Pressure Ulcer-Related Pelvic Osteomyelitis: Evaluation of a Two-Stage Surgical Strategy (Debridement, Negative Pressure Therapy and Flap Coverage) with Prolonged Antimicrobial Therapy

BACKGROUND: A two-stage surgical strategy (debridement-negative pressure therapy (NPT) and flap coverage) with prolonged antimicrobial therapy is usually proposed in pressure ulcer-related pelvic osteomyelitis but has not been widely evaluated. METHODS: Adult patients with pressure ulcer-related pelvic osteomyelitis treated by a two-stage surgical strategy were included in a retrospective cohort study. Determinants of superinfection (i.e.,, additional microbiological findings at reconstruction) and treatment failure were assessed using binary logistic regression and Kaplan–Meier curve analysis. RESULTS: Sixty-four pressure ulcer-related pelvic osteomyelitis in 61 patients (age, 47 [IQR 36–63]) were included. Osteomyelitis was mostly plurimicrobial (73%), with a predominance of S. aureus (47%), Enterobacteriaceae (44%), and anaerobes (44%). Flap coverage was performed after 7 (IQR 5–10) weeks of NPT, with 43 (68%) positive bone samples among which 39 (91%) were superinfections, associated with a high ASA score (OR, 5.8; P = 0.022). An increased prevalence of coagulase negative Staphylococci (P = 0.017) and Candida (P = 0.003) was observed at time of flap coverage. An ESBL Enterobacteriaceae was found in one (12%) patients, associated with fluoroquinolone consumption (OR, 32.4; P = 0.005). Treatment duration was as 20 (IQR 14–27) weeks, including 11 (IQR 8–15) after reconstruction. After a follow-up of 54 (IQR 27–102) weeks, 15 (23%) failures were observed, associated with previous pressure ulcer (OR, 5.7; P = 0.025) and Actinomyces infection (OR, 9.5; P = 0.027). CONCLUSION: Pressure ulcer-related pelvic osteomyelitis is a difficult-to-treat clinical condition, generating an important consumption of broad-spectrum antibiotics. Carbapenem should be reserved for ESBL at-risk patients only, including those with previous fluoroquinolone use. The uncorrelation between outcome and the debridement-to-reconstruction interval argue for a short sequence to limit the total duration of treatment. DISCLOSURES: T. Ferry, HERAEUS: Consultant, Speaker honorarium