Status Healthy volunteers Liver post-transplant Inpatient Outpatient Kidneypost-transplant Inpatient Outpatient Range of dlgoxin equIvalents, nmol/L 0.26-0.37 DUS posftlvtty rate 3 of 22 (14%) 10 of 14 (71%) 2 of 14 (14%) 5 of 8 (63%) 1 of 8 (13%) P valuV Uver KIdney Inpatient Outpatient InpatIent Outpatient -0.001 0.71 0.016 0.72 0.27 0.26-0.79 -0.003 0.82 0.012 0.27-0.33 0.32 0.71 0.28-0.46 0.059 Table 1. DUS Positivity Rates In Heaithy Volunteers and Transplant Recipients 172 CLINICAL CHEMISTRY, Vol. 36, No. 1, 1990 a Fishers exact test, one-tailed (P >0.05 suggests no significant difference between groups). kidney- and liver-transplant recipients, both as inpatients and outpatients. Sera from digoxin-free transplant patients and healthy volunteers, as well as protein-free filtrates from these groups prepared by ultraliltra. tion (Centrifree Micropartition Sys- tem; Amicon, Danvers, MA), were an- alyzed for DUB with a fluorescence polarization immunoassay, detection limit, 0.26 nmoIIL (Digoxin II; Abbott Laboratories, North Chicago, IL). We eyimined the potential cross-re- activity of putative DUB materials with the digoxin antibody in this assay. Sub- stances investigated (5) included nones- terified fatty acids (palmitic, palmi- toleic, stearic, linoleic, linolenic, oleic, and arachidonic), cortisol, and lysophos- phatidylcholine (Sigma Chemical Co., St. Louis, MO). The compounds, at their maximum physiological concentrations (6), were added to DUB-negative serum. As Table 1 shows, the DUB positivity rate in healthy volunteers was signifi- cantly lower than in either the liver- or kidney-transplant inpatient groups and significantly lower in liver-transplant outpatients than in liver-transplant in- patients. The DUB positivity rates in healthy volunteers and either outpa- tient group were not statistically dii. ferent. Moreover, no differences were seen in the rates between liver- and kidney-transplant inpatients or liver- and kidney-transplant outpatients. Al- though no conclusive-i.e., statistically significant-difference in DUB positivity rates was demonstrated between inpa- tient and outpatient kidney-transplant recipients at the 95% confidence level (P was 0.059), the results strongly suggest that detectable DUB is found more fre- quently among the inpatient than the outpatient groups. The statistical signif. icance of kidney-transplant inpatient and outpatient DUB positivity rates will be re-examined with an increased num- ber of kidney-transplant recipients. None of the proposed DUB candi- dates that we tested exhibited cross- reactivity in the assay. Thus, these compounds do not appear to be respon- sible for the DLIS activity detected in this study. The protein-free filtrates of serum from healthy volunteers as well as those from the two transplant groups were free of DUB activity, re- flecting the strongly protein-bound na- ture of this immunoreactive material. These results demonstrate an in- creased rate of DUB positivity in kid- ney- and liver-transplant inpatients vs the rate in comparable outpatient populations or healthy volunteers, as measured by our assay. Further, they suggest a possible role for this type of analysis in monitoring recovery from these transplant procedures or in cor- recting the underlying clinical condi- tions for which organ transplant was originally performed. References 1. Graves SW, Brown BA, Valdes R. Digoxin like substances measured in pa- tients with renal impairment. Ann Intern Med 1983;99:604-8. 2. Nanja AA, Greenway DC. Falsely raised plasma digoxin concentrationsin liver disease.Br Med J 1985;290:432-5. 3. Graves SW, Williams GH. Endogenous digitalis-like natriuretic factors [Review]. Annu Rev Med 1987;38:433-44. 4. Dasgupta A, Chou H, Saldana S. Dem- onstrationof digoxin like immunoreactive substancesin post liver transplant patients with positivefluid balance [Abstract].Clin Chem 1989;35:1144. 5. Kelly RA, O’Hare DS, Mitch WE, Smith TW. Identification of Na,K-ATPase inhibi- tors in plasma as nonesterifledfatty acids and lysophospholipid. J Biol Chem 1986;261:11704-11. 6. Tietx NW. Textbookof clinical chemistry. Philadelphia:WE Saunders, 1986;1810-56. Amitava Dasgupta1’2 David A. Dennen’ Sylvia Saldana’ Homer Chou’ Ronald W. McLawhon”2 Clin. Labs. and 2Dept. of Pathology University of Chicago Hospitals 5841 S. Maryland Ave., Box 146 Chicago, IL 60637 Effect of Sex-Hormone-Binding Giobuitn on No-Extraction Immunoassays for Testosterone To the Editor: The report of Masters and H#{228}linel (1) concerning errors in the radioim- munometry of testosterone (T) in so- rum interested us. We also have inves- tigated the influence of sex-hormone- binding globulin (SHBG) on the measurement of T in the Immunchem assay (2). We proved this effect by use of a slightly different method: Two speci- mens of pooled sera from women with low endogenous concentrations of T (<0.7 nmol/L) and SHBG concentra- tions of 253 and 10 nmol/L were used for analytical-recovery studies. A T stock solution with [icr, 2a(n)-3H]T as tracer was added to both sera, giving a final total T concentration of 87 njnol/ L. Then serial dilutions were per- formed with the same pooled sera. Beta-counting of the tritiated tracer confirmed the uniformity of the T con- tent in the different samples. Analyti- cal recovery of the added T was then measured by use of the Immunchem no-extraction RIA assay and an ex- traction RIA kit supplied by Baxter Dade. Figure 1 (top) illustrates the influ- ence of SHBG on the determination of T in the no-extraction assay (assay no. 1). The quantity of T measured in the serum with high SHBG content is di- minished by >50% in comparison with the serum with low SHBG content. This interference can be seen through- out the whole range of T concentra- tions. In contrast to this, the SHBG effect cannot be detected in the extrac- tion-assay (Figure 1, bottom). The di-