Development and validation of a liquid chromatography-tandem mass spectrometry assay for the simultaneous quantitation of 5 azole antifungals and 1 active metabolite

2.1. Materials
Burdick and Jackson LC-MS grade acetonitrile and methanol was purchased through Fisher Scientific. Ultrapure water was prepared with a
Milli-Q water purification system (Millipore Synergy). Fluka formic acid was
purchased through Sigma-Aldrich and certified ACS plus hydrochloric acid
from Fisher. The charcoal-stripped serum was from SeraCare. Fluconazole,
fluconazole-13C3, voriconazole, voriconazole-D3, posaconazole, posaconazole-
D4, itraconazole, itraconazole-D4, and hydroxyitraconazole-D4 were from
Cerilliant. Isavuconazole and hydroxyitraconazole were purchased from
Toronto Research Chemicals. Isavuconazole-D4 was purchased from Medical
Isotopes. Hydroxyitraconazole, hydroxyitraconazole-D4, and isavuconazole
were obtained as exact weight powders. Prior to use they were dissolved in
1% (v/v) 1 mol/l HCl in methanol to a concentration of 1 mg/ml. Calibrators
were prepared in charcoal-stripped serum, via serial dilution, at the following
concentrations: 0.2, 0.3, 0.6, 1.3, 2.5, 5.0, and 10.0μg/ml for voriconazole,
posaconazole, isavuconazole, itraconazole, and hydroxyitraconazole. The
fluconazole calibrators were prepared at 0.5, 0.9, 1.9, 3.8, 7.5, 15.0, and
30.0 μg/ml, via serial dilution, in charcoal-stripped serum. The protein precipitation
solution was acidified acetonitrile [0.1% (v/v) 1 mol/l HCl] containing
the internal standards for all 6 analytes (1 μg/ml).

2.2. Sample preparation
Fifty microliters of sample (serum or lithium heparin plasma) was
precipitated with 250 μl of the precipitation solution. After 60 s of
vortex mixing, the sample was centrifuged at 18,900 ×g for 10 min. A
50 μl aliquot was then diluted with 200 μl 0.1% (v/v) formic acid in
ultrapure water, which was then ready for LC-MS/MS analysis.

2.3. LC-MS/MS method
Utilizing a Transcend LC system (Thermo Scientific), 20 μl of the
diluted supernatant was injected onto a reversed-phase column
(Accucore RP-MS, 50 ×2.1 mm, 2.6 μm). Solvent A was comprised of
0.1% (v/v) formic acid in ultrapure water and solvent B was 0.1% (v/v)
formic acid in acetonitrile. At a flow rate of 0.8 ml/min, the gradient
was ramped from 40% B to 60% B over 20 s, held at 60% B for 30 s, and
finally equilibrated back at 40% B for 60 s. A Thermo Scientific TSQ
Vantage triple quadrupole mass spectrometer, operating in multiple
reaction monitoring (MRM) mode, was utilized as the detector. The
source was positive heated electrospray ionization (HESI) operating
with a spray voltage of 4000 V, a vaporizing temperature of 350 °C, a
sheath gas of 50, an auxiliary gas of 10, and a capillary temperature of
200 °C. Two mass-to-charge (m/z) transitions were monitored for each
analyte (quantifier and qualifier) and 1 transition for each internal
standard (Table 1).

2.4. Method validation
To validate the method, the following experiments were performed:
ion suppression, mixing study, interference, AMR, carryover, stability,
precision, and method comparisons. From the stability experiment
onwards, quality control samples at 2 levels were analyzed with every
batch. The levels for all analytes, except fluconazole, were 0.7 and
5.0 μg/ml for the low and high control, respectively. The fluconazole
controls were 1.9 and 15.0 μg/ml, for the low and high, respectively.
The use of leftover patient samples was approved by the Cleveland
Clinic Institutional Review Board.

2.4.1. Ion suppression and mixing study
The absolute ion suppression was evaluated by infusing a 2 μg/ml
solution of each analyte and the respective internal standard through a
connection tee post-column. While the solution was being infused at
25 μl/min, a solvent blank and 10 blank patient serum samples (5 males
and 5 females) were consecutively injected onto the LC column. The
patient blank samples were extracted according to procedure in the
sample preparation section, however, without internal standards. The
signal intensity from each analyte and internal standard were

monitored throughout the chromatogram. The chromatogram was observed
for a reduction or enhancement in signal intensity with the patient
samples versus the solvent blank for each analyte and internal
The acceptability of a candidate matrix for calibrator preparation
was determined via a mixing study. The candidate matrix (charcoalstripped
serum) spiked with 10.0 μg/ml of each analyte was mixed 1:1
with patient blank serum samples (3 males and 3 females). The candidate
matrix, the patient blank serum samples, and the mixture were
then extracted. The response ratio (analyte over internal standard) of
each mixture was then monitored. The candidate matrix was accepted if
the response ratio of each mixture was within 20% of the theoretical
response ratio.

2.4.2. Interferences
Interferences from both endogenous and exogenous sources were
investigated. Four endogenous conditions were evaluated for each
analyte: lipemic (L index: 689), hemolyzed (H index 303), icteric (I
index: 26), and uremic (blood urea nitrogen: 32 mg/dl) samples. This
was investigated at 2 analyte concentrations (1.0 μg/ml and 10.0 μg/
ml) in spiked charcoal stripped serum, by mixing the spiked serum 1:1
with the analyte-free interferent samples. Significant interference was
determined if the response ratio of any mixture was> 20% different
than the theoretical response ratio.
To evaluate the effects of potential exogenous interferences, commercial
controls [Lyphochek Quantitative Urine Quality Control, Liquid
Assayed Multiqual, Liquichek Immunoassay Control, and Liquichek
Urine Toxicology Quality Control (Bio-Rad)] were extracted. The controls
contained> 100 therapeutic drugs and common endogenous
substances. Interference was determined by observing the chromatograms
for any peaks with similar retention times as the analytes.

2.4.3. AMR
The AMR of each analyte was evaluated at 8 levels in pooled patient
blank serum: blank, 0.2, 0.3, 0.6, 1.3, 2.5, 5.0, and 10.0 μg/ml for
voriconazole, posaconazole, isavuconazole, itraconazole, and hydroxyitraconazole.
For fluconazole, the 8 levels were a blank, 0.5, 0.9, 1.9,
3.8, 7.5, 15.0, and 30.0 μg/ml. Each level was extracted in triplicate
and analyzed. The AMR was deemed acceptable if each level demonstrated
a recovery within 100 ± 20%, CV ≤20%, and a signal-tonoise
ratio > 10.

2.4.4. Carryover
Sample carryover was evaluated for each analyte by performing the
following sequence: low concentration sample ➔ a high concentration
sample ➔ 2nd injection of the low concentration sample in serum. The
high concentration sample was approximately double the upper limit of
quantitation (LOQ). The low concentration samples were approximately
0.3 μg/ml for all analytes except fluconazole, which was 1.0 μg/
ml. The samples were extracted and analyzed in triplicate. Sample

carryover was deemed acceptable if the 2nd injection of the low concentration
sample was within 20% of the first injection. Furthermore,
the mean of the 2nd low injection versus the mean of the initial injection
must be< 3 times the standard deviation of the initial injections.

2.4.5. Stability
The non-extracted stability of each analyte was evaluated at 3 different
storage conditions (ambient, 4 °C, and −20 °C) at 3 concentrations
in triplicate (i.e., 3 different storage containers). The levels for all
analytes except fluconazole were 0.3, 1.4, and 2.5 μg/ml, in pooled
patient serum. The concentrations used to assess fluconazole's stability
in pooled-patient serum were 0.9, 4.2, and 7.5 μg/ml. When the designated
time point was reached, the sample was stored at −70 °C until
analysis (Table 2). When last time point was reached (60 days from
time zero), all of the non-extracted samples were analyzed in a single
batch. The extracted stability was investigated by preparing the samples
and placing them into the instrument's autosampler (4 °C). The samples
were analyzed as each time point was reached (Table 2). The analyte
was considered stable at the specific time point if the mean value at the
experimental condition was within 20% of the mean value at day-0.

2.4.6. Precision
Utilizing the Clinical& Laboratory Standards Institute (CLSI) EP10-A3
guideline each analyte's intra-assay and total precision were determined at 3
levels in spiked, pooled patient serum. This consisted of extracting and
analyzing samples in the following sequence twice a day for 5 days:
mid ➔high➔low➔mid➔mid➔low➔low➔high➔high➔mid. For
fluconazole, the low, mid, and high mean levels were 1.9, 8.8, and 15.2 μg/
ml, respectively. For the other analytes, the low, mid, and high mean levels
were approximately 0.7, 3.0, and 5.0 μg/ml, respectively. The imprecision
was deemed acceptable if the highest total CV obtained was< 20%.

2.4.7. Method comparison
Approximately 40 samples per analyte were split for dual laboratory
comparisons. These samples consisted of leftover specimens (serum and
lithium heparin plasma), from patients who were prescribed an azole
medication, and spiked serum samples. For all of the analytes except
isavuconazole, an independent clinical reference laboratory, utilizing
LC-MS/MS methodology, was the comparator. Isavuconazole was not
available at that reference laboratory; therefore, the Fungus Testing
Laboratory at the University of Texas Health Science Center at San
Antonio was utilized. The Fungus Testing Laboratory's quantitation was
LC-MS-based. The results between our laboratory and the comparator
laboratories were plotted via Deming regression and bias plots. The
slopes, intercepts, correlation coefficients (R), and average percent
biases were evaluated.

2.4.8. Statistics
Statistical analysis was aided by EP Evaluator (; Data
Innovations) and Microsoft Excel.

Year prepared: