Biomedical Engineering Reference
In-Depth Information
Improper Use of Internal Standard
Initially, vidarabine was used as the internal standard for penciclovir in a method
based on mixed-mode strong cation exchange (MCX) solid-phase extraction due to
their similar properties, particularly hydrophobicity (Fig.
14a
). Quite stable IS
responses were obtained in a run consisted mainly of CS and QC samples with a CV
of 13.02 % (Fig.
14b
). However, 43 % of the CS and QC samples did not meet the
acceptance criterion of accuracy. On the other hand, when penciclovir-d
4
was used
as the internal standard, all the CS and QC samples met the acceptance criterion in
accuracy though the IS responses were more variable (the CV in IS responses was
23.81 %, Fig.
14c
).
A careful examination of the extraction procedure revealed the root cause, i.e.,
improper use of the internal standard. Specifically, despite the overall similarity in
hydrophobicity vs. pH characteristics, there is a relatively larger difference in hydro-
phobicity between vidarabine and penciclovir in acidic conditions (pH 1-2), in
which the mixture of a sample and the internal standards was loaded onto an MCX
plate and the loaded MCX plate was washed. For vidarabine, its hydrophobicity is
relatively less variable than that of penciclovir in this pH range. Because of this dif-
ference, minor change in pH caused differential recovery variation for the analyte
and its internal standard, i.e., potential impact on the accuracy of quantitation.
Solubility or Stability of Internal Standard
The solubility or stability of an internal standard in IS working solution could cause
IS response variations. Sometimes, an issue of solubility could appear as an issue of
stability [
42
]. For example, high interbatch IS response variation was observed for
a method of raloxifene glucuronides based on protein precipitation by acetonitrile
containing deuterated raloxifene glucuronides. However, internal standard responses
within a batch were relatively stable and no such variations in the responses of the
corresponding analytes were observed. It was further noticed that the internal stan-
dard responses of a batch appeared to be correlated to the operation speed of the lab
technician who performed the batch. The faster a lab technician's speed was, the
higher the IS responses of his run were.
Initially, it was suspected that the deuterated raloxifene glucuronides might be
unstable in acetonitrile. It was therefore replaced by methanol, despite the fact that
acetonitrile is usually a desirable solvent for protein precipitation. This suspicion
appeared to be supported by the results of a comparison test using both solvents
(Fig.
15
). Finally, the root-cause was identified as an issue of solubility by extensive
troubleshooting, instead of stability. Specifically, raloxifene glucuronides and their
deuterated internal standards are very hydrophilic and not very soluble in ace-
tonitrile. The seeming “stability” issue disappeared when 10 % of water was added
to acetonitrile. As to why the analytes (raloxifene glucuronides) did not show simi-
lar variable responses between batches as the corresponding deuterated internal
