Over the past few years multiple centers have reported high rates of acute kidney injury  (AKI) with patients who are receiving vancomycin (vanco) + piperacillin-tazobactam (pip-tazo) concomitantly.  This combination of antibiotics is commonly used in multiple inpatients settings including healthcare-acquired infections such as pneumonia or post-operative abdominal infections. Often two antibiotics are needed for empiric therapy, but in many cases, single agents can be used if appropriate cultures are obtained and reveal the cause of the infection.  However, this can take 72 hours or longer, depending how collection of specimens, to get this information. As a result patients can be on this combination for multiple days in the hospital, making these reports of acute kidney injury more concerning.

In January, a report from Detroit was published in Clinical Infectious Diseases by Navalkele and colleagues.  This was a multi-center matched cohort study comparing patients on vancomycin + pip-tazo to those receiving vancomycin + cefepime (which has a similar gram-negative spectrum to pip-tazo). This study was designed as a matched cohort study and included 279 patients on vancomycin and pip-tazo.  There were 279 (out of 803) patients on vancomycin and cefepime who were matched on five factors (severity of sepsis, ICU or non-ICU, duration of combination therapy, dose of vancomycin and number of other nephrotoxic agents).  All the patients had baseline normal renal function. The patients were relatively similar at baseline, though those receiving  pip-tazo were more likely to be male, Caucasian and transferred from another hospital than those receiving cefepime.

The important findings of the study were:

• Rates of AKI were 24-32% in the vanco + pip-tazo group vs 8-14% in the vanco + cefepime group
• Multivariate analysis indicated receipt of vanco + pip-tazo was an independent risk factor for AKI (Hazard Ratio 4.3)
• AKI rates in cefepime treated patient correlated with high vancomycin levels, but did NOT correlate with vancomycin levels in patients on pip-tazo
• Much of the AKI in patients on pip-tazo + vanco occurred within the first 4-5 days
• Patients on vanco + pip-tazo had longer duration a day by 2 days but no difference in mortality or rate of hemodialysis

The findings in this study are similar in magnitude to a recent meta-analysis published  also in CID.  This study compiled 11 prior studies and calculated an adjusted Odds Ratio of 3.11 for patients on Vanco+ pip-tazo  for increasing the risk of AKI. The analysis was limited by the quality of the primary data being heterogeneous and having different comparison groups.

One advantage of the meta- analysis is the larger numbers allowed for more sub-group analysis. Importantly the adjusted Odds Ratio for patients who were critically ill (i.e. in an intensive care unit) was not significantly higher with this combination of antibiotics over other critically ill patients. Indeed, in the Navalkele study, 80% of the patients were not critically ill,  so drawing conclusions about these patients from this data could be considered over-extrapolation.

While we do not have a randomized trial comparing these two agents (and do not except one in the near future), we can draw a few tentative conclusions from this literature:

1. Some hospitalized patients (non-icu) appear to be more vulnerable to AKI from vanco+piptazo that appears to be independent of vancomycin concentration level
2. It is unclear if this risk is similar in patients with baseline CKD, but it appears to not be as great in patients admitted to the ICU
3. It situations where there is equipoise between gram negative agents, choosing cefepime over pip-tazo in combination therapy may help decrease AKI and decrease length of stay
4. Duration of combination therapy may be important to the risk profile for AKI in combination therapy. The best intervention for patients may be to stop vancomycin if not indicated (which is often the case) after 48-72 if there is no MRSA or other clear indication for vancomycin. If there is a gram positive infection, stopping broad gram negative coverage may also be helpful.

The studies suggest a possible harm, but without firm outcomes (perhaps longer hospitalization, but no other clear morbidity and mortality consequences), it may be hard to justify a large shift in practice (such as never using these two medications in combination). It will be interesting to see how various physicians, hospitals and antibiotic stewardship programs integrate this data into practice.   Additionally, new guidelines for complicated infections may comment on this issue in the future.  It would be helpful to have more data in CKD patients and in ICU patients but controlling for confounding factors may be harder in these special populations.

NB: Some of these discussion points are drawn from a Journal Club from March 2017 given by Nate Shively, ID Fellow at Pitt.

Pubmed References

Navalkele B et al Risk of Acute Kidney Injury in Patients on Concomitant Vancomycin and Piperacillin-Tazobactam Compared to Those on Vancomycin and Cefepime. Clin Infect Dis (2017) 64 (2): 116-123.

Hammond DA et al Systematic Review and Metaanalysis of Acute Kidney Injury Associated With Concomitant Vancomycin and Piperacillin/TazobactamClin Infect Dis (2017) 64 (5): 666-674