María Dolores del Toro, Matteo Ferrari, Rafael San Juan, Cédric Arvieux, Benjamin M. Clark, Joshua S. Davis, Tristan Ferry, Trisha Peel, Jaime Lora-Tamayo, Eric Senneville, Louis Bernard, Efthymia Giannitsioti, Dace Vigante, Antonio Ramos, José Antonio Iribarren, Dolors Rodríguez-Pardo, Melchor Riera, L Guio, N. Benito, Daniëlle Neut, Rihard Trebše, Karina O'Connell, Craig A Aboltins, Michel Dupon, Alfredo Jover-Sáenz, H K Li, Peter F. M. Choong, Alberto Bahamonde, Josu Baraia-Etxaburu, Thomas Gottlieb, Jaime Esteban, M Jose G. Pais, Mauro José Costa Salles, Kaisa Huotari, Severine Ansart, Alex Soriano, Martin Clauss, Parham Sendi, Nathalie Asseray, Alba Ribera, Nina Gorisek Miksic, Mar Sánchez-Somolinos, Gábor Skaliczki, Lucía Gómez, Javier Ariza, Valérie Zeller, Juan Pablo Horcajada, Julián Palomino, Javier Cobo, Marta Fernandez-Sampedro, Alicia Rico, Ulrike Dapunt, Gwenael Le Moal, Ilker Uçkay, José Maria Barbero, and Werner Zimmerli
Background.: Streptococci are not an infrequent cause of periprosthetic joint infection (PJI). Management by debridement, antibiotics, and implant retention (DAIR) is thought to produce a good prognosis, but little is known about the real likelihood of success.Methods.: A retrospective, observational, multicenter, international study was performed during 2003-2012. Eligible patients had a streptococcal PJI that was managed with DAIR. The primary endpoint was failure, defined as death related to infection, relapse/persistence of infection, or the need for salvage therapy.Results.: Overall, 462 cases were included (median age 72 years, 50% men). The most frequent species was Streptococcus agalactiae (34%), and 52% of all cases were hematogenous. Antibiotic treatment was primarily using β-lactams, and 37% of patients received rifampin. Outcomes were evaluable in 444 patients: failure occurred in 187 (42.1%; 95% confidence interval, 37.5%-46.7%) after a median of 62 days from debridement; patients without failure were followed up for a median of 802 days. Independent predictors (hazard ratios) of failure were rheumatoid arthritis (2.36), late post-surgical infection (2.20), and bacteremia (1.69). Independent predictors of success were exchange of removable components (0.60), early use of rifampin (0.98 per day of treatment within the first 30 days), and long treatments (≥21 days) with β-lactams, either as monotherapy (0.48) or in combination with rifampin (0.34).Conclusions.: This is the largest series to our knowledge of streptococcal PJI managed by DAIR, showing a worse prognosis than previously reported. The beneficial effects of exchanging the removable components and of β-lactams are confirmed and maybe also a potential benefit from adding rifampin.