There are a few reasons why piperacillin/tazobactam (Zosyn) is not usually my first choice for a broad-spectrum gram-negative agent in the ED. First, at my institution, the Pseudomonas aeruginosa susceptibilities to pip-tazo are lower than that for cefepime. Second, do we really need the anaerobic coverage that pip-tazo provides for every sick patient? Pip-tazo is great for empiric treatment of intra-abdominal and severe diabetic foot infections, but may not be needed for a healthcare associated pneumonia. Third, with its frequent dosing (every 6 hours), too often the second dose is missed if the patient is still boarding in the ED.
Don’t get me wrong, pip-tazo is a great drug. I just want to have it around in the future to treat difficult gram-negative and anaerobic infections.
Link to Acute Kidney Injury
Two abstracts presented at the 2012 Society of Critical Care Medicine meeting suggested that patients who received vancomycin PLUS pip-tazo OR pip-tazo alone had a higher risk of developing acute kidney injury (AKI) compared to patients who received vancomycin alone. Dr. Haney Mallemat (@CriticalCareNow) provided a brief summary of the two abstracts for a UMEM pearl.
Hellwig T, et al. 
Retrospective evaluation of all adult patients admitted to Sanford USD Medical Center over a 6 month period who received vancomycin +/- pip-tazo for more than 48 hours. AKI was defined as an increase of serum creatinine > 0.5 mg/dL or a 50% increase from baseline. Of the 735 patients evaluated, the incidence of AKI for vancomycin alone, pip-tazo alone, and combination of vancomycin + pip-tazo were 4.9%, 11.1%, and 18.6%, respectively (vancomycin vs. pip-tazo, p = 0.014; vancomycin vs. combination, p = 0.005). When looking at just the ICU patients, a similar result was seen. Incidence of AKI was 6.0%, 12.2%, and 21.2%, respectively (vancomycin vs. pip-tazo, p = 0.279; vancomycin vs. combination, p = 0.005).
Min E, et al. 
Evaluation of 140 surgical ICU patients over the course of a year who received vancomycin +/- pip-tazo for at least 48 hours. AKI was defined as an increase in serum creatinine more than 1.5 times baseline during antibiotic therapy. The authors controlled for severity of illness and concomitant use of other nephrotoxic antibiotics. The incidence of AKI was higher in the vancomycin + pip-tazo group (40.5%) compared to the vancomycin alone group (9.0%, p < 0.001).
My thoughts: Two abstracts are certainly not enough to demonstrate causation, but it is an interesting association nonetheless.
Four New Studies
Is it really true? Could pip-tazo actually be increasing risk of AKI?
Moenster RP, et al. 
A retrospective cohort study was conducted of all diabetic patients with osteomyelitis treated with vancomycin plus either pip-tazo or cefepime for at least 72 hours at a VA Medical Center between January 2006 and December 2011. The primary outcome was development of AKI, defined as an increase in SCr of 0.5 mg/dL or 50% of baseline. 139 patients met inclusion criteria; 109 in the pip-tazo group and 30 in the cefepime group. In patients receiving vancomycin + pip-tazo, 29.3% (32/109) developed AKI compared to 13.3% (4/30) treated with vancomycin + cefepime (p=0.099). A multiple logistic regression analysis identified weight and average vancomycin trough as the only significant predictors of AKI. The authors were unable to detect a statistically significant difference in AKI between groups; however, power was not met.
Gomes DM, et al. 
A retrospective matched cohort of 224 patients receiving vancomycin + pip-tazo OR vancomycin + cefepime for more than 48 hours. Included patients had no preexisting kidney disease. AKI was defined according to the Acute Kidney Injury Network criteria. The incidence of AKI was higher in the pip-tazo + vancomycin group (34.8%) compared with the cefepime + vancomycin group (12.5%) in the unmatched analysis (p<0.0001). After adjusting for potential sources of bias through propensity score matched pairs and conditional logistic regression, pip-tazo + vancomycin combination therapy (p=0.003) was an independent predictor of AKI. There were no significant differences in time to AKI or hospital length of stay between groups.
Meaney CJ, et al. 
This group retrospectively evaluated 125 adult internal medicine patients receiving vancomycin treatment for at least 72 hours. Nephrotoxicity, defined as an increase in serum creatinine of 0.5 mg/dl or 50% above baseline (whichever was larger), occurred in 17 (13.6%) of 125 patients. On multivariable logistic regression analysis, after controlling for hypotensive episodes, Charlson Comorbidity Index, and baseline creatinine clearance, concomitant use of pip-tazo was associated with increased vancomycin-associated nephrotoxicity (adjusted odds ratio 5.36, 95% confidence interval 1.41-20.5). They concluded that vancomycin-associated nephrotoxicity is prevalent among internal medicine patients, with 5.36-fold higher odds if pip-tazo is concomitantly administered.
Burgess L, et al. 
A retrospective, single-center cohort of 191 internal medicine and ICU patients receiving vancomycin or vancomycin + pip-tazo for at least 48 hours. AKI was defined as an increase in serum creatinine more than 1.5 times baseline during antibiotic therapy. Nephrotoxicity developed in 8 (8.1%) of 99 patients in the vancomycin group and in 15 (16.3%) of 92 patients in the combination group (1-sided χ2 test, p=0.041). A steady-state vancomycin trough concentration of 15 μg/ml or greater was also associated with an increased risk of the development of nephrotoxicity. This is yet another study where it is difficult to conclude that pip-tazo is the cause of increased AKI compared to vancomycin alone. If the groups are equally sick, then what is the rationale for using two antibiotics compared to one?
My thoughts: Four retrospective studies are also not enough to demonstrate absolute causation, but they certainly add some strength to the association.
While there is no definitive link between pip-tazo and risk of AKI, there have been 6 different groups, with internal medicine and ICU patients, showing a possible association. To me, this adds one more reason to at least think twice about reflexively ordering pip-tazo for every sick patient.
Original: May 20, 2014; Updated: May 28, 2014