28 / 116
CHON-002
Capillary zone electrophoresis (CZE) was used for the analysis
of unsaturated CS disaccharides. Response factors were used
for the disaccharides (Di-0S, Di-4S, and Di-6S), for quantitation
of disaccharides in the samples. CS was calculated as the
amount of unsaturated disaccharides released by
chondroitinase ABC divided by the ratio of 1.036 (ratio of MW
of disaccharides and MW of disaccharide unit in CS). The
method was run ay 60C to avoid peak splitting of the
disaccharides.
Size exclusion chromatography (SEC) was used to analyze CS
using an RI detector. Dextrans (MW: 10K, 20K, 40K, 70K, and
150K g/mol) were used for calibration standards. The
relationship of the respective dextran and log MW was
evaluated. The separation was carried out using 3 columns in
series. Chondroitinase ABC was used to determine the
composition and content of the disaccharides. The authors
state they followed a published method with modifications.
The authors state the digestion time and inactivation method
used was sufficient but did not provide any supporting data.
The mixtures were then analyzed using CZE.
Safety Review
Analytical Range:
The range of the method was not defined by the authors. The
authors used statistical regression for linearity. Calibration
curve was evaluated for CS-A (20-200 mg/L). Routine samples
were evaluated from 20-100 mg/L. There is no mention of why
the curve range was varied.
LOQ:
Calculated as 3 times the LOD. LOD was calculated from the
calibration equation. LOD corresponds to the concentration of
CS which gives the minimum detectable zone length equal to
one second. The authors report LOQ at 9mg/mL, I calculated
the LOQ at 8 mg/mL based on the information given by the
authors. 8 mg/mL is at the 1% level. Its unclear if the LOD
concentration was properly evaluated since the calibration
curve used to extrapolate LOD concentration was much higher
(20-200 mg/L) than the lowest standard concentration (LOD
estimate implied at 3 mg/L). P. 59
Accuracy/Recovery
Assessed by the average of three replicates of sample spiked
with CS-A at the 50% and 100% level. 20 tablets composited,
weight equivalent to one tablet was used. Accuracy was
reported as 99.7 +/- 2.6% (97.1 – 102.3); therefore, not
meeting the SMPR requirement of 98-103% for >10% w/w.
Precision (RSDr)
Assessed by determination of six replicates in two different
samples. 20 tablets composited, weight equivalent to one