Early Switch From Vancomycin to Daptomycin in Methicillin-Resistant Staphylococcus aureus Bacteremia: Still More Questions Than Answers

To the Editor—Schweizer et al [1] report lower mortality (8.3% vs 17.4%; hazard ratio, 0.48) among patients with methicillin-resistant Staphylococcus aureus (MRSA) bacteremia who switch to daptomycin from vancomycin within 3 days vs remaining on vancomycin. This study follows other retrospective reports that suggest a benefit of switching to daptomycin for MRSA infections [2–4]. We thank the authors for presenting new data on an important topic and also suggest caution in drawing conclusions from these retrospective studies. A randomized, controlled trial (RCT) compared antistaphylococcal penicillin or vancomycin plus gentamicin to daptomycin in Staphylococcus aureus bacteremia and endocarditis. In that study, it was found that daptomycin vs standard of care yielded a 12.6% difference in treatment success (95% confidence interval, –7.4% to 32.6%), favoring daptomycin for MRSA; however, this was confounded by addition of gentamicin to the standard care group. Further, there was not a better rate of mortality (10.8% vs 11.3%; P = 1) or clinical failure (3.3% vs 3.5%; P = 1) and nonsignificantly higher microbiologic failure (15.8% vs 9.6%; P = .22) [5]. We wish to highlight 3 points in Schweizer et al’s work that may have predisposed to finding significantly different treatment effects.

First, defining the primary outcome as mortality within 30 days of receiving vancomycin introduced immortal time bias [6], as the median time to switch to daptomycin was 6 days. Excluding patients who died within ≥6 days of receiving vancomycin, rather than just 3 days, could have better addressed this bias, as would a matched cohort design that match patients based on survival to the time of daptomycin switch in the daptomycin group.

Second, the 2 groups comprised patient populations that substantially differed, likely influencing the outcome independent of exposure. For example, the patients in the daptomycin group were younger, more likely to get an infectious diseases consultation, more likely to receive multiple anti-MRSA drugs, and more likely to receive care at high-complexity facilities. The authors provide results of Cox regression analyses but do not describe the final models in detail. A propensity score-matched cohort design might have better accounted for between-group differences, and analyses stratified by potential confounders could have identified treatment effects in subpopulations with an a priori expectation of benefit.

Finally, the adequacy of vancomycin therapeutic dose monitoring (TDM) was not reported. It is unclear if patients in the vancomycin group promptly achieved and maintained goal vancomycin concentrations. Another concern deals with the pharmacodynamic exposure targeted (ie, troughs vs area under the curve [AUC]). The 2020 vancomycin dosing guidelines recommend AUC rather than trough monitoring for serious MRSA infections, with a goal AUC/minimum inhibitory concentration_{broth dilution} ratio of ≥400 [7]. Given that the study used data from 2007 through 2014, it is likely that all vancomycin was monitored via trough goals and may not reflect vancomycin outcomes with optimal TDM.

Given the points discussed, we are hesitant to conclude that early switch to daptomycin offers benefit for the treatment of MRSA bloodstream infections until future prospective/randomized studies can be performed. We do believe studies such as this should compel the funding of future RCTs to address real-world questions commonly encountered in infectious diseases.

Notes

Financial support. M. H. S. reports grants from the National Institutes of Health—National Institute of Allergy and Infectious Diseases (R21AI149026) during the conduct of the study.

Potential conflicts of interest. M. H. S. reports personal fees from Achaogen for consulting; personal fees from SIGA Technologies, Paratek, AbbVie, and Premier Healthcare Solutions; grants from Merck and Co, SuperTrans Medical, Allecra, and Nevakar; grants and personal fees/honoraria from various nonprofit organizations and universities; and personal fees from Hall, Booth, Smith, P.C. and Chambless, Higdon, Richardson, Katz & Griggs, LLP for consulting and expert testimony. In addition, M. H. S. has a patent US10688195B2 issued. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.