Abstract
Background: Guidelines recommend empirical therapy with piperacillin/tazobactam (TZP) for spontaneous bacterial peritonitis (SBP) with low risk of multidrug-resistant organisms. Whether coverage of beta-lactam-resistant Gram-positive bacteria, such as ampicillin-resistant Enterococcus faecium, provides clinical benefit in such situations is unknown. Methods: In this observational study, we investigated the real-world effectiveness of empirical therapy with TZP monotherapy versus TZP plus linezolid (LZD) combination therapy in patients with SBP from two centers. Treatment failure, defined as the need to escalate antibiotic therapy due to in vitro resistance, lack of neutrophil decrease in ascitic fluid, or clinical decision, and 30-day survival were retrospectively assessed. Results: In the first cohort, 100 SBP episodes were empirically treated with TZP + LZD combination therapy (n = 50) or TZP monotherapy (n = 50). Treatment failure was recorded in 48% with TZP monotherapy compared with 16% with TZP + LZD combination therapy (p = 0.001), and this difference persisted after stratification for community-acquired versus hospital-acquired SBP. Although treatment failure after TZP therapy was associated with lower 30-day survival (56% vs. 82%; p = 0.04), 30-day survival with empirical TZP + LZD combination therapy was not different from empirical TZP monotherapy (Kaplan-Meier estimates 74% vs. 69%; p = 0.87). TZP concentrations in ascitic fluid were >32 mg/L in 94% samples after continuous administration. In a second cohort of 41 patients empirically treated with TZP, treatment failure was observed in 37%, which was also higher than in episodes treated with TZP + LZD in cohort 1 (p = 0.03). Conclusion: In this retrospective analysis, empirical TZP + LZD combination therapy for SBP was associated with fewer treatment failures without impact on short-term survival.
