| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Oak Ridge Poster Sessions |
1
Dept. of Pathol. and Lab. Med., Univ. of Louisville, Louisville, KY 20292;
2
Dept. of Lab. Med. and Pathol., Univ. of Alberta, Edmonton, Canada;
a author
for correspondence: fax 502-852-1771, e-mail r0vald01{at}homer louisville.edu
We have observed significant analytical discrepancies (20% to 220%) in digitoxin (Crystodigin; DTN) immunoassay results for 11 serum samples from patients taking digitoxin. We used three methods—HPLC, immunoassay, and liquid chromatography electrospray mass spectrometry (LC/MS)—to investigate the possible sources of these discrepancies.
DTN is indicated for treatment of heart failure, atrial flutter, and supraventricular tachycardia (1). This drug is the most slowly excreted cardiac glycoside, with a half-life of 4 to 10 days (2). Administration and therapeutic monitoring of DTN are more popular in European countries than in the US. Plasma concentrations of 15–25 µg/L are considered therapeutic (3), and values >35 µg/L have been associated with toxicity in 80% of patients (4).
Biotransformation of DTN is complex. It is oxidatively cleaved to bisdigitoxoside (bis-DTN) in the liver, catalyzed by a specific cytochrome P-450 enzyme (5). Further deglycosylation to monodigitoxoside (mono-DTN) and digitoxigenin has been documented in mice, rabbits, and guinea pigs (5). DTN can be hydroxylated at position 12 to form digoxin. Glucuronidation of mono-DTN and, to a lesser extent, of digitoxigenin to form polar conjugates has also been documented (6). In isolated guinea pig atria, the positive inotropic actions of DTN and its metabolites have been ranked in order of decreasing activity as follows: mono-DTN, bis-DTN, digoxin, DTN, digitoxigenin (7).
Discrepant patients' samples had been analyzed by the following immunoassays: ACS:180® (Chiron Diagnostics), which utilizes a monoclonal antibody with chemiluminescence detection; TDx® (Abbott Labs.), which utilizes a polyclonal antibody and fluorescence polarization detection; and DPC (Diagnostic Products Corp.), which uses a polyclonal antibody with RIA technology. Patients' samples had been collected in Germany in accordance with the institutional human studies regulations of that country. All immunoassays had been performed according to their manufacturers' instructions.
Initially, we expected the deglycosylated congeners of digitoxin (i.e., bis-DTN, mono-DTN, and digitoxigenin) to be the sources of discrepancy. Therefore, we hypothesized that variations in serum DTN results by the immunoassays were the result of differential metabolism of DTN to its deglycosylated congeners in patients. Differences in generation of these congeners in patients led to variations in the apparent measured DTN.
We undertook two different approaches (HPLC-immunoassay and LC/MS) to detect and quantify DTN, its deglycosylated congeners, and any similar immunoreactive substance contributing to the differences in results. In the first approach (HPLC-immunoassay), 0.4–0.7 mL of serum was extracted on a Sep-Pak® Plus C18 column (Waters), being eluted with 500 mL/L methanol. After the eluate was evaporated, DTN and metabolites were reconstituted in the HPLC mobile phase (methanol, acetonitrile, and water), and the glycosides were separated on an HPLC system equipped with a 250 x 4.6 mm Microsorb C18 analytical column (Rainin). Fractions were collected every 30 s (or 1 min) with a Waters Millipore fraction collector, and the mobile phase was evaporated. The residue of each fraction was then reconstituted in DTN-free serum and analyzed by the ACS:180 and TDx DTN immunoassays.
The second approach involved quantification of DTN and congeners by LC/MS on the basis of each analyte's calibration curves. For the LC/MS analysis, 0.5 mL of patient's serum or a calibrator was extracted on 3-mL Supelco C18 columns with 500 mL/L methanol as the eluant. After evaporation of the solvent, the residue was reconstituted in 150 µL of 550 mL/L methanol, and 100 µL was injected onto the LC/MS. The calibrators (0, 10, 20, 40, and 80 µg/L) for DTN, bis-DTN, and mono-DTN were prepared in pooled DTN-free serum. The LC/MS system was composed of two tandem Nova-Pak® C18 columns (160 x 3.9 mm; Waters) installed in a Hewlett-Packard 1100 HPLC operating with a mobile phase of acetonitrile:methanol:water at 25:30:45 (by vol) from 0 to 20 min and changing to methanol alone by 23 min at a rate of 0.85 mL/min. The HPLC effluent was introduced through a post-column splitter (1:10 ratio) into a Hewlett-Packard 5989B mass spectrometer via a Hewlett-Packard 59987A electrospray interface operating in selected positive-ion mode. The mass to charge ratios (m/z) for ions monitored for sodium adducts of DTN, bis-DTN, mono-DTN, and digitoxigenin were 787.98, 657.83, 527.67, and 397.53, respectively. Concentrations of DTN and deglycosylated congeners were calculated from their peak areas as compared with those of the calibration curve for each respective analyte.
The HPLC-immunoassay method was able to separate DTN from its congeners as well as digoxin and its deglycosylated congeners in the same chromatographic run. The retention times (minutes) for these glycosides were as follows: DTN, 67; bis-DTN, 65; mono-DTN, 63; digitoxigenin, 60; digoxin, 57; bis-digoxin, 54; mono-digoxin, 47; and digoxigenin, 35. Analyses of HPLC fractions of a serum sample to which 60 µg/L DTN, 40 µg/L bis-DTN, 20 µg/L mono-DTN, and 60 µg/L digitoxigenin were added resulted in apparent DTN values of 52, 48, 16, and 27 µg/L by the ACS:180 DTN assay, respectively. Analyses of same fractions by the TDx DTN assay resulted in apparent DTN values of 38, 61, 32, and 107 µg/L, respectively. Lower recoveries of deglycosylated congeners of DTN by the ACS:180 can be attributed to their lower cross-reactivities in this assay (8).
Analyses of DTN and congeners by LC/MS resulted in the following elution times (minutes): DTN, 24; bis-DTN, 15; mono-DTN, 11; digitoxigenin, 8.5; and digoxigenin (internal standard), 5. Analyses of 10 and 30 µg/L DTN added to drug-free serum resulted in analytical recoveries of 65.4% ± 6.9% and 52.5% ± 5.4% by the LC/MS method.
Table 1
shows the results for 11 discrepant serum DTN samples by the
immunoassays, the HPLC-immunoassay method, and LC/MS. The extent of the
discrepancies (defined as the difference between the ACS:180 and the
TDx or DPC RIA results divided by the lowest result and multiplied by
100) ranged from 29% to 220%. Fig. 1
depicts the HPLC-immunoassay chromatogram for specimen 7 in
Table 1
. Immunoreactive substances in the early region of this
chromatogram (3–5 min) revealed a peak with apparent DTN
immunoreactivity of 36 µg/L by the ACS:180 and 145 µg/L by the TDx
digitoxin immunoassay (Fig. 1
). The same substances (possibly "polar
metabolites" of DTN) were responsible for discordances in results for
samples 8, 10, and 11. Notice that the LC/MS results for DTN
concentration in these samples agreed with the immunoreactivity
measured for the peak corresponding to DTN (Table 1
) except for sample
number 8. Limitations in sample volume kept us from further
investigating the cause of disagreement in DTN results for sample 8 by
LC/MS and HPLC-immunoassay.
|
|
As the investigation of discrepant samples in this study shows, deglycosylated metabolites of DTN were not the source of discord. However, our data suggest that suspected polar substances may have produced substantial discrepancy in results of at least four of the samples. HPLC chromatographic mobility and immunoreactivity of these substances suggest they may be the polar metabolites of DTN. Castle and Zavecs (9) have reported that purified glucuronide conjugate of mono-DTN from rat liver microsomes had an IC50 of 5.50 µmol/L for inhibition of canine kidney Na+,K+-ATPase catalytic activity. In the same study, IC50 values of 0.68 and 0.08 µmol/L were reported for DTN and digitoxigenin, respectively. Their data suggested that the glucuronide conjugate of mono-DTN was less biologically active than DTN and digitoxigenin. In a study of 29 different human liver microsomal preparations, the rate of glucuronidation of mono-DTN ranged from 18 to 87 pmol/min per milligram of protein (10).
We conclude that convergence of immunoassay, HPLC, and mass spectrometry was successful in resolution of discrepancy in DTN results. The deglycosylated congeners of DTN (mono-DTN, bis-DTN, and digitoxigenin) were not detected in serum of patients on DTN therapy and did not contribute to discrepancy in DTN results for samples investigated. Our preliminary data also suggest that suspected polar metabolites of DTN (possibly the conjugates of mono-DTN and others) were the sources of significant discrepancy in immunoassay results for the patients' samples tested. The ACS:180 assay cross-reacted less than the TDx to the polar immunoreactive substances.
We recommend that DTN immunoassays should be evaluated for cross-reactivity of possible polar metabolites of DTN. Structural characterization, clinical significance, and pharmacological activities of these substances should be further investigated.
Acknowledgments
Diagnostic Reference Laboratory is supported by a consortium of in vitro diagnostics manufacturers, including Chiron Diagnostics, Roche Diagnostics, Dade International, Bayer (Diagnostics Division), and Boehringer Mannheim.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  B. Al, P. Yarbil, M. Dogan, S. Kabul, and C. Yildirim A case of non-fatal oleander poisoning BMJ Case Reports, February 11, 2010; 2010(feb11_1): bcr0220091573 - bcr0220091573. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) digitoxin assays. Only
fractions eluting in the first 15 min are shown. Note the polar peak
eluting at 3–5 min, which measured 36 µg/L (ACS:180) and 146 µg/L
(FPIA) for apparent DTN. DTN in the discrepant sample eluted at 68–70
min (not shown) and measured 15 µg/L by the ACS:180, 7 µg/L by the
TDx DTN immunoassays.