A new high-performance liquid chromatographic method for determination of warfarin enantiomers

Warfarin is the most common agent used for control and prevention of venous as well as arterial thromboembolism. Although warfarin is administered as a racemic mixture of two stereoisomers (S and R), the S-form is mainly responsible for the anticoagulant effect. The anticoagulant effect of the drug is monitored by analysis of prothrombin complex (International Normalised Ratio,INR). In some cases, however, the measurements of plasma warfarin concentration are needed. Here, we present a new, rapid, sensitive and cost-effective HPLC-method for the determination of warfarin enantiomers in plasma. The chromatographic system consisted of Waters 616 gradient pump, Waters 996 photo diode array detector, Gilson 230 autoinjector and Pirkle (R,R) Whelk-O1 column (25 cmx4.6 mm I.D., 5 microm). An isocratic mobile phase of methanol/acetonitrile/water (50/10/40, v/v) with 0.1% glacial acetic acid was used. The follow rate was 1 mL/min. Data analysis was carried out with Waters Millennium32. The absorbance at 305 nm was measured with a total run-time of 15 min. Method linearity was studied by establishing regression data containing eight points over the range 0.08-10 microg/mL. In this range, warfarin showed to be linear (r2=0.9997 for S-warfarin and r2=0.9998 for R-warfarin). The limit of detection in plasma was 16 ng/mL for S-warfarin and 18 ng/mL for R-warfarin. Limit of quatitation was defined as 10xLOD. The extraction recovery was approximately 80%. Also the relation between INR and warfarin concentration was investigated. As expected, there was a low correlation between these two variables (r=0.23, y=0.3044x+0.9712). This method offers a rapid and cost-effective determination of warfarin enantiomers in human plasma.     

The pharmocogenomics of warfarin: closing in on personalized medicine

Warfarin, a coumarin anticoagulant, is used worldwide for the treatment and prevention of thromboembolic disease. Warfarin therapy, however, can be difficult to manage because of the drug's narrow therapeutic index and the wide interindividual variability in patient response. It is now clear that genetic polymorphisms in genes influencing metabolism (CYP2C9) and pharmacodynamic response (VKORC1) are strongly associated with warfarin responsiveness. Optimal warfarin dosing in turn drives other positive anticoagulation-related outcomes. Therefore, a strong basic science argument is emerging for prospective genotyping of warfarin patients. Effective clinical translation would establish warfarin pharmacogenomics as a heuristic model for personalized medicine.     

A comparative proteomic analysis of HepG2 cells incubated by S(−) and R(+) enantiomers of anti‐coagulating drug warfarin

Warfarin is a commonly prescribed oral anti‐coagulant with narrow therapeutic index. It interferes with vitamin K cycle to achieve anti‐coagulating effects. Warfarin has two enantiomers, S(?) and R(+) and undergoes stereoselective metabolism, with the S(?) enantiomer being more effective. We reported that the intracellular protein profile in HepG2 cells incubated with S(?) and R(+) warfarin, using iTRAQ‐coupled 2‐D LC‐MS/MS. In samples incubated with S(?) and R(+) warfarin alone, the multi‐task protein Protein SET showed significant elevation in cells incubated with S(?) warfarin but not in those incubated with R(+) warfarin. In cells incubated with individual enantiomers of warfarin in the presence of vitamin K, protein disulfide isomerase A3 which is known as a glucose‐regulated protein, in cells incubated with S(?) warfarin was found to be down‐regulated compared to those incubated with R(+) warfarin. In addition, Protein DJ‐1 and 14‐3‐3 Proteinσ were down‐regulated in cells incubated with either S(?) or R(+) warfarin regardless of the presence of vitamin K. Our results indicated that Protein DJ‐1 may act as an enzyme for expression of essential enzymes in vitamin K cycle. Taken together, our findings provided molecular evidence on a comprehensive protein profile on warfarin–cell interaction, which may shed new lights on future improvement of warfarin therapy.     

Metabolic profiling of HepG2 cells incubated with S(?) and R(+) enantiomers of anti-coagulating drug warfarin

Warfarin is a commonly prescribed oral anticoagulant with narrow therapeutic index. It achieves anti-coagulating effects by interfering with the vitamin K cycle. Warfarin has two enantiomers, S(−) and R(+) and undergoes stereoselective metabolism, with the S(−) enantiomer being more effective. We reported the intracellular metabolic profile in HepG2 cells incubated with S(−) and R(+) warfarin by GCMS. Chemometric method PCA was applied to analyze the individual samples. A total of 80 metabolites which belong to different categories were identified. Two batches of experiments (with and without the presence of vitamin K) were designed. In samples incubated with S(−) and R(+) warfarin, glucuronic acid showed significantly decreased in cells incubated with R(+) warfarin but not in those incubated with S(−) warfarin. It may partially explain the lower bio-activity of R(+) warfarin. And arachidonic acid showed increased in cells incubated with S(−) warfarin but not in those incubated with R(+) warfarin. In addition, a number of small molecules involved in γ-glutamyl cycle displayed ratio variations. Intracellular glutathione detection further validated the results. Taken together, our findings provided molecular evidence on a comprehensive metabolic profile on warfarin-cell interaction which may shed new lights on future improvement of warfarin therapy.     

Fluorometric and liquid chromatographic study of the binding of two coumarins to β-cyclodextrin

The fluorescences of warfarin and phenprocoumon are enhanced following complexation with β-cyclodextrin; (+)-(R)- and (?)-(S)-phenprocoumons have different affinities for this cyclodextrin, whereas the corresponding enantiomers of warfarin have similar binding constants. Apparently, hemiketal formation in the case of warfarin minimizes chiral discrimination. This is confirmed using a β-cyclodextrin bonded chromatography column on which the phenprocoumon enantiomers are separated, whereas those of warfarin are not.     

Determination of warfarin enantiomers and hydroxylated metabolites in human blood plasma by liquid chromatography with achiral and chiral separation

An assay comprising two simple, selective and isocratic HPLC methods with UV detection was developed and validated for measuring warfarin enantiomers and all five warfarin monohydroxylated metabolites in patient blood plasma. Following liquid/liquid extraction from 1 ml of blood plasma a baseline separation of analytes was achieved on chiral (alpha(1) acid glycoprotein - AGP) and achiral (C(18)) column. Both methods were consistent (R.S.D.<6.9% for warfarin enantiomers and<8.9% for monohydroxylated metabolites) and linear (r>0.998). The limits of detection were 25 ng/ml for warfarin enantiomers, 25 ng/ml for 4'-, 10-, 6- and 7-hydroxywarfarin, 35 ng/ml for 8-hydroxywarfarin and 50 ng/ml for racemic warfarin. In a clinical study in 204 patients, it was confirmed that the assay is appropriate for evaluation of influences of genetic polymorphisms, demographic factors and concomitant drug treatment on warfarin metabolism.     

Site I on human albumin: differences in the binding of (R)- and (S)-warfarin.

The binding of drugs known to interact with area I on human serum albumin (HSA) was investigated using a chiral stationary phase obtained by anchoring HSA to a silica matrix. In particular, this high-pressure affinity chromatography selector was employed to study the binding properties of the individual enantiomers of warfarin. The pH and composition of the mobile phase modulate the enantioselective binding of warfarin. Displacement chromatography experiments evidenced significant differences in the binding of the warfarin enantiomers to site I. The (S)-enantiomer was shown to be a direct competitor for (R)-warfarin, while (R)-warfarin was an indirect competitor for the (S)-enantiomer. Salicylate directly competed with (R)-warfarin and indirectly with (S)-warfarin. This behavior was confirmed by difference CD experiments, carried out with the same [HSA]/[drug] system in solution.     

Analysis of Oxcarbazepine and the 10‐Hydroxycarbazepine Enantiomers in Plasma by LC‐MS/MS: Application in a Pharmacokinetic Study

Oxcarbazepine is a second‐generation antiepileptic drug indicated as monotherapy or adjunctive therapy in the treatment of partial seizures or generalized tonic–clonic seizures in adults and children. It undergoes rapid presystemic reduction with formation of the active metabolite 10‐hydroxycarbazepine (MHD), which has a chiral center at position 10, with the enantiomers (S)‐(+)‐ and R‐(?)‐MHD showing similar antiepileptic effects. This study presents the development and validation of a method of sequential analysis of oxcarbazepine and MHD enantiomers in plasma using liquid chromatography with tandem mass spectrometry (LC‐MS/MS). Aliquots of 100 μL of plasma were extracted with a mixture of methyl tert‐butyl ether: dichloromethane (2:1). The separation of oxcarbazepine and the MHD enantiomers was obtained on a chiral phase Chiralcel OD‐H column, using a mixture of hexane:ethanol:isopropanol (80:15:5, v/v/v) as mobile phase at a flow rate of 1.3 mL/min with a split ratio of 1:5, and quantification was performed by LC‐MS/MS. The limit of quantification was 12.5 ng oxcarbazepine and 31.25 ng of each MHD enantiomer/mL of plasma. The method was applied in the study of kinetic disposition of oxcarbazepine and the MHD enantiomers in the steady state after oral administration of 300 mg/12 h oxcarbazepine in a healthy volunteer. The maximum plasma concentration of oxcarbazepine was 1.2 µg/mL at 0.75 h. The kinetic disposition of MHD is enantioselective, with a higher proportion of the S‐(+)‐MHD enantiomer compared to R‐(?)‐MHD and an AUC^0‐12 _{S‐(+)/R‐(?)} ratio of 5.44. Chirality 25:897–903, 2013. © 2013 Wiley Periodicals, Inc.     

In silico pharmacogenetics of warfarin metabolism

Pharmacogenetic approaches can be instrumental for predicting individual differences in response to a therapeutic intervention. Here we used a recently developed murine haplotype-based computational method to identify a genetic factor regulating the metabolism of warfarin, a commonly prescribed anticoagulant with a narrow therapeutic index and a large variation in individual dosing. After quantification of warfarin and nine of its metabolites in plasma from 13 inbred mouse strains, we correlated strain-specific differences in 7-hydroxywarfarin accumulation with genetic variation within a chromosomal region encoding cytochrome P450 2C (Cyp2c) enzymes. This computational prediction was experimentally confirmed by showing that the rate-limiting step in biotransformation of warfarin to its 7-hydroxylated metabolite was inhibited by tolbutamide, a Cyp2c isoform-specific substrate, and that this transformation was mediated by expressed recombinant Cyp2c29. We show that genetic variants responsible for interindividual pharmacokinetic differences in drug metabolism can be identified by computational genetic analysis in mice.     

High-performance liquid-chromatographic determination of warfarin enantiomers in plasma with automated on-line sample enrichment

A sensitive and selective chiral high performance liquid chromatographic method was developed for the direct determination of R- and S-warfarin enantiomers in human plasma. The method involved direct injection of human plasma onto a semipermeable surface (SPS) guard column, washing the proteins from the column with aqueous acetonitrile and back flushing the analytes onto a reversed phase ovomucoid silica HPLC column using switching valves. After separation, the analytes were simultaneously detected and quantitated with a fluorometer. The recoveries of R-warfarin from human plasma at 25 and 2500 ng/ml were 98.9% and 88.1%, respectively. The recoveries of S-warfarin at 25 and 2500 ng/ml were 105.4% and 93.9%, respectively. Using 100 microl of human plasma, the lower limit of quantification for both R- and S-warfarins was 25 ng/ml. Linear responses in analyte/internal standard peak height ratios were observed for analyte concentrations ranging from 25 to 2500 ng/ml for both enantiomers. Fluorescence chromatograms of drug-free human plasma showed no interfering peaks with retention times similar to those for R- and S-warfarins and the internal standard. Results from a 3-day validation study for both enantiomers demonstrated excellent precision (1.7-9.0%) and accuracy (97-109%) across the calibration range.     

Enantioselective assay for the determination of perhexiline enantiomers in human plasma by liquid chromatography

Effective use of the antianginal agent perhexiline is difficult because saturable metabolism by the polymorphic cytochrome P450 2D6 (CYP2D6) isoform produces elevated plasma perhexiline concentrations that have been associated with serious hepatic and neurological toxicity. Perhexiline is marketed for therapeutic use as a racemate and there is evidence for differences in the disposition of its enantiomers. The current study describes an enantioselective HPLC-fluorescent method utilising pre-column derivatization with (R)-(-)-1-(1-napthyl)ethyl isocyanate. Following derivatization, the enantiomers are resolved on a C18 column with gradient elution using a mobile phase composed of methanol and water. The method described is suitable for the quantification of (+)- and (-)-perhexiline in human plasma following clinical doses and demonstrates sufficient sensitivity, accuracy and precision between 0.01 and 2.00 mg/l for each enantiomer, with intra-assay coefficients of variation and bias <20% at 0.01 mg/l and <10% at 2.00 mg/l, and inter-assay coefficients of variation and biases <15% at 0.03 mg/l and <10% at 0.40 and 0.75 mg/l. The application of this method to plasma samples collected from a patient treated with perhexiline revealed that (+)-perhexiline concentrations were higher than (-)-perhexiline concentrations, confirming the stereoselective disposition of perhexiline. The current study describes an enantioselective method that utilises pre-column formation of fluorescent diastereomers that are resolved on a C18 HPLC column using a gradient of methanol and water.     

Comparison of warfarin therapy clinical outcomes following implementation of an automated mobile phone-based critical laboratory value text alert system

Background

Computerized alert and reminder systems have been widely accepted and applied to various patient care settings, with increasing numbers of clinical laboratories communicating critical laboratory test values to professionals via either manual notification or automated alerting systems/computerized reminders. Warfarin, an oral anticoagulant, exhibits narrow therapeutic range between treatment response and adverse events. It requires close monitoring of prothrombin time (PT)/international normalized ratio (INR) to ensure patient safety. This study was aimed to evaluate clinical outcomes of patients on warfarin therapy following implementation of a Personal Handy-phone System-based (PHS) alert system capable of generating and delivering text messages to communicate critical PT/INR laboratory results to practitioners' mobile phones in a large tertiary teaching hospital.

Methods

A retrospective analysis was performed comparing patient clinical outcomes and physician prescribing behavior following conversion from a manual laboratory result alert system to an automated system. Clinical outcomes and practitioner responses to both alert systems were compared. Complications to warfarin therapy, warfarin utilization, and PT/INR results were evaluated for both systems, as well as clinician time to read alert messages, time to warfarin therapy modification, and monitoring frequency.

Results

No significant differences were detected in major hemorrhage and thromboembolism, warfarin prescribing patterns, PT/INR results, warfarin therapy modification, or monitoring frequency following implementation of the PHS text alert system. In both study periods, approximately 80% of critical results led to warfarin discontinuation or dose reduction. Senior physicians' follow-up response time to critical results was significantly decreased in the PHS alert study period (46.3% responded within 1 day) compared to the manual notification study period (24.7%; P = 0.015). No difference in follow-up response time was detected for junior physicians.

Conclusions

Implementation of an automated PHS-based text alert system did not adversely impact clinical or safety outcomes of patients on warfarin therapy. Approximately 80% immediate recognition of text alerts was achieved. The potential benefits of an automated PHS alert for senior physicians were demonstrated.

     

Chromatographic profiles of tryptophan and kynurenine enantiomers derivatized with (S)‐4‐(3‐isothiocyanatopyrrolidin‐1‐yl)‐7‐(N,N‐dimethylaminosulfonyl)‐2,1,3‐benzoxadiazole using LC–MS/MS on a triazole‐bonded column

d ‐ and l ‐Tryptophan (Trp) and d ‐ and l ‐kynurenine (KYN) were derivatized with a chiral reagent, (S )‐4‐(3‐isothiocyanatopyrrolidin‐1‐yl)‐7‐(N,N‐dimethylaminosulfonyl)‐2,1,3‐benzoxadiazole (DBD‐PyNCS), and were separated enantiomerically by high‐performance liquid chromatography (HPLC) equipped with a triazole‐bonded column (Cosmosil HILIC) using tandem mass spectrometric (MS/MS) detection. Effects of column temperature, salt (HCO₂NH₄) concentration, and pH of the mobile phase in the enantiomeric separation, followed by MS detection of (S )‐DBD‐PyNCS‐d ,l ‐Trp and ‐d ,l ‐KYN, were investigated. The mobile phase consisting of CH₃CN/10 mM ammonium formate in H₂O (pH 5.0) (90/10) with a column temperature of 50–60 °C gave satisfactory resolution (R s) and mass‐spectrometric detection. The enantiomeric separation of d ,l ‐Trp and d ,l ‐KYN produced R s values of 2.22 and 2.13, and separation factors (α) of 1.08 and 1.08, for the Trp and KYN enantiomers, respectively. The proposed LC–MS/MS method provided excellent detection sensitivity of both enantiomers of Trp and KYN (5.1–19 nM).     

Clinical and genetic determinants of warfarin pharmacokinetics and pharmacodynamics during treatment initiation

Variable warfarin response during treatment initiation poses a significant challenge to providing optimal anticoagulation therapy. We investigated the determinants of initial warfarin response in a cohort of 167 patients. During the first nine days of treatment with pharmacogenetics-guided dosing, S-warfarin plasma levels and international normalized ratio were obtained to serve as inputs to a pharmacokinetic-pharmacodynamic (PK-PD) model. Individual PK (S-warfarin clearance) and PD (I(max)) parameter values were estimated. Regression analysis demonstrated that CYP2C9 genotype, kidney function, and gender were independent determinants of S-warfarin clearance. The values for I(max) were dependent on VKORC1 and CYP4F2 genotypes, vitamin K status (as measured by plasma concentrations of proteins induced by vitamin K absence, PIVKA-II) and weight. Importantly, indication for warfarin was a major independent determinant of I(max) during initiation, where PD sensitivity was greater in atrial fibrillation than venous thromboembolism. To demonstrate the utility of the global PK-PD model, we compared the predicted initial anticoagulation responses with previously established warfarin dosing algorithms. These insights and modeling approaches have application to personalized warfarin therapy.     

Chiral phase analysis of warfarin enantiomers in patient plasma in relation to CYP2C9 genotype

A direct chiral-phase high-performance liquid chromatographic method for measuring the ratio of S-warfarin/R-warfarin in patient plasma is described. Plasma samples are first extracted using solid-phase C₁₈ extraction columns, and the concentrated extracts analyzed using an (R,R) Whelk-O 1 column with a mobile phase of 0.5% glacial acetic acid in acetonitrile. The resulting chromatography provides baseline resolution of the warfarin enantiomers and internal standard (racemic ethylwarfarin), and is free from interference from other plasma components. Calibration curves were linear (mean r² of 0.999 for both enantiomers) over the concentration range 0.25–1.5 μg/ml. The intra-day and inter-day coefficients of variation for analysis of plasma spiked with 0.33 μg/ml S-warfarin and 0.67 μg/ml R-warfarin (S/R=0.5:1) was less than 7% for each enantiomer, with an accuracy of more than 93%. Plasma extracts from thirty-one patients homozygous for wild-type CYP2C9*1 provided an S/R ratio of 0.51±0.15. Two warfarin patients homozygous for the mutant CYP2C9*2 and CYP2C9*3 alleles exhibited elevated S/R ratios relative to the mean for individuals homozygous for the wild-type CYP2C9*1 allele. This method is suitable for population studies aimed at establishing the effect of polymorphic expression of CYP2C9 alleles on S-warfarin elimination in humans.     

HPLC separation of the enantiomers of nicotine and nicotine-like compounds

A sensitive and reproducible HPLC method utilizing a commercially available chiral α₁-acid glycoprotein (AGP) phase has been developed to separate and quantify the enantiomers of nicotine. The method is suitable for routine use as indicated by column life. The quantification of (R/S:0.05/99.95)-nicotine or (R/S:99/1)-nicotine was possible. In addition, the separation or at least partial separation of the enantiomers of nornicotine and nornicotine-derived compounds was achieved. © 1993 Wiley-Liss, Inc.     

Determination of MK-0767 enantiomers in human plasma by normal phase LC-MS/MS

A sensitive and selective analytical method for the enantioselective determination of MK-0767, a dual peroxisome proliferator-activated receptor (PPAR) alpha/gamma agonist, in human plasma has been developed and validated. The chromatography is based on normal-phase chiral separation on a Kromasil, 5 microm, CHI-DMB 250 mm x 4.6 mm column. The detection involves the direct introduction of the normal phase eluent into MS/MS without the addition of a post-column reagent. Atmospheric pressure chemical ionization (APcI) mode was selected as the ion source in this method. With proper sample handling and processing procedures, ex vivo interconversion of the enantiomers was kept to minimum during sample collection, preparation and short term storage of frozen human plasma samples. The method was successfully utilized to determine the concentrations of MK-0767 enantiomers in human plasma to support pharmacokinetic investigation in man.     

Determination of cyclophosphamide enantiomers in plasma by LC-MS/MS: Application to pharmacokinetics in breast cancer and lupus nephritis patients

This article describes the enantioselective analysis of cyclophosphamide (CPA) in human plasma using LC-MS/MS. CPA enantiomers were extracted from plasma using a mixture of ethyl acetate and chloroform (75:25, v/v). The enantiomers were separated on a Chiralcel(R) OD-R column, with the mobile phase consisting of a mixture of acetonitrile and water (75:25, v/v) plus 0.2% formic acid. The protonated ions and their respective product ions were monitored using two functions, 261 > 141 for CPA enantiomers and 189 > 104 for the internal standard (antipyrine). Recovery rates were higher than 95% and the quantification limit was 2.5-ng/ml plasma for both enantiomers. The coefficients of variation and the relative errors obtained for the validation of intra- and interassay precision and accuracy were less than 10%. The method was applied for the investigation of the enantioselective pharmacokinetics of CPA in a lupus nephritis patient treated with 1 g CPA infused over 2 h and in a breast cancer patient treated with 0.9 g infused over 1 h. No stereoselectivity in the pharmacokinetic parameters was observed for either patient. Clearance values of 2.63 and 2.93 l/h and of 3.36 and 3.61 l/h for (-)-(S) and (+)-(R)-CPA were obtained for the breast cancer and lupus nephritis patient, respectively.     

Enantiomerization and Enantioselective Bioaccumulation of Metalaxyl in Tenebrio molitor Larvae

The enantiomerization and enantioselective bioaccumulation of metalaxyl by a single dose of exposure to Tenebrio molitor larvae under laboratory condition were studied by high‐performance liquid chromatography tandem mass spectroscopy (HPLC‐MS/MS) based on a ChiralcelOD‐3R [cellulosetris‐tris‐(3, 5‐dichlorophenyl‐carbamate)] column. Exposure of enantiopure R‐metalaxyl and S‐metalaxyl in Tenebrio molitor larvae exhibited significant enantiomerization, with formation of the R enantiomers from the S enantiomers, and vice versa, which might be attributed to the chiral pesticide catalyzed by a certain enzyme in Tenebrio molitor larvae. Enantiomerization was not observed in wheat bran during the period of 21 d. In addition, bioaccumulation of rac‐metalaxyl in Tenebrio molitor larvae was enantioselective with a preferential accumulation of S‐metalaxyl. These results showed that enantioselectivity was caused not only by actual degradation and metabolism but also by enantiomerization, which was an important process in the environmental fate and behavior of metalaxyl enantiomers. Chirality 26:88–94, 2014. © 2013 Wiley Periodicals, Inc.     

Steady-state clearance rates of warfarin and its enantiomers in therapeutically dosed patients

Previous studies to identify the pharmacokinetics of R- and S-warfarin have not used steady-state area under the curve (AUC) data during therapeutic doses of racemic warfarin. Instead they have used high single doses of either racemic warfarin or a single enantiomer in volunteers or have taken a single blood sample from anticoagulated patients and assumed full compliance and a steady-state status. In this study, a series of steady-state racemic warfarin, R-warfarin, and S-warfarin serum concentrations, during a 24 h dosage interval, was measured in 10 compliant patients (5 females and 5 males) taking racemic warfarin. The anticoagulation status of all 10 patients according to the International Normalised Ratio (INR) was stable. Their mean (SD) age and weight were 67.0 (9.9) yr and 63.9 (15.4) kg. The mean (SD) clearances derived from steady-state AUC values, following therapeutic dosing, for racemic warfarin, R-warfarin, and S-warfarin were 2.40 (0.82), 2.30 (0.65), and 2.80 (1.17) ml/h/kg, respectively. The mean (SD) ratio of S-:R-warfarin clearance was 1.24 (0.40). Comparison of the clearance measured from the AUC, of these patients, to one point determinations assuming steady state for the samples drawn at either 6, 15, or 20 h after dosage (during the dosing interval) showed some statistical differences. Most single point determinations of warfarin clearance assume that a sample 12 h postdose is equivalent to that of the steady-state concentration, but in this study the steady-state concentration of only 6 patients occurred between 6 and 15 h postdose. This could explain why these studies demonstrate differences in the clearance of R- and S-warfarin compared to the values we have derived from steady-state AUC data using patients with proven compliance and therapeutic doses. Chirality 9:13–16, 1997. © 1997 Wiley-Liss, Inc.