Effects of echinocandin preparations on adult rat ventricular cardiomyocytes

Candida infections represent a relevant risk for patients in intensive care units (ICU; [1, 2]) resulting in increased mortality [3, 4, 5]. Adequate antifungal treatment is essential in patients with sepsis to prevent irreversible damage due to microbial load, systemic inflammation and organ failure [6].

The echinocandins are large (molecular weight ~ 1200 kDa) semisynthetic cyclic hexapeptides derived from various natural fungal products [7, 8]. The lipophilic side chain of the echinocandins intercalates with the phospholipid bilayer of the fungal cell membrane where it serves as a non-competitive inhibitor of β-1,3-D-glucan synthase. Deficiency of β-1,3-D-glucan in the fungal cell wall results in osmotic instability and fungal cell lysis. The echinocandins have become the antifungal agents of first choice due to the excellent activity against most Candida strains and the favorable safety profile. Guidelines recommend echinocandins as primary therapy for all forms of candidiasis, especially in severely ill patients with organ dysfunction [9, 10]. Development of septic cardiomyopathy reflects a crucial pathogenic part of hemodynamic instability in septic shock which aggravates tissue hypoperfusion and organ failure [11]. Due to cardiac effects following echinocandin administration seen in ICU patients [12] the in vitro effects of echinocandins and fluconazole in clinically relevant concentrations on isolated cardiomyocytes of the rat were examined.

In the experiments anidulafungin concentrations of 3.3, and 10 μg/ml showed a significant increase of contractility responsiveness in contrast to 0.1 and 1 μg/ml which were not different compared to controls (Fig. 1 a,b; Tab. 1). For caspofungin cell shortening was not different compared to controls for 0.1–3.3 μg/ml; however, incubation with 10 μg/ml showed a significant decrease. Micafungin concentrations of 0.1 and 1 μg/ml showed no effect in contrast to controls while incubation with 3.3–33 μg/m led to a significant increase in cell shortening. Measurement was not possible at 33 μg/ml for anidulafungin and caspofungin and at 100 μg/ml for all echinocandins due to a majority of round-shaped, non-contracting cardiomyocytes. Fluconazole showed no significant effect on cell shortening in all analyzed concentrations.

For all echinocandins the ratio of round-shaped, non-contracting versus rod-shaped normal-contracting cardiomyocytes increased dose-dependently. A stable rate of 44.3–53.5 % round-shaped cardiomyocytes was observed for anidulafungin concentrations from 0.1 to 10 μg/ml, whereas 86.6 % and 99.8 % of cardiomyocytes were round-shaped at 33 μg/ml and 100 μg/ml, respectively (Fig. 1 c; Tab. 2). Caspofungin concentrations of 0.1 and 1 μg/ml provided comparable proportions of rounded cardiomyocytes as controls while the proportion increased dose-dependently from 3.3 µg/ml to 100 µg/ml from 63.6 % to 73.1 %, 97.2 % and 100 %, respectively. At micafungin concentrations of 0.1–10 μg/ml equivalent fractions of round-shaped cardiomyocytes were seen as in controls; however, at concentrations of 33 and 100 μg/ml cardiomyocytes revealed a rounded shape in 66.4 and 95.3 %, respectively. Fluconazole showed no significant effect on shape at all analyzed concentrations.

The rate of round-shaped cardiomyocytes, following high-dose echinocandins given in culture medium containing 10 mg/ml albumin, proceeded according to assumed concentration gradients (Fig. 1 d). Preincubation of echinocandins with 10 mg/ml albumin led to cancellation of all of the described effects.

The results revealed for the first time that echinocandin preparations in contrast to fluconazole have a dose-dependent impact on function and shape of isolated rat cardiomyocytes in vitro. Nevertheless, there has already been evidence of an ex vivo cardiotoxicity in rats performed with a Langendorff heart experiment [16]. Some very recent clinical reports described severe hemodynamic instability during anidulafungin administration in a critically ill ICU patient [17] and a flash pulmonary edema of unknown origin in a 52-year-old male [18]. Cardiac effects following echinocandin administration have also been seen in septic ICU patients [12]. Additionally, Stover et al. [16] summarized four cases from the FDA adverse events reporting system (FAERS) which may reflect the cardiac impact of the echinocandins: two reports of arrhythmia (anidulafungin), one cardiac failure (anidulafungin) and one described as sudden cardiac death (caspofungin; [19]).

Regarding effective tissue concentrations, there are some limits in transferring the effects of isolated cardiomyocytes to an in vivo situation. Cardiomyocytes are protected in vivo by an endothelial barrier; however, disrupted endothelial layers in sepsis might promote increased antifungal tissue concentrations. Consequently trough concentrations of caspofungin in critically ill patients were found to be higher (mean 2.16 µg/ml) than in healthy subjects [20]. Additionally, even in healthy patients caspofungin already reaches peak plasma concentrations up to 20 µg/ml [21] and 9.94 µg/ml in steady state, while heart tissue reaches about 50 % of plasma concentrations [22, 23] anidulafungin reaches maximum levels up to 14 µg/ml [24, 25] and the peak micafungin plasma concentration is approximately 8.8 µg/ml [26]. Recommended dose regimens provide loading dosages for caspofungin and anidulafungin which could lead to even higher intramyocardial concentrations. Regarding this, all clinical echinocandin effects which were reported in the case reports [12, 17] were seen according to initial loading dosages (200 mg anidulafungin and 70 mg caspofungin) in critically ill septic patients.

Because echinocandins are highly protein-bound in plasma and thus the total concentration could be many times higher than the free concentration, albumin was added to mimic plasma conditions. Primarily, it has to be considered that serum albumin levels are frequently decreased in patients at risk for invasive fungal infections which would increase the active drug concentration. Additionally, no data concerning the kinetics of protein binding of echinocandins following bolus infusion in humans appear to have been published. Especially in situations when antifungal treatment is needed, it is mostly administered through a central line, which could be followed by high active cardiac concentrations due to less time to bind to proteins. Moreover, standard minimum inhibitory concentrations (MIC) of echinocandins used to predict treatment response are estimated in the absence of albumin. However, albumin addition to this microbiological testing significantly raises the MIC to values above the cut-off that is applied to classify microbiological resistance. Data from phase 2 and 3 studies have not described adverse events regarding cardiac failure following echinocandin administration. Product information for caspofungin from the European Medicines Agency (EMA) mentions occasional congestive heart failure. However, most studies on antifungal agents did not include a relevant rate of critically ill patients and therefore, hemodynamic reserves might have been sufficient in the overall study populations. In contrast in septic patients it is possible that a drug-related cardiac failure is wrongly attributed to a progress of illness. According to the drug evaluation documents from EMA and the U.S. Food and Drug Administration (FDA), hypotension and a hemodynamic breakdown are described for an experimental administration of caspofungin in rats. This was associated with vascular histamine release. It has not yet been examined if histamine release is a possible mechanism of the observed effects. Nevertheless, the very recent report from Fink et al. [17] and in vivo observations in three patients [12] lead to the assumption that there must be an additional non-histamine-related pathway which leads to cardiac breakdown. Additionally histamine-releasing cells (e.g. mast cells) as mediators were not observed in the preparations in the current study.

The results of the study do not give the explanation for the mechanism of the observed effects on isolated cardiomyocytes by echinocandins. As Stover et al. [16] based on other publications [27, 28] already mentioned, there are several possible mechanisms of drug-induced cardiomyopathy. The evidence of myocyte and mitochondrial damage was observed using transmission electron microscopy [16]. Furthermore, the observed rounded shape of isolated cardiomyocytes in the current experiments indicates a disturbed intracellular calcium homeostasis. However, the possibility of a receptor mediated mechanism or direct toxic effects, e.g. due to oxidative stress cannot be ruled out. Because of the rapidity in which the effects appeared, alterations in cardiomyocyte gene expression or protein synthesis are not very likely to be the reason for the observed effect. Because of the absence of endothelial cells in the experimental set-up any involvement of endothelial cells in the observed results can be ruled out. Remarkably, lower doses of anidulafungin (3.3 and 10 µg/ml) and micafungin (3.3, 10 and 33 µg/ml) caused an increase in contractility responsiveness whereas caspofungin (10 µg/ml) led to a decrease. However, the same rounded fibrillating shape of the isolated cardiomyocytes without a straightened contraction was observed when exposed to high doses of all echinocandins. Although the molecular structure of the scaffold is very similar, the echinocandins have some differences in the side chains what is assumed to be the reason for such differences [25]. This might also be the reason why the MICs of the echinocandins are not equal [8, 29, 30]. Nevertheless, the rate of cardiomyocytes with a round cell shape consistently increases for all echinocandins dose-dependently.

There are some limitations of the study: First of all, the results from isolated rat cardiomyocytes are not transferable one-to-one to the human clinical setup on the ICU. Isolated cells are not protected by endothelium and higher free echinocandin expositions are even implicated by the absence of albumin in the experiment. Additionally, clinical case reports described effects in septic critically ill patients whereas the experiments were performed in cells harvested from healthy animals. Due to the experimental approach, it is not possible to give decisive recommendations for the clinical choice of any echinocandin preparation or a required modification of the way of administration. Finally, it is not clear whether the observed effects are selective for cardiomyocytes.