Amitava Dasgupta, PhD, 1 David A. Biddle, MD, 1 Alice Wells, MT(ASCP), 1 and Pradip Datta, PhD 2. Abstract

Transcription

1 Clinical Chemistry / INTERFERENCE F CHAN SU INDIGXIN IMMUNASSAYS Positive and Negative Interference of the Chinese Medicine Chan Su in Serum Digoxin Measurement Elimination of Interference by Using a Monoclonal Chemiluminescent Digoxin Assay or Monitoring Free Digoxin Concentration Amitava Dasgupta, PhD, 1 David A. Biddle, MD, 1 Alice Wells, MT(ASCP), 1 and Pradip Datta, PhD 2 Key Words: Chan Su; Digoxin; Fluorescence polarization; Chemiluminescence; Microparticle enzyme immunoassay Abstract An over-the-counter Chinese medicine, Chan Su, is used as a cardiotonic agent. We demonstrated significant digoxin-like immunoreactivity in various organic and aqueous extracts of Chan Su. For example, when a 20-µL aliquot of an aqueous extract of Chan Su powder (1 mg/ml) was added to a 2-mL aliquot of a drug-free serum, the observed digoxin-like immunoreactivity was 2.76 ng/ml (3.53 nmol/l) digoxin equivalent using the fluorescence polarization immunoassay (FPIA). The magnitude of interference was much lower (0.94 ng/ml [1.20 nmol/l]) with the microparticle enzyme immunoassay (MEIA), and no interference was observed with the chemiluminescent assay (CLIA). We also observed a significant positive interference of the extract with the serum digoxin measurement using FPIA. In contrast, we observed a negative interference (falsely lowered digoxin concentration) of the extract in the serum digoxin measurement with the MEIA. The extract had no effect on the serum digoxin measurement with the CLIA. By taking advantage of the high protein binding of Chan Su and only 25% protein binding of digoxin, we further demonstrated that positive interference of Chan Su in the FPIA and negative interference of Chan Su in the MEIA of digoxin could be eliminated by monitoring the free digoxin concentration. Traditional Chinese medicines are readily available without prescription from herbal stores in local Chinese shopping areas. ne such Chinese medicine is Chan Su, which is prepared from the dried white secretion of the auricular glands and the skin glands of Chinese toads (Bufo melanostictus Schneider or Bufo bufo gargarzinas Gantor). Chan Su is also a major component of traditional Chinese medicines Liu-Shen-Wan and kyushin. 1,2 These medicines are used for the treatment of conditions such as tonsillitis, sore throat, furuncle, and palpitations because of their anesthetic and antibiotic action. Traditional use of Chan Su given in small doses also includes stimulation of myocardial contraction, anti-inflammatory effects, and pain relief. 3 The cardiotonic effect of Chan Su is due to its major bufadienolides, such as bufalin, cinobufagin, and resibufogenin. 4 Bufalin is known to block vasodilation and increase vasoconstriction, vascular resistance, and blood pressure by inhibiting Na+,K+-ATPase. 5 Bufalin also induces apoptosis in human leukemia HL60 cells and human tumor cells. 6,7 A recent report 8 also indicated that administration of bufalin in male rats diminished the luteinizing hormone response to gonadotropin-releasing hormone and the secretion of testosterone. The authors concluded that Chan Su might possess hypogonadal effects in male rats. 8 Although the general population considers traditional Chinese medicines safe, toxic effects of Chan Su have been documented in the literature. At high dosages, Chan Su causes cardiac arrhythmia, breathlessness, seizure, and coma. Death of a Chinese woman after ingestion of Chinese herbal tea containing Chan Su has been reported Am J Clin Pathol 2000;114: American Society of Clinical Pathologists

2 Clinical Chemistry / RIGINAL ARTICLE Structural similarities between bufadienolides and digoxin account for the toxic effects and serum digoxin-like immunoreactivity of Chan Su. Fushimi and Amino 10 reported a serum concentration of 0.4 ng/ml (0.51 nmol/l) in a healthy volunteer after ingestion of kyushin tablets containing Chan Su as the major component. Panesar 11 reported an apparent digoxin concentration of 0.88 ng/ml (1.124 nmol/l) in healthy volunteers who ingested Lu- Shen-Wan pills. The author used the fluorescence polarization immunoassay (FPIA; Abbott Laboratories, Abbott Park, IL) and the FLX/TDx analyzer (Abbott Laboratories) for the study. Chan Su is the major component of Lu-Shen- Wan pills. An apparent digoxin concentration of 4.9 ng/ml (6.3 nmol/l; measured by the FPIA) was reported in 1 woman who died of ingestion of Chinese herbal tea containing Chan Su. 9 The availability of Chinese medicines in the United States is increasing, and the use is no longer limited to the Asian population. A recent study concerning the prevalence of unconventional therapies found that 1% of the general population in the United States has used herbal medicine. 12 f those who also saw a physician, almost 75% did not inform the physician of their alternative medicine intake. Because ingestion of Chan Su may cause digoxin-like immunoreactivity in serum, we studied the effect of Chan Su in serum digoxin measurements. We used the FPIA and the microparticle enzyme immunoassay (MEIA; Abbott Laboratories) for digoxin to study this potential interference because these assays are used widely in clinical laboratories for routine monitoring of digoxin levels. We also compared our results with a new chemiluminescent assay (CLIA) for digoxin available from Bayer Diagnostics, Tarrytown, NY. We now report negative interference of Chan Su in serum digoxin measurement by the MEIA and positive interference in measurement by the FPIA. Materials and Methods Chan Su was purchased from a Chinese herbal store in Houston, TX. High-performance liquid chromatography grade methanol, chloroform, ethyl acetate, and acetone were obtained from Aldrich Chemical, Milwaukee, WI. The FPIA for digoxin was obtained from Abbott Laboratories, and the assays were run on an FLX/TDx analyzer. The MEIA for digoxin also was obtained from Abbott Laboratories, and the assays were run on an AxSYM analyzer (Abbott Laboratories). The CLIA for digoxin was run using an ACS:180 Plus analyzer (Bayer Diagnostics). The ultrafiltration devices (Amicon Micropartition System, Amicon, Danvers, MA) contain an anisotropic, hydrophilic membrane (YMT, Taichung, Taiwan) with narrow pore size distribution. Typically, this device retains 99.9% serum protein and less than 5% L-thyroxine. The MEIA and CLIA for digoxin do not require sample pretreatment. n the other hand, the FPIA requires sample pretreatment in which 200 µl of serum is treated with 200 µl of 50% sulfosalicylic acid in 50% methanol. After protein precipitation, the specimen is centrifuged at a high speed, and the clear supernatant is analyzed for digoxin concentration. The assay is linear for a serum digoxin concentration up to 5.0 ng/ml (6.4 nmol/l), and the sensitivity of the assay is 0.2 ng/ml (0.26 nmol/l) of serum digoxin concentration. The MEIA is linear up to a serum digoxin concentration of 4.0 ng/ml (5.1 nmol/l), and the sensitivity of the assay is 0.3 ng/ml (0.38 nmol/l) of serum digoxin concentration. The CLIA is linear up to a serum digoxin concentration of 5.0 ng/ml (6.4 nmol/l), and the sensitivity of the assay is 0.15 ng/ml (0.19 nmol/l) of serum digoxin concentration. To find a proper solvent for extraction of Chan Su, we studied 5 solvents: methanol, chloroform, ethyl acetate, acetone and water (1:1 by volume), and deionized water. We added 2 ml of each solution to test tubes containing 2 mg of a fine powder of Chan Su. After vortex mixing, specimens were allowed to stand overnight at 4 C. Undissolved residue then was separated from the solvent by centrifugation. Then 20 µl of each extract was added to 2-mL aliquots of drugfree serum to study the digoxin-like immunoreactivity by FPIA, MEIA, and CLIA. In the next experiment, we studied the effect of Chan Su on serum digoxin measurements. Several serum pools were prepared from patients receiving digoxin. Then 6.0 and 2 µl of the aqueous extract of Chan Su were added to 4-mL aliquots of digoxin serum pools. We supplemented 2 different serum digoxin pools, pool 1 and pool 2, with aqueous extract of Chan Su for this part of the study. Serum digoxin concentrations were measured using all 3 digoxin immunoassays (FPIA, MEIA, and CLIA). In another experiment, we studied the serum protein binding of digoxin-like immunoreactive components of Chan Su extract. For this purpose, 4.0, 8.0, and 14.0 µl of aqueous extract of Chan Su were added to 4-mL aliquots of a drugfree serum pool (no digoxin). We measured total and free digoxin-like immunoreactivity using the FPIA and MEIA. The free digoxin-like immunoreactivity was measured in the protein-free ultrafiltrate. We prepared protein-free ultrafiltrate by centrifuging 0.8 to 1.0 ml of serum with the Amicon Micropartition System for 30 minutes at 1,500g. We also studied the possibility of eliminating the interference of Chan Su in serum digoxin measurement by the FPIA and MEIA by monitoring the free digoxin concentration instead of the total digoxin concentration. Again, we supplemented 4-mL aliquots of the serum digoxin pool with no Chan Su extract (control) and 4.0, 8.0, 14.0, and 2 µl of aqueous American Society of Clinical Pathologists Am J Clin Pathol 2000;114:

3 Dasgupta et al / INTERFERENCE F CHAN SU IN DIGXIN IMMUNASSAYS Chan Su extract. Total and free digoxin concentrations were measured by the FPIA. We repeated the same experiment with another digoxin pool and used the MEIA for measuring total and free digoxin concentrations. Statistical analyses were performed using the independent 2-tailed t test. We considered a difference statistically significant only at a 95% confidence interval or higher (P <.05). Results Chemical structures of bufalin, cinobufotalin (major components of Chan Su), and digoxin are shown in Figure 1. We observed significantly higher digoxin-like immunoreactivity in the acetone and water and the aqueous extracts of Chan Su compared with other extracts. For example, when 20 µl of aqueous extract was added to 2 ml of drug-free serum and subsequently analyzed for digoxinlike immunoreactivity, the observed concentration was 2.7 ng/ml (3.53 nmol/l) digoxin equivalent using the FPIA. The lowest immunoreactivity of 0.5 ng/ml (0.67 nmol/l) digoxin equivalent was observed when ethyl acetate extract was added to drug-free serum and subsequently analyzed by the FPIA. We observed significantly lower cross-reactivity of Chan Su with the MEIA and observed no cross-reactivity with the CLIA Table 1. We observed significant increases in serum digoxin concentration (positive interference) when the aqueous extract of Chan Su was added to aliquots of serum digoxin pools and subsequently analyzed by the FPIA. In contrast, we observed negative interference (falsely lower serum digoxin concentrations) when same specimens were analyzed by the MEIA for digoxin. Interestingly, we observed no interference of Chan Su in serum digoxin measurements using the CLIA. For example, when 20 µl of dilute aqueous extract of Chan Su was added to 4-mL aliquots of digoxin serum pool (5.0 µl of Chan Su extract per milliliter of serum), the observed digoxin concentration was 3.5 ng/ml (4.45 nmol/l) when measured by the FPIA. The original digoxin concentration was 2.0 ng/ml (2.53 nmol/l). Interestingly, the observed digoxin concentration was significantly decreased to 1.6 ng/ml (2.07 nmol/l) when measured by the MEIA. The CLIA was free of this type of interference because the serum digoxin concentration was 2.0 ng/ml (2.56 nmol/l) when measured by the CLIA Table 2. The control value was 2.0 ng/ml (2.50 nmol/l). The components of Chan Su responsible for digoxinlike immunoreactivity are significantly bound to serum proteins. When 4-mL aliquots of drug-free serum were supplemented with 4.0, 8.0, and 14.0 µl of aqueous extract of Chan Su, we observed digoxin-like immunoreactivity of 0.4, 0.8, and 1.6 ng/ml (0.50, 1.05, and 2.00 nmol/l), H H C 18 H 31 9 H Bufalin H Cinobufotalin H Digoxin respectively, when measured by the FPIA. However, we observed no digoxin-like immunoreactivity in the proteinfree ultrafiltrate in aliquots containing 4.0 and 8.0 µl of H H H C Figure 1 Chemical structures of bufalin, cinobufotalin, and digoxin. 176 Am J Clin Pathol 2000;114: American Society of Clinical Pathologists

4 Clinical Chemistry / RIGINAL ARTICLE Table 1 Digoxin-Like Immunoreactivity in Various Extracts of Chan Su as Measured by the FPIA, MEIA, and CLIA Mean (SD) Digoxin Equivalent Concentration, nmol/l * (n = 3) Solvent FPIA MEIA CLIA Methanol 2.40 (3) 0.97 (2) None detected Chloroform 2.27 (9) 1.16 (8) None detected Ethyl acetate 0.67 (2) 0.59 (3) None detected Acetone and water 3.99 (3) 1.23 (4) None detected Water 3.53 (1) 1.20 (3) None detected CLIA, chemiluminescent assay; FPIA, fluorescence polarization immunoassay; MEIA, microparticle enzyme immunoassay. * To convert nanomoles per liter to nanograms per milliliter, divide by Table 2 Effect of Chan Su Extract on Serum Digoxin Measurements by FPIA, MEIA, and CLIA Mean (SD) Digoxin Concentration, nmol/l* (n = 3) Specimen FPIA MEIA CLIA Drug-free serum None detected None detected None detected µl Chan Su extract per ml serum 0.82 (4) 0.53 (7) None detected µl Chan Su extract per ml serum 2.23 (5) 0.87 (6) None detected Digoxin serum pool (6) 2.70 (6) 2.50 (6) µl Chan Su extract per ml pool (7) 2.43 (2) 2.54 (9) µl Chan Su extract per ml pool (2) 2.07 (3) 2.56 (5) Digoxin serum pool (7) 1.53 (8) 1.42 (7) µl Chan Su extract per ml pool (6) 1.41 (5) 1.43 (6) µl Chan Su extract per ml pool (5) 1.34 (3) 1.51 (9) CLIA, chemiluminescent assay; FPIA, fluorescence polarization immunoassay; MEIA, microparticle enzyme immunoassay. * To convert nanomoles per liter to nanograms per milliliter, divide by Significantly greater than the digoxin concentration in the absence of Chan Su extract by independent 2-tailed t test (P <.05). Significantly less than the digoxin concentration in the absence of Chan Su extract by independent 2-tailed t test (P <.05). Chan Su extract. nly the specimen containing 14.0 µl of Chan Su extract (3.5 µl of extract per milliliter of serum) showed a digoxin-like immunoreactivity of 0.5 ng/ml (0.64 nmol/l) in the protein-free ultrafiltrate. We also studied the possibility of eliminating interference of Chan Su in serum digoxin measurements by the FPIA and MEIA by monitoring free digoxin instead of total digoxin. We added various amounts of aqueous Chan Su extract in aliquots of a serum digoxin pool and measured total and free digoxin concentrations using the MEIA. The total digoxin concentrations declined with increasing amounts of Chan Su extract. However, the free digoxin concentration did not change even in the specimen containing the highest amount of Chan Su extract Figure 2. Therefore, measuring the free digoxin concentrations can eliminate interference of Chan Su in serum digoxin measurements by the MEIA. We also added various amounts of aqueous Chan Su extract in another serum pool containing digoxin and measured total and free digoxin concentrations using the FPIA. We observed significant increases in total digoxin concentrations with increasing amounts of Chan Su extract. The free digoxin concentration did not change from the control value when a small amount of Chan Su extract was added. However, the free digoxin concentrations increased significantly from the control value when a higher volume of Chan Su extract was added to the pool Figure 3. Therefore, interference of Chan Su in serum digoxin measurements by the FPIA can be eliminated only partially by monitoring the free digoxin concentration instead of the total digoxin concentration. Discussion Although acetone and water (1:1 by volume) is a better extraction solvent for Chan Su, we used aqueous extract of Chan Su for our study because adding aqueous extract to the serum digoxin pool should not change the matrix. n the other hand, acetone is an organic solvent and may slightly change the matrix or may cause microprecipitation of serum proteins. Digoxin has a narrow therapeutic range ( ng/ml [ nmol/l]), and therapeutic monitoring of digoxin is essential to avoid the toxic effects of digoxin. Because of the widespread use of Chinese medicines by the general American Society of Clinical Pathologists Am J Clin Pathol 2000;114:

5 Dasgupta et al / INTERFERENCE F CHAN SU IN DIGXIN IMMUNASSAYS Digoxin (nmol/l) Digoxin (nmol/l) Chan Su (µl/ml) Chan Su (µl/ml) Figure 2 Total (white bars) and free (black bars) digoxin concentrations as measured by the microparticle enzyme immunoassay in serum digoxin pool supplemented with various concentrations of Chan Su extract. To convert nanomoles per liter to nanograms per milliliter, divide by Figure 3 Total (white bars) and free (black bars) digoxin concentrations as measured by the fluorescence polarization immunoassay in serum digoxin pool supplemented with various concentrations of Chan Su extract. To convert nanomoles per liter to nanograms per milliliter, divide by population, it is possible that some patients receiving digoxin also are taking Chan Su for its cardiotonic properties. The MEIA and the FPIA for digoxin are used widely in clinical laboratories in the United States for measuring serum digoxin concentrations. We observed significant positive interference of Chan Su with the serum digoxin measurement by the FPIA. Apparent serum digoxin-like immunoreactivity in the range of 0.4 to 0.9 ng/ml ( nmol/l) in healthy volunteers after ingesting Chinese medicine containing Chan Su as the major component has been reported. We mimicked this situation in our experiments and showed that Chan Su had an additive effect on serum digoxin measurements by the FPIA, causing significant falsely elevated digoxin concentrations. For example, when 6.0 µl of dilute aqueous Chan Su extract was added to 4 ml of serum, the observed digoxinlike immunoreactivity was 0.6 ng/ml (0.82 nmol/l) digoxin equivalent by the FPIA. Such a concentration is expected to occur in a patient taking Chan Su. When we added 6.0 µl of dilute aqueous Chan Su extract to 4-mL aliquots of a serum digoxin pool containing 2.0 ng/ml (2.53 nmol/l) of digoxin, the observed digoxin concentration increased to 2.5 ng/ml (3.23 nmol/l) when measured by the FPIA. n the other hand, the observed digoxin concentration significantly declined to 1.9 ng/ml (2.43 nmol/l) when measured by the MEIA (Table 2). This situation would be confusing clinically, especially if a patient were transferred from one hospital in which FPIA was used for serum digoxin measurements to another hospital in which the MEIA was used. Similarly, an outpatient could be transferred from one clinic to another clinic in which one laboratory uses an FPIA and the other uses a MEIA for determining serum digoxin concentrations. Moreover, false-positive serum digoxin concentrations as measured by the FPIA may lead the physician to reduce the digoxin dosage, resulting in suboptimal therapy. Conversely, falsely lowered serum digoxin concentrations due to the presence of Chan Su may lead the physician to increase the digoxin dosage, which may result in toxic effects. We also explored the possibility of eliminating this interference in the MEIA and FPIA for digoxin by monitoring free digoxin concentrations instead of total digoxin concentrations. Digoxin is only 25% bound to protein. In contrast, active components of Chan Su responsible for digoxin-like immunoreactivity are approximately 70% bound to serum protein. Therefore, these compounds may be absent or present in much lower concentration in the protein-free ultrafiltrate. A similar approach was used to eliminate the interference of digoxin-like immunoreactive substances in the serum digoxin assay, taking advantage of the high protein binding of the digoxin-like immunoreactive substances. 13,14 We demonstrated that monitoring the free digoxin concentration could eliminate interference of Chan Su in serum digoxin measurements by the MEIA. For example, in a digoxin pool containing 1.3 ng/ml (1.66 nmol/l) digoxin (measured by the MEIA), the concentration of mean free digoxin was 1.0 ng/ml (1.29 nmol/l, SD = 4 nmol/l). In the presence of even 5.0 µl of Chan Su extract per milliliter 178 Am J Clin Pathol 2000;114: American Society of Clinical Pathologists

6 Clinical Chemistry / RIGINAL ARTICLE of serum, the mean free digoxin concentration was 1.0 ng/ml (1.28 nmol/l, SD = 2 nmol/l). These differences were not statistically significant (Figure 2). The magnitude of interference of Chan Su with the FPIA for digoxin was higher than with the MEIA. Therefore, the interference of Chan Su in the serum digoxin measurement by the FPIA can be eliminated by monitoring free digoxin concentrations only if a low concentration of Chan Su is present. For example, in the serum digoxin pool containing 0.95 ng/ml (1.22 nmol/l) of digoxin, the mean free digoxin concentration was 0.73 ng/ml (0.94 nmol/l, SD = 3 nmol/l). When 1.0 µl/ml of dilute Chan Su extract was added, the mean total digoxin concentration was increased significantly to 1.17 ng/ml (1.50 nmol/l). However, the mean free digoxin concentration was 0.76 ng/ml (0.97 nmol/l, SD = 2 nmol/l). This value was not significantly different from the control value of 0.94 ng/ml (1.20 nmol/l). When increasing amounts of Chan Su extract were added, total and free digoxin concentrations increased significantly. When 5.0 µl/ml of Chan Su extract was added to another aliquot of the serum pool, the observed mean digoxin concentration was 2.73 ng/ml (3.50 nmol/l, SD = 4 nmol/l). The free digoxin concentration also increased significantly to 1.76 ng/ml (2.25 nmol/l, SD = 6 nmol/l) (Figure 3). The exact mechanism of suppression of the total digoxin concentration by Chan Su extract in the MEIA has not been studied. A possible mechanism is the faster dissociation rate of Chan Su antibody complex compared with the digoxin-antibody complex. We conclude that Chan Su can cause significant positive interference in serum digoxin measurement if measured by the FPIA and significant negative interference if measured by the MEIA. The CLIA, which uses monoclonal antibody, showed no interference from Chan Su. From the 1 Department of Pathology and Laboratory Medicine, University of Texas-Houston Medical School, Houston, and 2 Bayer Diagnostics, Tarrytown, NY. Address reprint requests to Dr Dasgupta: Dept of Pathology and Laboratory Medicine, University of Texas-Houston Medical School, 6431 Fannin, MSB 2.292, Houston, TX References 1. Hong Z, Chan K, Yeung HW. Simultaneous determination of bufadienolides in traditional Chinese medicine preparations, Liu-Shen-Wan by liquid chromatography. J Pharm Pharmacol. 1992;44: Chan WY, Ng TB, Yeung HW. Examination for toxicity of a Chinese drug, the toad glandular secretion product Chan SU in pregnant mice and embryos. Biol Neonate. 1995;67: Chen KK, Kovarikove A. Pharmacology and toxicology of toad venom. J Pharm Sci. 1967;56: Morishita S, Shoji M, guni Y, et al. Pharmacological actions of kyushin, a drug containing toad venom: cardiotonic and arrhythmogenic effects, and excitatory effect on respiration. Am J Chin Med. 1992;20: Pamnani MB, Chen S, Bryant HJ, et al. Effect of three sodium-potassium adenosine triphosphate inhibitors. Hypertension. 1991;18: Masuda Y, Kawazoe N, Nakajo S, et al. Bufalin induces apoptosis and influences the expression of apoptosis-related genes in human leukemia cells. Leuk Res. 1995;19: Kawazoe N, Aiuchi T, Masuda Y, et al. Induction of apoptosis by bufalin in human tumor cells is associated with a change of intracellular concentration of Na+ ions. J Biol Chem (Tokyo). 1999;126: Wang SW, Chiao YC, Tsai SC, et al. Inhibition of bufalin on pituitary and testicular functions in rats. J Pharmacol Exp Ther. 1997;283: Ko R, Greenwald M, Loscutoff S, et al. Lethal ingestion of Chinese tea containing Chan SU. West J Med. 1996;164: Fushimi R, Amino N. Digoxin concentration in blood [in Japanese, abstract in English]. Rinsho Byori. 1995;43: Panesar NS. Bufalin and unidentified substances in traditional Chinese medicine cross react in commercial digoxin assay. Clin Chem. 1992;38: Eisenberg DM, Kessler RC, Foster C, et al. Unconventional medicine in the United States: prevalence, cost, and patterns of use. N Engl J Med. 1993;328: Dasgupta A, Schammel D, Limmany A, et al. Estimating concentrations of total digoxin and digoxin-like immunoreactive substances in volume expanded patients treated with digoxin. Ther Drug Monit. 1996;18: Dasgupta A, Trejo. Suppression of total digoxin concentrations by digoxin-like immunoreactive substances in the MEIA digoxin assay: elimination of negative interference by monitoring free digoxin concentrations. Am J Clin Pathol. 1999;111: American Society of Clinical Pathologists Am J Clin Pathol 2000;114: