Analog or Homolog Internal Standards

Poor precision due to matrix effect variability within a multi-sample analytical run, that is not accounted for by the SIS, is an example of what is often referred to as curve divergence . Curve divergence can result from response drift or from variable matrix effects on the analyte and/or SIS. A simple test to determine whether the divergence is associated with the SIS is to perform linear regression on the raw data ignoring the SIS (i.e. simply analyte response vs concentration rather than ratio of analyte/SIS responses vs concentration). If the precision and accuracy for the calibration curve improve, then modifications to the method and/or finding another SIS will be required. 

If the particular analog happens to be a structural isomer (Table 2.1), chromatographic resolution is necessary to distinguish the analytical signals for analyte and SIS (identical molecular masses) unless MS/MS is used and the isomers yield different product ions. Since the molecular mass of a homologous structure differs from that of the analyte by at least 14 Da, there is seldom any problem of cross-contributions arising from overlap of the isotopolog distributions (Section 8.5.2c). 

A different potential problem with use of an analog or homolog of the analyte as SIS arises in the case of biomedical analyses in which metabolites can have a molecular mass indistinguishable (at unit mass resolution) from that of the SIS (Matuszewski 1998 Jemal 2002). This problem is more fully discussed in Sections 9.4.7b and 10.2.9c in the context of analysis of incurred samples during method validation or to assess a previously validated method. 

---