-hazards regression models. Functional outcome was evaluated by pain intensity and the Knee injury and Osteoarthritis Outcome Score (KOOS). Among the 131 patients, 55 (42%) were treated with biofilm-active antibiotics and 76 (58%) were treated with non-biofilm-active antibiotics. The median follow-up period was 3.7 years (range, 2.0–7.6 years), and the infection-free survival probability was 74% (95% CI 61–85%) after 1 year and 56% (95% CI 47–66%) after 2 years. Infection-free survival after 1 year was better for patients who received biofilm-active antibiotics compared with those who did not (83% vs. 70%; P = 0.040) and remained superior after 2 years (67% vs. 48%; P = 0.038). In addition, biofilm-active antibiotic treatment was associated with lower pain intensity (P = 0.006) and higher KOOS on all five subscales. In patients with knee PJI, biofilm-active antibiotic therapy was associated with better infection outcome, lower pain intensity and better joint function.