Does Piperacillin-Tazobactam Minimum Inhibitory Concentration for Pseudomonas aeruginosa Predict Outcome of Pseudomonal Bacteremia?

Background: Pseudomonal bacteremia has a high mortality rate.  The Clinical and Laboratory Standards Institute (CLSI) breakpoint of piperacillin/tazobactam (PT) for P aeruginosa (PA) is higher compared to other Enterobacteriaecae [minimum inhibitory concentration (MIC) ≤64 mcg/ml vs. ≤16 mcg/ml].  Insufficient PT dosing for bacteremia has been linked to poor outcomes.

Objective: To review outcomes in a cohort of pseudomonal bacteremic patients, with an emphasis on those with PA bacteremia with higher susceptible PT MICs (eg, 32-64 mcg/ml). 

Methods: Retrospective review of all adult patients admitted to our two main campuses (bed size >1000) with PA bacteremia from 1/07 to 7/09 who received ≥48 hours of antibiotic.  The main outcomes were in-hospital mortality and/or clinical/microbiologic failure in 4 subcohorts including 1 split by PT MICs.

Results: A total of 152 patients with positive blood cultures for PA were identified of which 119 were analyzed (33 were excluded due to mortality <48 hours of therapy).  The mean age was 65 years with 51% male.  Median length of stay to bacteremia was 9 days, range 0-189 (39 were bacteremic on admission).  The most common sources were line (57%) and pulmonary (26%); 28% had an unknown source.  Eighty-seven percent (104/119) were susceptible to PT; 16 (13%) susceptible isolates had a higher MIC of 32-64 mcg/ml.  Overall in-hospital mortality rate was 27% and the recurrence rate was 7%.  Seventy-one (68%) patients with PT susceptible isolates were on PT at some time in their hospital stay; 51 of those patients had an empiric regimen that include PT (42 were monotherapy). Of all the PT therapy in PT susceptible isolates, 32 (45%) of the 71 were suboptimally dosed based on renal function and indication.  Suboptimally dosed patients were younger (p = 0.03), more likely to have cardiovascular disease (p = 0.003) or line source (p = 0.06), and less likely to be on dialysis (p = 0.005) compared to those optimally dosed.  No difference in mortality/clinical-microbiologic failure was observed comparing the 4 cohorts of: 1) PT vs. other antipseudomonal drugs; 2) PT alone vs. PT in combination with other antipseudomonal drugs; 3) optimal vs. suboptimal PT dosing; 4) treatment of patients with PT MIC 32-64 mcg/ml with PT vs. other antipseudomonal drugs. 

Conclusions: A large portion of our hospitalized pseudomonal bacteremic patients did not have optimal PT dosing.  There were few PA bacteremia with higher susceptible PT MICs treated with PT.  Our study was retrospective with a small number of patients, which may account for the lack of difference in outcomes between subcohorts.  We plan to expand our review to 5 years, encompassing a larger sample size, to try address these questions. Our institution is in discussion regarding changes to the breakpoints of PT for PA.  Additional plans to improve PT dosing with stewardship interventions and a pilot of extended infusion is ongoing.