FREE-THYROXINE MEASURED IN UNDILUTED SERUM BY DIALYSIS AND ULTRAFILTRATION - EFFECTS OF NONTHYROIDAL ILLNESS, AND AN ACUTE LOAD OF SALICYLATE OR HEPARIN

In vitro dilution of serum during processing of a free T4 assay explai ns to some extent the divergent results obtained in non-thyroidal illn ess. If serum from such patients contains low affinity T4 protein bind ing inhibitors, as has been suggested, in vitro dilution will result i n spuriously reduced serum free T4 measurements. If these inhibitors c ross the dialysis membrane in an equilibrium dialysis assay, their inh ibitory effect will be weakened, and in vitro free T4 levels will decr ease, even in undiluted serum. In contrast, ultrafiltration methods on undiluted serum seem accurate. We have compared a new, commercially a vailable dialysis technique with an in-house ultrafiltration method fo r free T4 measurements in undiluted serum. Control subjects (n = 41) h ad 14% higher free T4 (P < 0.02) by ultrafiltration. Non-thyroidally i ll patients not receiving glucocorticoids or dopamine (n = 54) had una ltered free T4 levels, 28.4 +/- 10.3 pmol/l (dialysis) and 31.0 +/- 10 .3 pmol/l (ultrafiltration). Dopamine infusion in somatic ill patients (n = II) resulted in reduced free T4 in both assays but only signific antly for dialysis, and subjects with familial dysalbuminemic hyperthy roxinemia (n = 8) had unaltered free T4 levels in both assays. Salicyl ate (1.5 g) given orally 09:00 h. (n = 5) resulted within 30 min. in i ncreased (P < 0.01) free T4 as measured by both techniques, although m ore pronounced and sustained as measured by ultrafiltration. Serum TSH decreased concomitantly (P < 0.01). These findings were confirmed whe n salicylate was administered at 13:00 h. (n = 8). The dialysis proced ure resulted in a decrease in serum salicylate of 14% (P < 0.001). Hep arin (5,000 units) given intravenously at 09:00 h. (n = 7) tended to i ncrease (P < 0.10) free T4 after 15 min. in both assays, and decreased TSH (P < 0.001). In conclusion the two methods demonstrated qualitati vely, but not always quantitatively quite similar changes in situation s with (1) the possible existence of circulating low affinity T4 bindi ng inhibitors (somatic illness), (2) pituitary suppression (dopamine), (3) protein binding capacity changes (dysalbuminemic hyperthyroxinemi a), and (4) acute protein binding displacement by salicylate or hepari n. The dialysis method seems therefore to be a valuable reference assa y for the measurement of serum free T4, although from a theoretical po int of view it seems not ideal.