XENOBIOTICA,1990, VOL. 20, NO. 6, 635-643
Comparison of digoxin analysis by highperformance liquid chromatography/post-column derivatization and fluorescence-polarization immunoassay L. EMBREE and K. M. McERLANEt
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia V6T lWS,Canada
Received 12 June 1989, revised 5 December 1989; accepted 12 Jmnutry 1990
1. This study compared the analysis of digoxin using a high-performance liquid chromatographic post-column derivatization (HPLC-PC) assay and the TDx fluorescence polarization immunoassay (FPIA). 2. Serum obtained from 15 digitalizedpatients showed higher mean digoxin levels with the FPIA method as compared to the HPLC-PC procedure such that the mean HPLCPC/FPIA ratio was 0.91 +@14 (meanf SD). Demonstrated cross-reactivity of digoxin metabolites with the FPIA is probably responsible for this observation. 3. Cross-reactivity of the immunoassay towards endogenous material present in serum samples from certain patient groups was an even greater problem, with apparent ‘digoxin’ serum concentrations in untreated hepatic failure patients being within the therapeutic range for digoxin. 4. The HPLC-PC method did not suffer from such interference and would therefore provide more accurate values for patients where high levels of interference could contribute to false digoxin levels.
Introduction Digitalis glycosides form one of the most beneficial groups of drugs available to aid the failing heart. Digoxin, the most commonly used digitalisglycoside, is still an important drug for the treatment of congestive heart failure and certain disturbances in cardiac rhythm despite the advent of new therapies (Doherty and Kane 1975). Several immunoassay procedures and a few chromatographicmethods have been used for digoxin analysis. Clinical laboratories use digoxin immunoassays due to their speed, sensitivity and cost. These methods, however, lack specificity for digoxin in the presence of its metabolites (Gault et al. 1982, Valdes et al. 1984) and some co-administered drugs (DiPiro et al. 1980, Silber et al. 1979). High levels of digoxin-like immunoreactive substance(s) (DLIS), which also cross-react with the antisera used in immunoassay methods, have been reported in serum from hypertensive patients (Cloix et al. 1987, Wilkins 1985), renal failure patients (Kramer et al. 1985, Yatscoff et al. 1984), hepatic failure patients (Greenway and Nanji 1985, Nanji and Greenway 1985, 1986), umbilical cord blood and placental homogenates (Diamandis et al. 1985, Koren et al. 1988, Scherrmann et al. 1986a, 1986 b) as well as normal adult subjects who never received digoxin (Diamandiset al. 1985, Balzan et al. 1984, Valdes and Siegfried 1983).
T o whom correspondence should be addressed. 0049-8254/90 5300 0 1990 Taylor St Francis Ltd.
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
636
L. Embree and K . M . McErlane
The presence of DLIS in patient samples, particularly those with potentially high DLIS levels, seriously compromises the accuracy and interpretation of the results from immunoassay methods. Fluorescence polarization immunoassays (FPIA) for digoxin have shown less interference from DLIS than other immunoassay methods (Yatscoff et al. 1984) but do not completely eliminate the interference (Bianchi 1986, Soldin et al. 1986). It has been suggested that the FPIA may therefore be better than the radioimmunoassay (RIA) when DLIS is present (Yatscoff et al. 1984). This paper reports the comparison of a previously described high-performance liquid chromatographic post-column (HPLC-PC) fluorogenic digoxin assay (Embree and McErlane 1989) with the FPIA for interference from endogenous compounds present in serum from patient groups where high levels of DLIS have been found.
Materials and methods Materials and reagents Digoxin and its metabolites: digoxigenin bisdigitoxoside, digoxigenin monodigitoxoside, digoxigenin, dihydrodigoxin and dihydrodigoxigenin, as well as the internal standard digitoxigenin,were obtained from Boehringer (Mannheim, GFR). L-Ascorbic acid, hydrochloric acid (HCI) and hydrogen peroxide (30%) were purchased from BDH Chemicals (Toronto, Ontario, Canada). Reagent grade absolute ethanol was purchased from Commercial Alcohols Ltd. (Toronto, Ontario, Canada). Acetone, methanol and n-propanol were reagent grade and obtained from BDH Chemicals (Toronto, Ontario, Canada). Isooctane (2,2,4-trimethylpentanentane)was glass distilled quality and purchased from BDH Chemicals (Toronto, Ontario, Canada). HPLC grade water was produced using the Milli-Q Water System (Millipore Corp., Milford, MA, USA). The remaining HPLC grade solvents were Omnisolv grade from BDH Chemicals (Toronto, Ontario, Canada). HPLC-PC assay of digoxin The HPLC-PC assay employed has been previously described (Embree and McErlane 1989). It consisted of an HPLC system for separation of digoxin from its metabolites and interfering compounds, a post-column fluorogenic reactor and a fluorometer equipped with a modified quartz flow cell. Excitation (360nm) and emission filters (425 nm) were used with the fluorometer for detection of digoxin. Digoxin standards and digitoxigenin, the internal standard used, were prepared in ethanol for addition to serum samples as described previously (Embree and McErlane 1989)and a solvent-solvent extraction procedure was used for preparation of serum samples prior to HPLC-PC analysis. Preparation of &orsin metabolites in serum for FPIA analysis Stock solutions were prepared by weighing the individual metabolites of digoxin (1 mg), dissolving them in ethanol (100ml). Dilutions of this stock solution were prepared so that the final concentration was 1ng metabolite/lOpl. Each of the five metabolites (10~1)was added separately to serum (1 ml) and these were assayed by the TDx FPIA (Abbott Laboratories Diagnostics Division, North Chicago, IL, USA). Comparison of the HPLC-PC and FPIA assays of serum sampks spiked m-tk digoxin T o seven serum samples (4ml) the digoxin standard solutions in ethanol were added to provide serum samples containing from 0-3.4 ng digoxinlml. After mixing, aliquots of 3 ml were removed for analysis using the HPLC-PC fluorogenic assay previously described (Embree and McErlane 1989). An aliquot of the remaining sample was then analysed for digoxin using the TDx FPIA following the directions supplied with the commercial kit. Comparative assay of serum sampks from digitahed and UndigitaIlzed patiCnts Serum samples from various different patient groups were obtained from local hospitals. Undigitalized hypertensive patients with blood pressure values greater than 150/100mmHg were included in this study. Undigitalized renal failure patients who required dialysis and alcoholic cirrhosis patients were also evaluated for ‘digoxin’levels. One digitalized hepatic failure patient was also included in the study. Mixed arterial and venous unbilical cord blood samplesobtained from undigitalized mothers were likewise assayed by both the HPLC-PC and FPIA methods. Regrettably for reasons of confidentiality, patient demographics and details of any concomitant medication were unavailable.
Digoxin HPLC-PC and FPIA: Analysis of patient serum
637
Results HPLC-PC assay of digoxin
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
For the HPLC-PC assay used in this study, the response to digoxin was linear from 05-3.3 ng/ml extracted from 3 ml of serum with a mean coefficient of variation of 5.6% and a minimum quantifiable limit of 0 5 nglinjection (it was however possible to detect but perhaps not quantify digoxin to 025ng/ml). The chromatographic system provided complete separation of digoxin from digoxigenin, dihydrodigoxigenin, digoxigeninmonodigitoxosideand digoxigeninbisdigitoxoside (Embree and McErlane 1989). The resolution between digoxin and dihydrodigoxin (I = 0899) was sufficient for peak height quantitation of digoxin.
FPIA of digoxin metabolites in serum Serum samples containing from 1*17-1-70nmol/l of each metabolite were assayed by the FPIA method. The results from the FPIA method shown in table 1 indicated apparent ‘digoxin’concentrations in the range 0 7 4 8nmol/l, although no digoxin was present in these samples and the blank serum used in their preparation did not give false-positive digoxin levels with this immunoassay. Comparison of HPLC-PC and FPIA methods with serum spiked with digoxin Seven blank serum samples spiked with 0-3.4 ng digoxinfmlwere prepared and assayed once by both methods. The observed relationship between the two methods is presented in figure 1 and shows that under these conditions the methods are comparable. Comparison of HPLC-PC and FPIA methods with serum from digitalized patients Serum samples from 15 patients who had received digoxin therapeutically were assayed by both HPLC-PC and FPIA methods and the results are shown in table 2. The mean concentration obtained using the HPLC-PC assay was 099 f 0 5 6 and that found with the FPIA method was 1.13 f 0.73.
Table 1. Digoxin analysis of serum, spiked only with metabolites, using fluorescence polarization immunoassay.
Metabolite
Blank Serum Digoxigenin Digoxigenin monodigitoxoside Digoxigenin bisdigitoxoside Dihydrodigoxin Dihydrodigoxigenin
FPIA Known nanomolar assay results concentration (nmol/l)
Cross-reactivity reported by FPIA manufacturer (%)
none 1.41 (0.55)*
none 14
> 100
1.48 (0.77)
48
> 100
1-17 (0.76) 1.27 (0.99) 1.70 (0.67)
36 0.8 0.7
> 100 91
Figures in parentheses represent concentration expressed as ng/ml.
L. Embree and K . M . McErlane
638
0 0
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
4-
1.54
Slops = 0.821
0.5
y-Intercept r
04 0
, 0.5
= 0.140 t = 1.26
= 0.9897 1
1
1.5
1
2
2.5
3
3.5
4
4.5
FPIA METHOD ng/rnL
Figure 1 . Correlation between HPLC-PC and TDx FPIA methods for the analysisof digoxin in human serum samples containing known quantities of digoxin. Table 2. Comparison of the HPLC-PC and FPIA methods for digoxin analysis in digitalized patients. Digoxin concentration (nglml) HPLC-PC/FPIA Patient
HPLC-PC
FPIA
ratio
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1.1 1 1.41 0.69 0.98 1.01 089 0.94 1.07 093 280 053 081 0.62 0.69
1.2 1.6 0.8 1 .o 1 .o 0.9 1.1 0.9 06 1.1 3.6 09 0.9 0.6 07
092 088 086 098 1.01 099 086 1.19 0.78 085 0.78 0.59 0-90 1.03 099
099 +056 (n=15)
1.13 +073 (n=15)
091 k014
mean & S D
0-47
(n=15)
Assay of serum samples from various patient groups Hypertensive patients. Serum samples from five undigitalized hypertensive patients were obtained and were evaluated by both the HPLC-PC and TDx FPIA methods. None of the samples were within the detection range of either method.
Digoxin HPLC-PC and FPIA: Analysis of patient serum
639
Renal failure patients. Serum samples were obtained from 20 undigitalized renal failure patients who were on haemodia1ysis.-The renal failure patients’ serum samples were evaluated by both the HPLC-PC and TDx FPIA. ‘Digoxin’ was not detected in any of the serum samples by either method.
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
Hepatic failure. Ten serum samples from nine hepatic failure patients with diagnosed alcoholic cirrhosis were evaluated by both HPLC-PC and FPIA methods. The results are given in table 3 and apparently show evidence of ‘digoxin’ in five of the eight undigitalized patients using the FPIA method. Only the digitalized patient (0.125mg digoxin daily) gave positive digoxin levels by the HPLC-PC assay. For this patient ( # 9) the HPLC-PC assay indicated lower levels of digoxin than the corresponding FPIA results. Umbilical cord blood samples. A total of 17mixed arterial and venous umbilical cord blood samples from 11 patients were obtained and assayed by both HPLC-PC and FPIA methods. The results are shown in table 4. For those samples assayed by HPLC-PC, none indicated that ‘digoxin’was present at levels above the sensitivity of this assay (0.5 nglml). The FPIA method indicated that ‘digoxin’ was apparently present in all but two samples investigated.
Discussion Interference from digoxin metabolites is presently a significant problem with the TDx FPIA method. The cross-reactivity values reported with the FPIA method for digoxigenin, digoxigenin monodigitoxoside and digoxigenin bisdigitoxoside were as great or greater than those observed previously with the RIA method (Valdes et al. 1984). For dihydrodigoxigenin, the reported cross-reactivity for the FPIA method (91%) was considerably greater than that reported for RIA methods (Gault et al. 1982). The FPIA method requires serum proteins to be precipitated with 5-sulphosalicylicacid (Skogen et al. 1987)which has been suggested to result in the transformation of digoxin and its metabolites to digoxigenin and/or dihydrodigoxigenin (Sonobe et al. 1980). It was therefore not surprising that the metabolites were not differentiated from the drug itself using this procedure. Table 3. Digoxin assay of serum samples from hepatic failure patients by HPLC-PC and TDx FPIA methods. Apparent digoxin concentration (nglml) Patient
HPLC-PC
FPIA
1
ND ND ND ND ND ND ND ND 062 1-06
06 02 (36
2
3 4 5 6
7 8 9*
0.5
NF NF 02 NF 1.2 1.3
Digitalized. ND =none detected (
Comparison of digoxin analysis by highperformance liquid chromatography/post-column derivatization and fluorescence-polarization immunoassay L. EMBREE and K. M. McERLANEt
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia V6T lWS,Canada
Received 12 June 1989, revised 5 December 1989; accepted 12 Jmnutry 1990
1. This study compared the analysis of digoxin using a high-performance liquid chromatographic post-column derivatization (HPLC-PC) assay and the TDx fluorescence polarization immunoassay (FPIA). 2. Serum obtained from 15 digitalizedpatients showed higher mean digoxin levels with the FPIA method as compared to the HPLC-PC procedure such that the mean HPLCPC/FPIA ratio was 0.91 +@14 (meanf SD). Demonstrated cross-reactivity of digoxin metabolites with the FPIA is probably responsible for this observation. 3. Cross-reactivity of the immunoassay towards endogenous material present in serum samples from certain patient groups was an even greater problem, with apparent ‘digoxin’ serum concentrations in untreated hepatic failure patients being within the therapeutic range for digoxin. 4. The HPLC-PC method did not suffer from such interference and would therefore provide more accurate values for patients where high levels of interference could contribute to false digoxin levels.
Introduction Digitalis glycosides form one of the most beneficial groups of drugs available to aid the failing heart. Digoxin, the most commonly used digitalisglycoside, is still an important drug for the treatment of congestive heart failure and certain disturbances in cardiac rhythm despite the advent of new therapies (Doherty and Kane 1975). Several immunoassay procedures and a few chromatographicmethods have been used for digoxin analysis. Clinical laboratories use digoxin immunoassays due to their speed, sensitivity and cost. These methods, however, lack specificity for digoxin in the presence of its metabolites (Gault et al. 1982, Valdes et al. 1984) and some co-administered drugs (DiPiro et al. 1980, Silber et al. 1979). High levels of digoxin-like immunoreactive substance(s) (DLIS), which also cross-react with the antisera used in immunoassay methods, have been reported in serum from hypertensive patients (Cloix et al. 1987, Wilkins 1985), renal failure patients (Kramer et al. 1985, Yatscoff et al. 1984), hepatic failure patients (Greenway and Nanji 1985, Nanji and Greenway 1985, 1986), umbilical cord blood and placental homogenates (Diamandis et al. 1985, Koren et al. 1988, Scherrmann et al. 1986a, 1986 b) as well as normal adult subjects who never received digoxin (Diamandiset al. 1985, Balzan et al. 1984, Valdes and Siegfried 1983).
T o whom correspondence should be addressed. 0049-8254/90 5300 0 1990 Taylor St Francis Ltd.
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
636
L. Embree and K . M . McErlane
The presence of DLIS in patient samples, particularly those with potentially high DLIS levels, seriously compromises the accuracy and interpretation of the results from immunoassay methods. Fluorescence polarization immunoassays (FPIA) for digoxin have shown less interference from DLIS than other immunoassay methods (Yatscoff et al. 1984) but do not completely eliminate the interference (Bianchi 1986, Soldin et al. 1986). It has been suggested that the FPIA may therefore be better than the radioimmunoassay (RIA) when DLIS is present (Yatscoff et al. 1984). This paper reports the comparison of a previously described high-performance liquid chromatographic post-column (HPLC-PC) fluorogenic digoxin assay (Embree and McErlane 1989) with the FPIA for interference from endogenous compounds present in serum from patient groups where high levels of DLIS have been found.
Materials and methods Materials and reagents Digoxin and its metabolites: digoxigenin bisdigitoxoside, digoxigenin monodigitoxoside, digoxigenin, dihydrodigoxin and dihydrodigoxigenin, as well as the internal standard digitoxigenin,were obtained from Boehringer (Mannheim, GFR). L-Ascorbic acid, hydrochloric acid (HCI) and hydrogen peroxide (30%) were purchased from BDH Chemicals (Toronto, Ontario, Canada). Reagent grade absolute ethanol was purchased from Commercial Alcohols Ltd. (Toronto, Ontario, Canada). Acetone, methanol and n-propanol were reagent grade and obtained from BDH Chemicals (Toronto, Ontario, Canada). Isooctane (2,2,4-trimethylpentanentane)was glass distilled quality and purchased from BDH Chemicals (Toronto, Ontario, Canada). HPLC grade water was produced using the Milli-Q Water System (Millipore Corp., Milford, MA, USA). The remaining HPLC grade solvents were Omnisolv grade from BDH Chemicals (Toronto, Ontario, Canada). HPLC-PC assay of digoxin The HPLC-PC assay employed has been previously described (Embree and McErlane 1989). It consisted of an HPLC system for separation of digoxin from its metabolites and interfering compounds, a post-column fluorogenic reactor and a fluorometer equipped with a modified quartz flow cell. Excitation (360nm) and emission filters (425 nm) were used with the fluorometer for detection of digoxin. Digoxin standards and digitoxigenin, the internal standard used, were prepared in ethanol for addition to serum samples as described previously (Embree and McErlane 1989)and a solvent-solvent extraction procedure was used for preparation of serum samples prior to HPLC-PC analysis. Preparation of &orsin metabolites in serum for FPIA analysis Stock solutions were prepared by weighing the individual metabolites of digoxin (1 mg), dissolving them in ethanol (100ml). Dilutions of this stock solution were prepared so that the final concentration was 1ng metabolite/lOpl. Each of the five metabolites (10~1)was added separately to serum (1 ml) and these were assayed by the TDx FPIA (Abbott Laboratories Diagnostics Division, North Chicago, IL, USA). Comparison of the HPLC-PC and FPIA assays of serum sampks spiked m-tk digoxin T o seven serum samples (4ml) the digoxin standard solutions in ethanol were added to provide serum samples containing from 0-3.4 ng digoxinlml. After mixing, aliquots of 3 ml were removed for analysis using the HPLC-PC fluorogenic assay previously described (Embree and McErlane 1989). An aliquot of the remaining sample was then analysed for digoxin using the TDx FPIA following the directions supplied with the commercial kit. Comparative assay of serum sampks from digitahed and UndigitaIlzed patiCnts Serum samples from various different patient groups were obtained from local hospitals. Undigitalized hypertensive patients with blood pressure values greater than 150/100mmHg were included in this study. Undigitalized renal failure patients who required dialysis and alcoholic cirrhosis patients were also evaluated for ‘digoxin’levels. One digitalized hepatic failure patient was also included in the study. Mixed arterial and venous unbilical cord blood samplesobtained from undigitalized mothers were likewise assayed by both the HPLC-PC and FPIA methods. Regrettably for reasons of confidentiality, patient demographics and details of any concomitant medication were unavailable.
Digoxin HPLC-PC and FPIA: Analysis of patient serum
637
Results HPLC-PC assay of digoxin
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
For the HPLC-PC assay used in this study, the response to digoxin was linear from 05-3.3 ng/ml extracted from 3 ml of serum with a mean coefficient of variation of 5.6% and a minimum quantifiable limit of 0 5 nglinjection (it was however possible to detect but perhaps not quantify digoxin to 025ng/ml). The chromatographic system provided complete separation of digoxin from digoxigenin, dihydrodigoxigenin, digoxigeninmonodigitoxosideand digoxigeninbisdigitoxoside (Embree and McErlane 1989). The resolution between digoxin and dihydrodigoxin (I = 0899) was sufficient for peak height quantitation of digoxin.
FPIA of digoxin metabolites in serum Serum samples containing from 1*17-1-70nmol/l of each metabolite were assayed by the FPIA method. The results from the FPIA method shown in table 1 indicated apparent ‘digoxin’concentrations in the range 0 7 4 8nmol/l, although no digoxin was present in these samples and the blank serum used in their preparation did not give false-positive digoxin levels with this immunoassay. Comparison of HPLC-PC and FPIA methods with serum spiked with digoxin Seven blank serum samples spiked with 0-3.4 ng digoxinfmlwere prepared and assayed once by both methods. The observed relationship between the two methods is presented in figure 1 and shows that under these conditions the methods are comparable. Comparison of HPLC-PC and FPIA methods with serum from digitalized patients Serum samples from 15 patients who had received digoxin therapeutically were assayed by both HPLC-PC and FPIA methods and the results are shown in table 2. The mean concentration obtained using the HPLC-PC assay was 099 f 0 5 6 and that found with the FPIA method was 1.13 f 0.73.
Table 1. Digoxin analysis of serum, spiked only with metabolites, using fluorescence polarization immunoassay.
Metabolite
Blank Serum Digoxigenin Digoxigenin monodigitoxoside Digoxigenin bisdigitoxoside Dihydrodigoxin Dihydrodigoxigenin
FPIA Known nanomolar assay results concentration (nmol/l)
Cross-reactivity reported by FPIA manufacturer (%)
none 1.41 (0.55)*
none 14
> 100
1.48 (0.77)
48
> 100
1-17 (0.76) 1.27 (0.99) 1.70 (0.67)
36 0.8 0.7
> 100 91
Figures in parentheses represent concentration expressed as ng/ml.
L. Embree and K . M . McErlane
638
0 0
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
4-
1.54
Slops = 0.821
0.5
y-Intercept r
04 0
, 0.5
= 0.140 t = 1.26
= 0.9897 1
1
1.5
1
2
2.5
3
3.5
4
4.5
FPIA METHOD ng/rnL
Figure 1 . Correlation between HPLC-PC and TDx FPIA methods for the analysisof digoxin in human serum samples containing known quantities of digoxin. Table 2. Comparison of the HPLC-PC and FPIA methods for digoxin analysis in digitalized patients. Digoxin concentration (nglml) HPLC-PC/FPIA Patient
HPLC-PC
FPIA
ratio
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1.1 1 1.41 0.69 0.98 1.01 089 0.94 1.07 093 280 053 081 0.62 0.69
1.2 1.6 0.8 1 .o 1 .o 0.9 1.1 0.9 06 1.1 3.6 09 0.9 0.6 07
092 088 086 098 1.01 099 086 1.19 0.78 085 0.78 0.59 0-90 1.03 099
099 +056 (n=15)
1.13 +073 (n=15)
091 k014
mean & S D
0-47
(n=15)
Assay of serum samples from various patient groups Hypertensive patients. Serum samples from five undigitalized hypertensive patients were obtained and were evaluated by both the HPLC-PC and TDx FPIA methods. None of the samples were within the detection range of either method.
Digoxin HPLC-PC and FPIA: Analysis of patient serum
639
Renal failure patients. Serum samples were obtained from 20 undigitalized renal failure patients who were on haemodia1ysis.-The renal failure patients’ serum samples were evaluated by both the HPLC-PC and TDx FPIA. ‘Digoxin’ was not detected in any of the serum samples by either method.
Xenobiotica Downloaded from informahealthcare.com by University of Calgary on 02/02/15 For personal use only.
Hepatic failure. Ten serum samples from nine hepatic failure patients with diagnosed alcoholic cirrhosis were evaluated by both HPLC-PC and FPIA methods. The results are given in table 3 and apparently show evidence of ‘digoxin’ in five of the eight undigitalized patients using the FPIA method. Only the digitalized patient (0.125mg digoxin daily) gave positive digoxin levels by the HPLC-PC assay. For this patient ( # 9) the HPLC-PC assay indicated lower levels of digoxin than the corresponding FPIA results. Umbilical cord blood samples. A total of 17mixed arterial and venous umbilical cord blood samples from 11 patients were obtained and assayed by both HPLC-PC and FPIA methods. The results are shown in table 4. For those samples assayed by HPLC-PC, none indicated that ‘digoxin’was present at levels above the sensitivity of this assay (0.5 nglml). The FPIA method indicated that ‘digoxin’ was apparently present in all but two samples investigated.
Discussion Interference from digoxin metabolites is presently a significant problem with the TDx FPIA method. The cross-reactivity values reported with the FPIA method for digoxigenin, digoxigenin monodigitoxoside and digoxigenin bisdigitoxoside were as great or greater than those observed previously with the RIA method (Valdes et al. 1984). For dihydrodigoxigenin, the reported cross-reactivity for the FPIA method (91%) was considerably greater than that reported for RIA methods (Gault et al. 1982). The FPIA method requires serum proteins to be precipitated with 5-sulphosalicylicacid (Skogen et al. 1987)which has been suggested to result in the transformation of digoxin and its metabolites to digoxigenin and/or dihydrodigoxigenin (Sonobe et al. 1980). It was therefore not surprising that the metabolites were not differentiated from the drug itself using this procedure. Table 3. Digoxin assay of serum samples from hepatic failure patients by HPLC-PC and TDx FPIA methods. Apparent digoxin concentration (nglml) Patient
HPLC-PC
FPIA
1
ND ND ND ND ND ND ND ND 062 1-06
06 02 (36
2
3 4 5 6
7 8 9*
0.5
NF NF 02 NF 1.2 1.3
Digitalized. ND =none detected (
