Cardiac toxicity of the echinocandins: chance or cause and effect association?
K. R. Stover*,† PharmD, S. T. King*,† PharmD and J. D. Cleary† PharmD
*Department of Pharmacy Practice, University of Mississippi School of Pharmacy, Jackson, MS, and †Division of Infectious Diseases, School of Medicine, University of Mississippi Medical Center, Jackson, MS, USA
Received 2 August 2013, Accepted 8 October 2013
Keywords: cardiotoxicity, echinocandins
What is known and objective: Fungal infections pose a constant risk to critically ill and immunosuppressed patients. The echinocandin antifungals give practitioners an arsenal of agents with apparently lower toxicity relative to older agents. The objective of this commentary is to review the cardiac toxicity of the echinocandin antifungals in the light of recent evidence and published case reports.
Comment: Three case reports detail cardiac decompensation following the initiation of anidulafungin and caspofungin and corroborate ex vivo laboratory results, in which rat hearts exposed to anidulafungin and caspofungin had significantly decreased cardiac contractility. Our hypothesized mechanism of toxicity of anidulafungin and caspofungin is mitochondrial toxicity.
What is new and conclusion: The clinical corroboration of the ex vivo work presented above highly suggests that the cardiac toxicity seen with some of the echinocandin antifungals is a cause and effect pattern, not a chance finding.
WHAT IS KNOWN AND OBJECTIVE
Fungal infections pose a constant risk to critically ill and immunosuppressed patients. Pharmacotherapy of such infections is more challenging than with bacterial infections due to greater cellular similarity between the eukaryotic host and pathogen. Therapy against these infections has always been only marginally effective and riddled with adverse events. The discovery and subsequent marketing of the echinocandin antifungals gave practitioners an arsenal of agents with apparently lower toxicity relative to older agents. Nonetheless, as these agents have become more widely used in ICUs and other at-risk populations, novel adverse reactions are surfacing. The objective of this commentary is to review the cardiac toxicity of the echinocandin antifungals in light of recent evidence and published case reports.
Hindahl and Wilson,1 Fink et al.2 and Lichtenstern et al.3 should be commended for their astute and vigilant adverse event monitoring of their patients on echinocandin therapy. Their recognition of the
adverse effects, thorough assessment and prompt clinical inter- vention have provided valuable insight into the adverse cardio- pulmonary effects of two echinocandin agents.
These cases detail a syndrome of cardiac decompensation following the initiation of caspofungin and anidulafungin and corroborate our clinical observations and the Antimycotic Research Center’s ex vivo laboratory results. In an ex vivo Langendorff model, we exposed hearts from Harlan Sprague Dawley rats to anidula- fungin, caspofungin and micafungin. We utilized concentrations from 0ti 5 to 10 times clinical (serum) concentrations achieved after peripheral administration.4 In our model, anidulafungin (10–80 lg/
mL) and caspofungin (6–48 lg/mL) significantly decreased cardiac contractility over all concentrations tested. At concentrations achievable in humans, both anidulafungin 10 lg/mL prepared
with and without ethanol (ti55ti 0 ti 20ti 7% and ti 56ti 3 ti 38ti 6%, respectively) and caspofungin 6 lg/mL (ti23ti 3 ti 24ti 9%) 4–10and 12 lg/mL (ti17ti8 ti 16ti 6%) led to reductions in contractility. Similarto thiscaseseries, cardiacdecompensationin ourstudies was seen within the first few minutes of medication administration.
Micafungin was tested at exposure concentrations from 7ti 2 to 72 lg/mL and did not result in a change in cardiac contractility. Preliminary results demonstrate similar results in vivo.11
There are a few points that we feel warrant attention. First, these cases report cardiac dysfunction with anidulafungin and caspo- fungin similar to the effects seen in our studies and clinically.4 No cases have been reported to date with micafungin. It is possible that the difference in toxicity is a result of the difference in lipophilicity of these agents. Anidulafungin and caspofungin are highly lipophilic echinocandins and have been associated with toxicity, whereas micafungin is more hydrophilic and has not been associated with cardiac toxicity.4 Structural similarities to oligo- mycin may also contribute to the activity.
Important shared characteristics between the patients are worth a brief discussion (Table 1). Most concerning is the fi nding that each of the patients reported by Lichtenstern et al.3 had pre- existing left ventricular hypertrophy and/or dysfunction. Given the excess metabolic demands during critical illness, as well as the fragile Starling curve of a failing heart, these two populations of patients warrant close monitoring while receiving an echinocan- din. This is even more important for those patients receiving loading doses of anidulafungin and caspofungin through central venous catheters, given the proximity of infusion to the heart. Our anecdotal experience suggests that central venous administration
Correspondence: K. R. Stover, Department of Pharmacy Practice, University of Mississippi School of Pharmacy, 2500 North State Street, Jackson, MS 39216, USA. Tel.: +1 601 984 2615; fax: +1 601 984 2751; e-mail: [email protected]
is instrumental in the reaction. Prior use of cardiotoxic medications was not documented in the majority of the cases. Nonetheless, a history or concomitant use of certain medications (anthracycline agents, thiazolidinediones, itraconazole, trastuzumab, anti-ar-
© 2013 John Wiley & Sons Ltd 1
Table 1. Summary of patient case series
Patient Age (years) Gender Ethnicity Cardiac disease Liver dysfunction Renal dysfunction Vasopressor support ICU care
11 52 Male Caucasian N Y N N N
22 41 Male Caucasian N Y Y Y Y
33a 81 Female Not reported Y N Y Y Y
43a 71 Male Not reported Y N Y Y Y
53a 66 Male Not reported Y N N Y Y
aPatients 3, 4 and 5 correspond to cases 1, 2 and 3, respectively, from Lichtenstern et al.
Fig. 1. Mitochondrial oxidative phosphorylation pathway. Substrates for each complex are displayed at the point of entry ( ). Known and proposed sites of inhibition are displayed ( ). FMN and FAD = oxidized fl avin nucleotides; FE-S = iron–sulphur protein;
Q = coenzyme Q/ubiquinone; Cyt = cytochrome; BAL = dimercaprol, TTAF = iron-chelating agent.
rhythmics, amphotericin B) may further exacerbate the cardiotoxic effects of the echinocandins.
Finally, only one of the published reports hypothesizes a potential mechanism of toxicity. This report hypothesizes that histamine release was the cause of the fl ash pulmonary oedema seen after anidulafungin administration.1 As histamine release is a reported adverse event, we conducted a complete dose-ranging study on histamine in our model and observed no dose-associated decrease in cardiac contractility.4 Instead, we feel that the rapid nature of the effects seen clinically and in our ex vivo studies may refl ect damage to the mitochondria.12 We grossly examined hearts exposed to high doses of anidulafungin and caspofungin with a transmission electron micrograph. The images showed enlarged mitochondria and disintegrating myofi brils, among other changes.4,12 Samples exposed to high concentrations of micafungin showed some enlarged mitochondria, but no evidence of disintegrating myofi- brils. In isolated mitochondrial studies, we tested the effects of anidulafungin, caspofungin and micafungin on oxidative phos- phorylation. Initial results display dose-dependent inhibition of oxidative phosphorylation with all agents tested.13 Preliminary data suggest that these agents inhibit several steps in the mitochondrial oxidation pathway.13 Most specifically, we hypothesize that anidu- lafungin and caspofungin inhibit complex III, a coenzyme Q-
dependent cytochrome c reductase, and complex IV, a cytochrome c oxidase, as the primary mechanism.14 (See Fig. 1). The specifi c action of micafungin on the mitochondrial oxidation pathway has not yet been determined. These data appear similar to the recently described mechanism for itraconazole-associated heart failure.15
Moving forward, it is clear that a randomized controlled trial assessing cardiotoxicity would be unethical. Prospective, observa- tional trials will prove to be invaluable in further delineating the population at greatest risk of cardiotoxic reactions to the echino- candins. Practitioners should be cognizant of these effects and routinely and continually monitor their patients. These observa- tions can be optimized by high-level assessments, such as the pulse contour cardiac output monitor (PiCCO) utilized by Lichtenstern et al.
WHAT IS NEW AND CONCLUSION
The clinical corroboration of the ex vivo work presented above highly suggests that the cardiac toxicity seen with some of the echinocandin antifungals is a cause and effect relationship and not a chance finding. Preliminary data suggest that the mechanism of toxicity is mitochondrial damage. Until observational studies delineate the populations most at risk, clinicians should continue
to be vigilant with cardiac monitoring of patients receiving the echinocandin antifungals.
CONFLICT OF INTEREST
KS and JC have received grant funding from Astellas Pharma, Inc.,
The in vitro work described in this commentary was supported by an investigator-initiated protocol from Astellas Pharma, Inc. (Northbrook, IL).
for the research cited above.
1.Hindahl CB, Wilson JW. Flash pulmonary oedema during anidulafungin administra- tion. J Clin Pharm Ther, 2012;37:491–493.
2.Fink M, Zerlauth U, Kaulfersch C et al. A severe case of haemodynamic instability during anidulafungin administration. J Clin Pharm Ther, 2013;38:241–242.
3.Lichtenstern C, Wolff M, Arens C et al. Cardiac effects of echinocandin prepara- tions – three case reports. J Clin Pharm Ther, 2013;38:429–431.
4.Stover KR, Farley J, Kyle P, Cleary JD. Cardiotoxicity of some echinocandin anti- fungals. Expert Opin Drug Saf, 2013; doi:10.1517/14740338.2013.829036. [Epub ahead of print].
5.Lui P, Ruhnke M, Meersseman W, Paiva JA, Kantecki M, Damle B. Pharmacokinetics of anidulafungin in critically ill patients with candidemia/invasive candidiasis. Antimic- rob Agents Chemother, 2013;57:1672–1676.
6.Weiler S, Seger C, Pfisterer H et al. Pharma- cokinetics of caspofungin in critically ill patients on continuous renal replacement
therapy. Antimicrob Agents Chemother, 2013;57:4053–4057.
7.Kuse ER, Demeyer I, Undre N. Pharmaco- kinetics of micafungin in adult patients with invasive candidiasis and candidaemia. OJMM, 2012;2:84–90.
8.Stone JA, Holland SD, Wickersham PJ et al. Single- and multiple-dose pharmaco- kinetics of caspofungin in healthy men. Antimicrob Agents Chemother, 2002;46:739– 745.
9.Pfizer Inc. Eraxis [package insert]. New York, NY: Pfizer Inc., 2006.
10.Keirns J, Sawamoto T, Holum M, Buell D, Wisemandale W, Alak A. Steady-state phar- macokinetics of micafungin and voriconaz- ole after separate and concomitant dosing in healthy adults. Antimicrob Agents Chemother, 2007;51:787–790.
11.Stover KR, Cleary JD. In vivo evaluation of cardiac toxicity with caspofungin. American College of Clinical Pharmacy 2013 Annual Meeting. Abstract 114. Albuquerque, NM. Pending October 2013.
12.Stover KR, Hosler JP, Henry GM, Harkins M, Cleary JD. Mitochondrial toxicity with caspofungin. American College of Clinical Pharmacy 2011 Annual Meeting. Abstract 130. Pittsburgh, PA. October 2011.
13.Stover KR, Hosler JP, Cleary JD. Inhibition of mitochondrial oxidative phosphorylation by ani- dulafungin. American College of Clinical Pharmacy 2012 Annual Meeting. Abstract 113. Hollywood, FL. October 2012. Encore: UM SOP Research Poster Day. 8 November 2012.
14.Hosler J, Shapleigh JP. Encyclopedia of environmental microbiology [Internet]. Aerobic Respiration, 1st edn. New York, NY: John Wiley & Sons, Inc, 2003. Available at: http://onlinelibrary.wiley.com/doi/
10.1002/0471263397.env158/abstract (accessed 28 October 2013).
15.Cleary JD, Stover KR, Farley J, Daley W, Kyle PB, Hosler J. Cardiac toxicity of azole antifungals. Pharmacol Pharm, 2013;3:362– 368.MK-0991