ESPE Abstracts

Background: We have previously shown radioimmunoassay (RIA) and gas chromatography-tandem mass spectrometry (GC-MS/MS) to be comparable when analyzing estradiol and testosterone concentrations in prepubertal and pubertal children. However, the reliability for steroid hormone determination with RIA in umbilical cord blood is not known. In general, older studies using RIA show higher values of testosterone, than more recent ones using MS/MS.

Patients and methods: Umbilical cord blood was collected from 236 infants (133 boys, 103 girls). Estradiol and testosterone concentrations were analyzed using both RIA (Spectria, Orion Diagnostica, Espoo, Finland) and GC-MS/MS (Agilent, Montreal, Canada). For estradiol the limit of detection (LOD) was 9 pmol/l for RIA and 2 pmol/l for GC-MS/MS. With RIA the intra-assay coefficient of variation (CV) was <8% whereas the interassay CV was <13% for concentrations ≥40 pmol/l. With GC-MS/MS the intra-assay CV was <4% and the interassay CV was 7% for concentrations ≥300 pmol/l. For testosterone the LOD was 0.1 nmol/l for both RIA and GC-MS/MS. With RIA the intra- and interassay CV was <7% for concentrations of ≥0.9 nmol/l. With GC-MS/MS the intra-assay CV was <4% and the interassay CV was 16% for 0.2 nmol/l and ≤8% for concentrations of ≥5 pmol/l.

Results and discussion: For estradiol, there was a good correlation between RIA and GC-MS/MS (r=0.92, P=0.000). The concentrations of estradiol were similar, RIA (range 413–71 765, median 14 044 pmol/l) vs GC-MS/MS (range 499–68 825, median 15 754 pmol/l). For testosterone, there was also a correlation between RIA and GC-MS/MS (r=0.44, P=0.000). Umbilical cord blood testosterone concentrations determined with RIA were higher (range 0.7–15.0, median 3.3 nmol/l) than with GC-MS/MS (range 0.1–8.1, median 0.3 nmol/l), with linear fit functions; RIA=0.99×GC-MS/MS +3.10 nmol/l. Although RIA is as reliable as GC-MS/MS when analyzing serum from prepubertal and pubertal children, that is not the case in umbilical cord blood. We speculate that this is due to cross reactivity with other androgens, testosterone metabolites, or estrogens with the RIA.

Conclusion: We conclude that RIA is sufficient for determination of estradiol, but not testosterone, in umbilical cord blood.