|Home | About | Journals | Submit | Contact Us | Français|
According to the product insert for Cancidas (caspofungin acetate), the drug must not be diluted in solutions containing glucose as this decreases caspofungin stability. The aim of this study was to compare caspofungin MICs for a collection of yeasts by means of EUCAST method EDef7.1 but using two different concentrations of glucose: 2% versus 0.2%. MICs were identical or within one 2-fold dilution for 93 out of 95 strains (97.9%), showing that glucose does not interfere with susceptibility.
In a previous work, a comparison of the major susceptibility testing methods for echinocandins was performed using a well-characterized panel of Candida strains (1). EUCAST method EDef 7.1 performed as one of the best methods for discriminating wild-type strains from isolates with mutations in FKS hot spot regions. However, for caspofungin the MICs obtained by EUCAST method EDef 7.1 were in general higher than those obtained using the CLSI M27A3 methodology. In addition, the number of very major errors (number of FSK hot spot mutants classified as susceptible according the wild-type upper limit values) was 50% for EUCAST method EDef 7.1 (1), while 7% of very major errors were obtained by means of CLSI M27A3 at 24 h (1, 2). Finally, this previous work also found that caspofungin MICs obtained by means of EUCAST method EDef 7.1 and CLSI M27A3 were, in general, higher than those obtained by Pfaller et al. in a previous work using CLSI M27A3 methodology (4), indicating a possible variability associated with caspofungin antifungal susceptibility testing (1). Thus, it was of interest to explore which variables influence the reproducibility of the susceptibility tests for caspofungin. Although it has been proven that both the EUCAST EDef 7.1 and CLSI M27 A3 (2, 5) methodologies generate similar results for amphotericin B and azole drugs (3, 6), this does not appear to be the case for caspofungin. The main differences between EUCAST method EDef 7.1 and CLSI M27A3 include different inocula (105 versus 103 CFU/ml, respectively) and glucose concentrations (2% versus 0.2%, respectively) (3, 6).
The product insert for Cancidas states that the compound must not be diluted in solutions containing glucose because it decreases drug stability. As stated previously, the glucose concentration in the growth medium recommended for method EDef 7.1 in RPMI 1640 is 10 times higher than that recommended by the CLSI (5). As this concentration could have influenced the MIC values obtained by means of EUCAST method EDef 7.1 (5), the Antifungal Susceptibility Testing Subcommittee of EUCAST decided to set up a comparison test to examine whether the increased glucose concentration affects the MIC results. The test compared caspofungin MICs for a collection of well-characterized yeast isolates by means of EUCAST method EDef 7.1 (5) but using growth medium with two different concentrations of glucose: 2% versus 0.2%.
Ninety-five strains of Candida were used throughout the study. Sixty-seven were considered susceptible and 28 resistant, including 10 FKS wild-type and 10 fks hot spot mutant Candida albicans isolates, 9 FKS wild-type and 10 fks hot spot mutant Candida glabrata isolates, 1 FKS wild-type and 1 fks hot spot mutant Candida dubliniensis isolate, 13 FKS hot spot wild-type and 3 fks hot spot mutant Candida krusei isolates, 19 FKS wild-type Candida parapsilosis isolates, and 15 FKS hot spot wild-type and 4 fks hot spot mutant Candida tropicalis isolates. This collection of strains has been described and used in a previous work (1).
C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were used as quality control strains in all experiments.
EUCAST microdilution methodology was performed by strictly following the EDef 7.1 standard guidelines (5). Plates were prepared in one batch, sealed in aluminum foil, and stored at −80°C for no longer than 2 months before use. Microtiter plates were read spectrophotometrically at 530 nm after 24 h, and the MIC was determined using 50% growth inhibition. One set of microplates contained a glucose concentration of 0.2%, whereas the other set contained the recommended glucose concentration of 2%.
Statistical analysis was done using PASW statistics, version 18.0 (PASW, S.L., Madrid, Spain). MIC values were transformed to log2.
The reproducibility of the results obtained by reference techniques was calculated to determine the percentage of essential agreement (EA) between MIC values. Agreement was defined as discrepancies in MIC results of no more than ±1 2-fold dilution.
Linear regression analysis for both methods was done to test the linearity of the relationship between the MICs obtained with 2% of glucose versus those obtained with 0.2% glucose.
In addition, a two-way random-effect model was utilized to calculate the intraclass correlation coefficient (ICC) with a 95% confidence interval. The ICC is a reverse measurement of the variability of the counting values. The ICC was calculated using the formula ICC = (group mean square − error mean square)/(group mean square + error mean square) and thus has a maximum value of 1 if there is a perfect correlation and a minimum value of −1 if there is a complete absence of correlation. The ICC evaluates the correlation between values offering statistical significance since it takes into account the number of cases and absolute value of the counting. The ICC is the analysis which exhibits the highest statistical power for correlation studies.
The comparison between the MICs obtained with 0.2% glucose versus 2% glucose (EDef 7.1) showed that 61 strains (64.2%) had identical MICs, 32 (33.7%) had one 2-fold dilution difference (25 [78.1%] had one 2-fold dilution lower and 7 [21.8%] had one 2-fold dilution higher), and 2 (2.1%) had two 2-fold-dilution differences (one lower and one higher). The Pearson's coefficient obtained by means of linear regression was 0.913. The ICC using a random model of consistency for both variables, 2% glucose versus 0.2% glucose, was 0.954, with a 95% confidence interval of 0.931 to 0.969. When the same model was stressed with an absolute agreement model, the ICC was 0.950, with a 95% confidence interval of 0.922 to 0.968. Thus, statistical analysis shows that glucose concentration does not have a significant influence on the MIC values. MICs were essentially the same for 93 strains (97.8%) because the MICs were identical or had one 2-fold dilution difference. If the usual definition of essential agreement had been used (±2 2-fold dilutions), 100% agreement would have been obtained. ICC values obtained by means of essential agreement or absolute consistency were very close—0.954 versus 0.950—with very narrow 95% confidence intervals.
In summary, the Cancidas product insert includes a warning concerning the stability of caspofungin if dissolved in a glucose-containing solution. However, the concentration used in method EDef 7.1 does not influence the susceptibility values, as the MICs obtained with a glucose concentration of 0.2% (CLSI) of glucose were essentially the same as those obtained with a 2% (EUCAST) concentration. Alternative factors must account for the variability observed for differential MIC results obtained using EUCAST and CLSI susceptibility testing procedures.
Potential conflicts of interest are as follows. M.C.A. has been a consultant for Astellas, Merck, Pfizer, and SpePharm and an invited speaker for Astellas, Cephalon, Merck Sharp & Dohme, Pfizer, Schering-Plough, and Swedish Orphan. C.L.-F. has been a consultant for Pfizer and Schering Plough and an invited speaker for Pfizer, Gilead, Schering Plough, and Merck Sharp & Dohme and has received research funding from Pfizer, Gilead, Merch Sharp & Dohme, and Schering Plough. D.S.P. is a shareholder in Merck; has acted as a consultant for Merck, Pfizer, and Astellas; is an advisory board member for Merck, Pfizer, Astellas, and Myconostica; has a patent for assays for resistance to echinocandin-class drugs (application no. 07763-O69WO1); has received research funding (although not for this particular study) from Merck, Pfizer, Astellas, and Myconostica; and has been an invited speaker for Merck, Pfizer, Astellas, and Myconostica. G.G.-E. has no conflicts of interest. A.G.L. has no conflicts of interest. In the past 5 years, J.L.R.T. has received grant support from Astellas Pharma, Gilead Sciences, Merck Sharp & Dohme, Pfizer, Schering Plough, Soria Melguizo SA, the European Union, the Spanish Agency for International Cooperation, the Spanish Ministry of Culture and Education, the Spanish Health Research Fund, Instituto de Salud Carlos III, the Ramon Areces Foundation, and the Mutua Madrileña Foundation. He has been an advisor/consultant to the Panamerican Health Organization, Gilead Sciences, Merck Sharp & Dohme, Myconostica, Pfizer, and Schering Plough. He has been paid for talks on behalf of Gilead Sciences, Merck Sharp & Dohme, Pfizer, and Schering Plough. In the past 5 years, M.C.E. has received grant support from Astellas Pharma, bioMérieux, Gilead Sciences, Merck Sharp & Dohme, Pfizer, Schering Plough, Soria Melguizo SA, the European Union, the ALBAN Program, the Spanish Agency for International Cooperation, the Spanish Ministry of Culture and Education, the Spanish Health Research Fund, the Instituto de Salud Carlos III, the Ramon Areces Foundation, and the Mutua Madrileña Foundation. He has been an advisor/consultant to the Panamerican Health Organization, Gilead Sciences, Merck Sharp & Dohme, Pfizer, and Schering Plough. He has been paid for talks on behalf of Gilead Sciences, Merck Sharp & Dohme, Pfizer, and Schering Plough.
Published ahead of print on 17 May 2010.