63 / 71
662
S
chneider
&
A
ndersen
:
J
ournal of
AOAC I
nternational
V
ol
. 98, N
o
. 3, 2015
I. Quantification
(
a
)
Internal standards
.—MG-D5 is used as the internal
standard for both MG and BG. All other analytes have their
corresponding isotopically labeled internal standards (LMG-D5,
CV-D6, and LCV-D6) incorporated into the method.
(
b
)
Calibrationcurves
.—Foragivenanalyte,thequantification
ion peak area ratios for analyte:corresponding internal standard
(y-axis) are plotted versus concentration (x-axis) for the matrix
calibrant samples. The resultant linear relationship (R
2
≥ 0.95) is
used to calculate the concentration of the analyte in test samples
using the equation y = mx + b, where m is the slope and b is the y
intercept of the calibration curve
.
J. Identification
Acceptable identification of an analyte can be determined
according to either EU (10) or FDA (11) criteria. An analyte is
considered to be present in a sample when:
(
a
) Its chromatographic retention time is ±2.5% (EU) or
±5% (FDA) of the average retention time for the corresponding
non-zero matrix calibrant samples
(
b
) Its peak area ratio of qualitative ion:quantification ion is
within the acceptable range of the corresponding average ratio
for the non-zero extracted matrix calibrant samples. For the EU,
this range is dependent on the peak ion ratio, ranging from ±20
to ±50 relative % (10). For the FDA, the acceptable range is
±10% absolute (11).
(
c
) The S/N must be ≥3 for both SRM transitions.
Results and Discussion
Method Performance
AOAC First Action Method
2012.25
proved to be fairly
straightforward for participants, and all were able to complete
the study and submit the required data. Participants were
requested to perform their three sets of extractions and analyses
within 3 weeks. Two laboratories completed their analyses in the
first week from sample receipt, and the majority of laboratories
completed or initiated their sample analysis within the second
week. One laboratory completed the analyses in the fifth week.
Performance of the method was evaluated based on the results
from all 14 laboratories with regard to quantification and
identification of each of the five analytes. Overall, the results
of the study were excellent, with trueness generally ≥90% and
RSDr generally ≤10%, with HorRat <1.
Ruggedness
A few deviations from, and variations within, the study
protocol were noted by study participants and served to illustrate
the ruggedness of the method. Deviations included differences
in standard and sample storage temperatures, varying speeds of
centrifugation, and differences in the pore size and material used
for the final extract filtration. One laboratory stored standard
solutions at –6°C, and three laboratories stored tissue samples
at –20, –50, or –70°C, instead of the recommended –20°C for
standards and –80°C for tissue samples. Six laboratories did
not centrifuge samples at 2000 ×
g
and eight laboratories did
not microcentrifuge at 20000 ×
g
, but instead used a range of
speeds (700 to 6000 ×
g
for centrifuge and 10000 to 30 000 ×
g
for microcentrifuge), which were likely a function of available
laboratory equipment. Final filtration was generally completed
with PVDF syringe filters with 0.45 μm pore size as indicated in
Method
2012.25
. Three laboratories reported that PVDF filters
with 0.22 μm pore size were used, and one laboratory reported
that 0.45 μm PTFE filter vials were used. None of the variations
for storage temperature, centrifuge speed, or filtration appeared
to have influenced method performance.
A variety of liquid chromatographic systems, including three
ultra-HPLC (UHPLC) systems, were used in this study. Method
2012.25
and the study protocol provided for HPLC conditions,
however, participants were given flexibility to design their
own chromatographic separation to ensure that all analytes
were retained sufficiently on their column. Variations were
observed for injection volume, mobile phase gradient, flow
rate, and column temperature. The primary concerns voiced by
participants during the method familiarization phase involved
the high percentage of acetonitrile (approximately 99% by
volume) in reconstituted samples compared to the initial
mobile phase composition of 40% acetonitrile. Participants
were encouraged to adjust the gradient used and/or, given their
sensitive instrumentation, decrease the injection volume in order
to ensure analytes were suitably retained on the chromatographic
column and detected. Five laboratories slowed down the initial
(0–1 min) 40 to 90% acetonitrile gradient in Method
2012.25
(Table 1) by holding the initial acetonitrile composition at 10
or 20% for 0.5 to 3 min, then ramping up to 90% acetonitrile
over 2 to 12 min. One laboratory used the mobile phase gradient
described in Method
2012.25
gradient until 6 min, then dropped
Table 2. MS/MS parameters for the Waters Corp. Quattro
LCZ system
SRM,
m/z
Collision
energy, eV
Cone
voltage, V
Retention
time, min
MG
329
→
313
a
35
43
5.1
329
→
208
35
43
5.1
MG-D5
334
→
318
40
30
5.1
CV
372
→
356
a,b
40
25
5.6
372
→
251
b
35
25
5.6
CV-D6
378
→
362
40
25
5.6
BG
385
→
341
a
35
35
6.0
385
→
297
50
35
6.0
LMG
331
→
239
a
25
25
7.8
331
→
316
20
25
7.8
LMG-D5
336
→
239
25
25
7.8
LCV
374
→
358
a
30
25
7.9
374
→
239
25
25
7.9
LCV-D6
380
→
364
35
25
7.9
LBG
387
→
342
a
30
25
10.9
387 → 281
30
25
10.9
a
Product ion transition used for quantification.
b
An additional transition (
m/z
372 → 340) was used by five laboratories
for either quantification or for identification.