Compound W, a 3,3=-diiodothyronine sulfate (T 2 S) cross-reactive material in maternal serum, was found to be useful as a marker for fetal hypothyroidism. In the present report, we explored its biochemical properties and studied its concentrations in cord and in maternal serum obtained from various gestational periods and at term from different continents. Mean W concentrations, expressed as nmol/L T 2 S-equivalent, in maternal serum during gestation showed a moderate increase at 20 -26 wk (1.57 nmol/L) and an accelerated increase to 34 -40 wk (3.59 nmol/L). The mean serum level was relatively low in nonpregnant women (0.17 nmol/L). Compound W levels in cord and maternal serum at term were not significantly different among samples obtained from Taiwan compared with samples from the United States. The mean cord serum "corrected" (by hot acid digestion) concentrations of W were significantly higher than maternal serum concentrations at birth and were also higher in venous than in paired arterial samples, suggesting that the placenta may play a role in its production. We compared a total of 45 iodothyronine analogs by antibody, gel filtration, and HPLC chromatographic studies and found only one compound, N,N-dimethyl- E mploying a sensitive RIA for T 2 S, we found high concentrations in fetal and pregnant serum in humans (1). We further identified that this T 2 S-cross-reactive material is not authentic T 2 S and does not co-chromatograph with synthetic T 2 S on HPLC. Thus, the name Compound W was coined to represent this material in fetal and maternal serum (1). Levels increased with the progression of pregnancy and rapidly peaked before parturition. At delivery, a 20-fold increase in "T 2 S" was found compared with nonpregnant women and returned to baseline within 7-10 d. Serial measurements of serum W in pregnant women have been found to be useful as a noninvasive technique for the diagnosis of fetal hypothyroidism (2,3).The present study was undertaken to determine the normal distribution of elevated Compound W levels in different stages of pregnancy as well as from various geographic areas and correlate between paired maternal and cord blood samples, and to explore its origin and chemical structure.
MATERIALS AND METHODS
T 2 S RIA. 3,3=-T 2 S and [125 I]T 2 S were prepared by the method of Eelkman-Rooda and co-workers (4,5). T 2 S was further purified and quantitatively recovered by reverse-phase HPLC with a preparative column, as described previously (1,2,4).The T 2 S RIA procedure was a modification of the RIA described previously (1,2). Serum samples (0.2-1.0 mL) were extracted with 2 vol 95% ethanol (final concentration, 63%) before assay. Preliminary experiments showed that the extraction efficiency of T 2 S in serum exceeded 96%. Final T 2 S concentrations were not corrected for recovery. The lower limit of detection of the assay was 3.3 fmol (2 pg), or 33.1 pmol/L in a 300 L ethanol extract of serum. Intra-assay variations were 1.9 -9.1% and interassay variations were 6.0 -19.5%, depending on ...