Internal Standard Comparison
In the biphenyl separation method, all analytes performed best with their matched ISs, though RSD and percent recovery remained in the acceptable range for several substitutions (Table 3). For 17OHP4, the T-13C3 and AE-13C3 ISs yielded good percent recoveries (98.0 % and 108.9 % respectively) and RSDs (4.3 % and 1.4 % respectively), though measurements of endogenous concentrations decreased, especially at lower concentrations. For calculating T, the nearby 17OHP4-13C3 IS peak performed well as an IS substitute (percent recovery = 110 % RSD = 3.7 %). For AE, the T-13C3 and 17OHP4-13C3substitutions showed good recovery (99.4 % and 107 % respectively) and consistent endogenous AE measurements (Figure 3), but RSDs were not considered acceptable (18.9 % and 21.0 % respectively). For quantifying P4, AE-13C3 was a good IS substitute, with acceptable percent recovery (109 %), RSD (5.3 %), and P4 measurements consistent with concentrations determined with the matched IS. When quantified with T-13C3 and 17OHP4-13C3, P4 concentrations were 28.1 to 67.0 % lower than the concentration determined with P4-13C3 (Figure 3).
In the C18 method, the four 13C3 ISs outperformed the two d 4 standards for quantification of every adrenal steroid, except ALD (Figure 4). The B-d 4 peaks had small peak areas compared to the other analytes, which likely caused higher variability in the IS ratios used to calculate concentrations. This is likely reflected by high RSD values (18.5 % - 23.0 %) when using B-d 4 as an IS for other adrenal steroids (Table 4). As a target compound, B was best quantified with 13C3 ISs, which were characterized by larger, more distinct peaks than B-d 4. In comparison to an RSD of 18.8 % when calculated with B-d 4, RSDs for B fell below 4.5 % when calculated with S-13C3, F-13C3, E-13C3, and 11DOC-13C3. Though ALD performed best with ALD-d 4, this IS also had a relatively small and broad peak, leading to more variability when used as an IS substitute, with analyte RSDs ranging from 14.8 % to 16.5 %. Except for quantifying 11DOC with S-13C3, the accuracies and RSDs calculated for E, F, B, S, 11DOC, and DHEA using13C3 ISs all fell within acceptable ranges.
For the most part, individual differences were the primary drivers in analyte concentration. However, even when percent recovery and RSD fell within acceptable thresholds, IS substitutions altered observed analyte concentrations (Figure 4). In the female sample, concentrations of B were close to the RL, 0.869 ng/g. When calculated with B-d 4, concentrations fell below RL but exceeded RL when calculated with S-13C3, 11DOC-13C3, and E-13C3. In both methods, substitutions that failed to meet acceptable RSD and/or percent recovery thresholds altered concentrations substantially. In the C18 method, concentrations varied from those calculated with matched ISs by as much as 63.3 %. On average, good substitutions varied much less from matched IS analyte concentrations (mean = 15.5 % median = 11.3 %) than poor substitutions (mean = 31.7 % median = 30.1 %).