Reply to Koehler et al

To the Editor—We appreciate the interest Koehler et al expressed in our review of combination antimicrobial therapy against Enterococcus faecalis [1], particularly as it pertains to infective endocarditis (IE) [2]. The authors raise important clinical points on the optimal treatment options and the necessity of combination therapy in IE and highlight a key conclusion of our review paper: current data are limited and further studies are required [1].

First, Koehler et al question the necessity of combination therapy for E. faecalis IE. Early clinical data suggested that clinical cures in the treatment of IE were only 50% compared to those in streptococci [3]. Subsequent experiments showed that the addition of gentamicin or streptomycin was bactericidal in vitro and increased cure rates, becoming the standard of care [3–11]. Moreover, current data that support the use of ampicillin monotherapy for IE are limited to case reports and animal studies. A case report of high-level aminoglycoside-resistant E. faecalis IE demonstrated cure in a 75-year-old male (nonsurgical candidate) after the continuous infusion ampicillin dose was increased from 16 g to 24 g/day to achieve bactericidal activity [12]. While case reports may serve as hypothesis-generating literature, they are at risk of publication bias as many negative studies (ie, treatment failures) are unreported. Likewise, animal models have demonstrated some treatment success, but findings need to be corroborated in humans [13]. Furthermore, vegetations in animal models are developed over the course of 1–2 days, whereas human vegetations may develop over weeks and have varying degrees of biofilm vs planktonic bacteria, not comparable to that of animal studies.

Second, Koehler et al raise questions regarding the mechanistic basis of the synergism. The mechanism of synergism between aminoglycosides and cell wall agents were elegantly studied by Moellering et al, suggesting that the alteration in cell wall synthesis conferred by the B-lactam agent (and also vancomycin) increased the amount of aminoglycoside that reached its intracellular ribosomal target yielding a bactericidal effect [14]. This effect has been shown to be beneficial both in vivo and in vitro [14]. On the other hand, the double B-lactam synergism stems from a penicillin-binding protein (PBP) effect, in which saturation of 2 different PBPs creates a synergistic bactericidal effect despite lack of susceptibility to individual agents by conventional minimum inhibitory concentrations (MIC). This phenomenon also reminds us that MIC determination has important limitations, and interpretation of these values is not straightforward in situations when antibiotic options are limited. This issue has become overwhelmingly apparent in the treatment of multidrug-resistant gram-negative and gram-positive infections in the last decade. Only a full understanding of the mechanistic basis of resistance allows for novel strategies to be implemented and tried in clinical scenarios.

Finally, Koehler et al question the validity of the clinical studies on E. faecalis IE. We agree that prospective, randomized clinical trials would be ideal to provide robust recommendations for the treatment of all infections. However, the reality is different, and in many instances, only limited data provide bases for recommendations. The Spanish cohort described by Fernandez-Hidalgo et al [15] is the best evidence of the efficacy of the ampicillin–ceftriaxone combination. Although some of the data were collected retrospectively, the study attempted to identify patients in a prospective manner. The overwhelming result of that study was that the ampicillin–gentamicin combination was associated with statistically higher frequency of renal toxicity without major differences in clinical outcomes compared to the ampicillin–ceftriaxone combination. Moreover, the combination was selected in patients after careful in vitro and animal experiments that suggested that this combination could be effective. Although, we agree that the use of ceftriaxone could pose collateral damages to the microbiota and increased colonization by Clostridium difficile and others, the nephrotoxicity of the ampicillin–gentamicin combination seems to be a major clinical limitation for patients who require prolonged therapy. The recent publication of the POET trial [16] raises the interesting possibility of using an early switch to oral agents in selected cases of E. faecalis IE, decreasing the toxicity of current regimens.

In summary, we agree that more clinical data are needed to validate our approaches for the treatment of E. faecalis IE. However, in the absence of such data, judicious translation of studies with robust mechanistic basis along with strong translational science is the best strategy to improve the care of our patients.

Note

Potential conflicts of interest. K. L. L. has received research funding or acted as an advisor or consultant for Merck, Davol/BARD, Actavis, Melinta Therapeutics, the Medicines Company, and Pfizer. C. A. is employed by the University of Texas Health Science Center; has received grants from Merck and MeMed Diagnostics; and has received royalties from UpToDate and Harrison Principal of Internal Medicine. All other authors reported 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.