Rapid and Sensitive Liquid Chromatography/Mass Spectrometry Assay for Caspofungin in Human Aqueous Humor

Current topical treatment for fungal keratitis is inadequate (13) given the lack of favorable outcomes with existing antifungal eye drops (i.e., fluconazole, amphotericin B, and natamycin) (7, 14). Accordingly, it is important to investigate the topical application of newer antifungal agents that are available only as injections. In rabbit models (5, 8), caspofungin eye drops were effective in inhibiting the progression of fungal keratitis. It remains unknown if caspofungin eye drops are able to penetrate the human eye. Studies in this area are impeded by the absence of a simple and sensitive analytical assay to quantify caspofungin in human aqueous humor.

Currently, high-performance liquid chromatography (HPLC) with fluorescence or amperometric detection (4, 11, 12, 15) and liquid chromatography/tandem mass spectrometry (LC/MS/MS) (1-3, 9, 16) are used to quantify caspofungin in biological samples. The majority of these methods, however, involve liquid-liquid extraction and require large sample volumes. Because aqueous humor samples usually have small volumes (<150 μl), these methodologies are not suitable. This study aimed to develop a rapid and sensitive LC/MS assay for caspofungin that involves simple preparation and low sample volumes.

Caspofungin diacetate was provided by Merck Sharp & Dohme. Azithromycin dihydrate (internal standard [IS]) was purchased from Kopran Ltd. (Maharashtra, India). Calibration standard and quality control (QC) stock solutions of caspofungin (1 mg/ml) were prepared separately in water and stored at −80°C. IS stock solution (1 mg/ml) was freshly prepared in acetonitrile for each analysis. Working solutions of caspofungin and IS were freshly prepared by diluting the stock solutions with water and acetonitrile, respectively.

The calibration curve was constructed using water due to the limited availability of blank aqueous humor (6, 10). Calibration standards were prepared by adding 30 μl IS working solution to 30 μl caspofungin working solutions. The final concentrations of calibration standards were 10, 50, 100, 500, 1,000, 2,000, and 5,000 ng/ml with the IS (500 ng/ml). All QC samples (30, 300, and 4,000 ng/ml) were freshly prepared. The blank aqueous humors were pooled to prepare aqueous humor-based low (30 ng/ml; n = 3) and high (4,000 ng/ml; n = 3) QC samples (30 μl each) to validate the water-based calibration curve.

The Shimadzu LC system (Shimadzu, Japan) comprised a high-pressure gradient unit (LC-20AD pump and LC-20ADsp pump), DGU-20A₃ degasser, and CTO-10A column oven with a FCV-12AH switching valve. A Synergi Hydro-RP C₁₈ column (80 Å, 50 by 2 mm, 4 μm; Phenomenex) and guard column (4 by 2 mm, 4 μm; Phenomenex) were used. The column temperature was 30°C, and the flow rate was 0.5 ml/min. The mobile phase was Milli-Q water-formic acid (A; 100:0.1 [vol/vol]) and methanol-formic acid (B; 100:0.1 [vol/vol]). The gradient elution program was as follows: 0.0 to 2.8 min from 20% B to 30% B, 2.8 to 3.0 min from 30% B to 95% B, 3.0 to 5.0 min at 95% B, 5.0 to 5.5 min from 95% B to 20% B, and 5.5 to 8.0 min at 20% B. The mobile phase was directed to an MS system between 3.0 and 6.0 min. The run time was 8.0 min. The injection volume was 10 μl.

An MS-2010EV single-quadrupole mass analyzer coupled with an electrospray ionization interface (Shimadzu, Japan) was used. The tuning voltages were fixed for the interface, the curved desolvation line (CDL), and Q-array. The detector voltage was 1.5 kV. Flow rates were 1.5 liters/min for nitrogen gas and 10 liters/min for both the nebulizer and the drying gas. The temperatures for CDL and heat block were 200°C. Data were acquired and processed using the LCMSsolutions software program (version 3.4). The protonated ions of caspofungin (m/z = 547.5 [M+2H]²⁺) and IS (m/z = 749.4 [M+H]⁺) were monitored.

Caspofungin eluted at 4.68 ± 0.01 min, and the IS eluted at 4.48 ± 0.01 min (Fig. 1). Blank aqueous humor samples (n = 6) showed no interfering peaks with endogenous products or coadministered drugs used in eye surgery (phenylephrine, cyclopentolate, tropicamide, and oxybuprocaine).

Calibration curves plotted the peak area ratio of caspofungin to IS (y) against the caspofungin concentration (x), with y = 0.005x − 0.074 (n = 5; r² = 0.999 ± 0.003). Good linearity over 10 to 5,000 ng/ml was observed, with 1/x weighting employed. The limit of detection (LOD) calculated using the equation LOD = 3.3σ/S was 2.11 ng/ml (σ is the standard deviation of the calibration curve, and S is the slope of the calibration curve). The lower limit of quantification (LLOQ), defined as the reproducible peak of the lowest concentration in the calibration curve, was 10 ng/ml (n = 6), with an accuracy of 105% and a coefficient of variation (CV) of 5.44%.

For intraday analysis, six determinations of each QC concentration were assayed. For interday assessment, three replicates at each QC concentration were run on five separate days. The intra- and interday accuracies (percent bias) were within 11.0%. The intra- and interday precisions (CV) were within 6% (Table 1). The aqueous humor-based low (mean = 28.8 ng/ml; CV = 4.21%) and high (mean = 4,111 ng/ml; CV = 3.03%) QC samples gave results similar to those of the corresponding water-based QC samples. The retention times for caspofungin in aqueous humor- or water-based samples were identical.

Caspofungin was stable in water (i.e., it remained at ±15% of the initial concentration) under the following conditions: (i) after two freeze-thaw cycles, (ii) at −80°C for 60 days (Fig. 2), and (iii) at 2 to 8°C for 3 days. Caspofungin and the IS in a 1:1 mixture of acetonitrile-water were stable for 24 h when left in the LC/MS autosampler at 4°C. The caspofungin stock solution was stable for 2 months when stored at −80°C (Fig. 2).

Although our LLOQ was not as low as that reported by Rochat et al. (9), it was impossible to adapt their LC/MS/MS method, which involved liquid-liquid extraction, given the limited volume of aqueous humor. The sensitivity of the current assay could be further improved with tandem MS. Caspofungin could nonspecifically adsorb to plastic and glass surfaces, thus reducing assay recovery (9). A simple sample pretreatment with an equal volume of acetonitrile, as per the current assay, prevents the nonspecific adsorption. This is in contrast to published methods which required additional preanalysis processing (2, 9, 11, 15, 16).

This study has established a simple and rapid LC/MS assay which is sensitive, accurate, and reproducible for quantifying caspofungin in a limited volume of biological samples.

• Open in new tab
• Download powerpoint

FIG. 1.

Representative chromatogram of blank aqueous humor sample (A) or two aqueous humor samples containing a caspofungin concentration of 95.1 ng/ml or 63.8 ng/ml (with IS at 500 ng/ml) (B1 and B2, respectively) from two consenting patients who were administered one drop of 0.5% caspofungin eye drop hourly for 4 h prior to the surgery. The aqueous humor samples were removed from the participants 1.67 h and 1.17 h after the last caspofungin eye drop was administered for B1 and B2, respectively.

• Open in new tab
• Download powerpoint

FIG. 2.

Stability of caspofungin in samples and stock solution when stored at −80°C.

• Copyright © 2010 American Society for Microbiology