Search

Crotethamide and cropropamide, both components of the respiratory stimulant prethcamide, are metabolized in humans by demethylation of the [(dimethylamino)-carbonyl]-propyl moiety. The resulting metabolites are characterized by gas chromatography-mass spectrometry of urinary extracts. The use of HCl to prevent losses by volatilization during the evaporation step, combined with methanol as solvent, complicates gas liquid chromatographic analysis of prethcamide. The resulting artifacts are identified.

A quantitative determination of physiological levels of all-trans-retinoic acid (vitamin A acid) in human serum has been developed. A double-phase extraction of 3.5 ml of serum followed by a specific and sensitive high performance liquid chromatography procedure allowed measurement of levels down to 1 ng/ml. Serum concentrations in 37 fasting volunteers ranged from 2.7 to 4.2 ng/ml and fitted a normal-Gaussian distributional shape with a mean value of 3.5 ng/ml and SD of 0.4 ng/ml, as demonstrated by the Kolmogorov-Smirnov test.

A sensitive and rapid method for the gas chromatographic (with electron-capture detection) confirmation of derivable sympathomimetic amines is described. Extractive derivatization with pentafluorobenzoyl chloride is performed on 2-ml urine or plasma samples. Especially for primary amines, the method appears to be very sensitive. Mass spectral data allowed confirmation of the monobenzoylation of all congeners.

A new, sensitive and selective detection method for vitamin K1 in human serum is reported. After the chromatographic separation, vitamin K is converted to its hydroquinone form in a wet-chemical post-column reduction reaction with tetramethylammonium octahydridotriborate. The reaction proceeds in an open tubular knitted reaction coil at elevated temperature. This system, when combined with a two-step multidimensional liquid chromatographic separation, has proven useful for the quantitative determination of serum concentrations of vitamin K1 in normal, healthy individuals.

Bromhexine hydrochloride was administered to dogs i.v. and p.o. Plasma determinations of the unchanged drug were carried out by gas-liquid chromatography with electron-capture detection. The plasma levels obtained after i.v. administration were consistent with a two-compartment open model. Comparison of the results of i.v. and oral administration show a low bioavailability, probably due to a first pass effect.

A specific GLC method was developed for the determination of floxuridine in plasma using the thermionic nitrogen detector. The method involves the isolation of the compound and internal standard from plasma on a strong anion-exchange column at pH 10, followed by elution with 0.3 M acetic acid in methanol. The eluate is evaporated to dryness, and the residue is dissolved in dimethyl sulfoxide and permethylated with potassium tert-butoxide and methyl iodide. The permethylated compounds are reextracted from the reaction mixture with cyclohexane-methylene dichloride (9:1). The organic solution is evaporated to dryness, the residue is dissolved in ethyl acetate, and an aliquot is analyzed by GLC on a 3% OV-17 column. The extraction recovery from spiked plasma was 93.2 +/- 2.1% (SD), whereas linearity for the overall procedure was in the 0-1-microgram/ml range. The detection limit of the thermionic nitrogen detector was 50 ng/ml. The within-run and within days precision (CV) were 4.0 and 6.2%, respectively, at 300 ng/ml.

Weanling rats were fed vitamin A deficient diets (-A) or diets supplemented with vitamin A (+A) (4.4 mg retinol equivalents/kg diet) for a period of 7 or 6 wk, respectively. In liver tissues of these two groups of animals both the subcellular localization as well as the fatty acid composition of the retinyl esters was studied. During vitamin A supplementation or deprivation, the kinetics of the different ester forms were investigated. Results indicate that the subcellular localization of all retinyl esters is similar and dependent on age. Two pools exist, ie one consisting of the nuclear/cell debris and mitochondrial-lysosomal fractions and the other containing the microsomal and cytosol fractions. HPLC analysis showed retinyl palmitate as the predominating (80%) form of the various retinyl esters. By supplementation clearly two kinetic behaviours can be demonstrated: one being a relatively stable storage of the palmitate and stearate, increasing with time and the second one being a more labile pattern for the ester forms with other saturated and unsaturated fatty acids. By vitamin A depletion all retinyl esters are affected indicating that the ester forms other than palmitate and stearate are also storage forms of vitamin A.

As a substantial part in the development of a reference material in clinical chemistry, a highly accurate and precise isotope dilution mass spectrometric method has been worked out for the determination of cortisol in human serum. The candidate definitive method consists of addition of 4-[14C]cortisol and, following equilibration, solvent extraction of cortisol and its internal standard from the serum matrix. After conversion into methoxime-trimethylsilyl derivatives, the extract is purified by gel chromatography. Measurements are made by combined capillary gas chromatography mass spectrometry. The ratio of peak heights at m/z 605 and 607 for each sample and calibration mixtures is used for quantification. To ensure maximum accuracy each set of calibration mixtures was composed from two stock solutions and closely bracketed the anticipated serum concentration. An analysis of variance, involving comparison of within-run and between-run variability, gave a total coefficient of variation (CV) of 0.38% for a serum pool containing 90.79 ng cortisol ml-1 serum.

Serum levels of 5-fluoro-2'-deoxyuridine in cancer treated patients were measured by gas chromatography mass spectrometry under chemical ionization conditions; 1-(2-deoxy-beta-D-lyxofuranosyl)-5-fluorouracil (3'-epi-5-fluoro-2'-deoxyuridine) was used as an internal standard. The drug and internal standard were quantitatively isolated from the serum sample by a mini-column anion exchange method and the extract permethylated using potassium-tert-butoxide in dimethylsulphoxide and methyl iodide. The derivatized nucleosides were then re-extracted from the reaction mixture and analysed on a glass capillary column coated with Superox-4. The column was coupled directly to the chemical ionization source of the mass spectrometer; NH3 was used as the reagent gas. The gas chromatographic effluent was monitored at m/z 289, the [MH]+ ion of the N,O-permethyl derivatives of 5-fluoro-2'-deoxyuridine and the internal standard. Recovery of 5-fluoro-2'-deoxyuridine from serum in the 0-1 microgram ml-1 concentration range averaged 93 +/- 2% (SD); a linear detector response was observed up to 50 ng 5-fluoro-2'-deoxyuridine ml-1. At the 10 ng ml-1 level, a within-run assay precision of 10% (CV) (n = 5) was found, while a detection limit of about 1 ng 5-fluoro-2'-deoxyuridine ml-1 of serum was attained. The method was applied to the measurement of disappearance curves of 5-fluoro-2'-deoxyuridine in the serum of treated patients.

For this sensitive RIA for 1 alpha,25-dihydroxyvitamin D, we used antibodies to 1 alpha,25-dihydroxycholecalciferol-3-hemisuccinate conjugated to bovine serum albumin, raised in rabbits. These antibodies show a high affinity for 1 alpha,25-dihydroxyvitamin D3 but cross react with other vitamin D metabolites as well. Extraction and liquid chromatography are required to isolate the 1 alpha,25-dihydroxyvitamin D from human serum or plasma with benzene (analytical recovery, 96%) and chromatographed by "high-performance" liquid chromatography on a silica column. The sensitivity of the assay is 2 pg/tube. The between-assay CV is 12.6% (means = 61.8 ng/L, n = 27) and the within-assay CV is 7.4% (means = 61.8 ng/L, n = 15). The reference interval, obtained from data on serum of healthy persons, is 51.8 +/- 15.1 ng/L (means +/- SD, n = 91), with no sex-related difference. Patients with chronic renal failure have lower concentrations in serum, and patients suffering from calcium nephrolithiasis have increased concentrations. Our assay offers the advantage of a relatively short prepurification procedure and the production of large amounts of stable gamma-globulins that can be used for several years without loss of binding properties.

Untransformed retinoic acid has never been demonstrated in human excreta under normal physiological conditions. We have developed a two-step liquid chromatographic system for the demonstration of subnanogram amounts of this compound in human urine without administration of any precursor.

A liquid chromatographic procedure for the quantitative determination of the thiazine dye azure B, the principal constituent of Romanowsky stains, is presented. Unlike previous methods relying on peak area normalization, the present approach involves real quantitation through calibration with the reference standard azure B tetrafluoroborate. The method has been used for the quality control of commercial azure B preparations and to study their stability in stock and staining solutions, either or not in the presence of eosin Y. Results suggest that highly pure azure B perchlorate meets the requirements of a reference material, useful for standardization of Romanowsky-Giemsa staining in haematology.

We describe a sensitive and specific liquid-chromatographic assay for pentoxifylline and three of its metabolites in human plasma and urine. Addition of hydrochloric acid to the sample before extraction, and incorporation of acetic acid in the chromatographic eluent, allow the simultaneous determination of the four compounds plus an internal standard in one chromatographic run. Unlike gas-chromatographic procedures, this method does not involve derivatization no similar analysis of serum or plasma samples has been described before now. The method has been applied successfully to routine analysis and to pharmacokinetic studies.

Chromatographic procedures have been developed for determination of the stabilizers N-acetyl-DL-tryptophan and octanoic acid in human albumin solutions. N-Acetyl-DL-tryptophan and the internal standard, N-formyl-DL-tryptophan, were separated by liquid chromatography on a reversed-phase column with UV detection at 280 nm. Deproteinization and extraction were carried out with methanol. The extraction recovery at the level of 4.9 mM was 92.5 +/- 2.5% (S.D.) (n = 10), and the average coefficient of variation (C.V.) for replicate analyses of albumin solutions (mean = 2.57, 10.44 and 17.10 mM) was 1.10% (n = 27). Octanoic acid was determined gas chromatographically as its methyl ester, with nonanoic acid as the internal standard. The sample pretreatment included acidification, extraction with hexane and derivatization with methanol-sulphuric acid. The relative recovery from albumin solutions was 89.7 +/- 5.8% (S.D.) (n = 6), and replicate determinations of the compound yielded a C.V. of 5.5% (mean = 14.82 mM, n = 9).

Several conflicting observations in the literature raised considerable doubt about the metabolic fate of doxycycline, which, like other tetracyclines, has been claimed to be metabolically inert. A double liquid chromatographic approach was used in an attempt to demonstrate the polar metabolites and/or conjugates in excreta of human volunteers who ingested the drug. Both ion-exchange and reversed-phase chromatography failed to reveal significant by-products in urine and feces, except for minor amounts of 4-epidoxycycline. In addition, enzymatic hydrolysis procedures did not present any evidence of the conjugates. Thus, the different excretion behavior of doxycycline, compared to other analogs, cannot be explained in terms of increased metabolism.

A method is described for the quantification of tulobuterol in human plasma, based upon selective extraction and high resolution capillary gas chromatography-mass spectrometry. The mass spectrometric determination is based on selected ion monitoring of two fragment ions derived from the N,O-heptafluorobutyryl derivatives of tulobuterol and its internal standard, deschlorotulobuterol. The plasma work-up procedure consists of extraction of 1.0 ml alkalinised plasma with dichloromethane, back-extraction into an acidified aqueous phase, followed by a final extraction with dichloromethane after realkalinisation. The method gave interference-free and linear results. The between-assay variability is 4.7% CV at the 3.0 micrograms l-1 plasma concentration level. The assay permitted quantitative measurements down to 170 ng tulobuterol per litre of plasma and hence provided sufficient sensitivity for quantification of plasma levels after therapeutic doses.

Combined gas chromatography-mass spectrometry with multiple-ion detection is used to quantitate vitamin D3 in plasma samples. Extensive cleanup of the plasma extract prior to analysis is required before mass fragmentographic analysis is possible, i.e., solvent extraction, digitonin precipitation, Lipidex column chromatography, and derivatization. Low resolution mass fragmentography is performed by monitoring simultaneously the molecular ions of the heptafluorobutyric esters of the all-trans isomer of vitamin D3 and dihydrotachysterol2, which is used as an internal standard. This new method uses 5 ml of plasma and it is specific and sensitive to 600 pg of vitamin D3. This assay shows that in plasma of normal adult subjects there is a vitamin D3 concentration of 5 to 11 ng/ml, lower than those values reported for biological and competitive protein binding assays. Specifically the method described offers potential as an eventual reference method.

Human plasma and serum levels for 1 alpha,25-dihydroxyvitamin D were determined by a cytosol radioreceptor assay (RRA) and a radioimmunoassay (RIA). For both assays, 1.5 ml of human serum or plasma is used. Prior to RRA or RIA, extraction with benzene is performed followed by 'high-performance' liquid chromatography (HPLC) on a silica column (25 X 0.46 cm) with hexane/isopropanol (9/1 by vol), to isolate 1 alpha,25-dihydroxyvitamin D from the other vitamin D metabolites. The cytosol receptor was isolated from the intestine of healthy chickens. The antisera were raised in rabbits to 1 alpha,25-dihydroxyvitamin D3-3-hemisuccinate coupled to bovine serum albumin. The standard curves for RRA and RIA are prepared with 1 alpha,25-dihydroxyvitamin D3. 1 alpha,25-dihydroxy[3H]vitamin D3 of high spec act (158 kCi/mol) is used as tracer. The reactants are incubated for 16 h at 4 degrees C. Then, bound and free ligand are separated after the addition of dextran-coated charcoal. Both assays have a sensitivity of 2 pg/tube. The cytosol receptor and the antibodies have about the same absolute affinity for 1 alpha,25-dihydroxyvitamin D3 but the cytosol receptor has a higher relative affinity for 1 alpha,25-dihydroxyvitamin D3 (compared with other vitamin D metabolites). Reproducibility and precision are better for the RIA. The between- and within-assay CVs are 16.0% (mean = 58.7 ng/l, n = 16) and 11.2% (mean = 52.1 ng/l, n = 15), respectively, for RRA and 12.6% (mean = 61.8 ng/l, n = 27) and 7.4% (mean = 61.8 ng/l, n = 15), respectively using RIA. Reference values obtained by both assays on healthy males and healthy premenopausal females are the same for both sexes; 53.9 +/- 31.0 ng/l (n = 46) using RRA and 51.8 +/- 30.2 ng/l (n = 91) for RIA (mean +/- 2 SD).

We could not confirm the results of a previous research group which stated that retinal (vitamin A aldehyde) is present in corpus luteum of bovine origin. On the contrary we establish the occurrence of retinol (vitamin A alcohol) in both human and bovine corpora lutea. Identity of the compound is confirmed by its chromatographic behavior on high performance liquid chromatographic systems (adsorption and reversed phase) and by spectrometric analysis (UV, fluorimetry).

A new method, based upon a selective extraction and gas chromatographic/mass spectrometric analysis, was developed to monitor the effect of combined haemoperfusion-haemodialysis treatment in a case of propericiazine poisoning. The method relies on the selected ion monitoring of the acetate derivatives of propericiazine and its internal standard fluphenazine, after their extraction from 1 ml of alkalinized plasma with n-hexane:isopropanol (8:2, v/v), back-extraction into an acidified water phase, realkalinization and extraction with n-hexane: isopropanol, derivatization with acetic anhydride and gas chromatography on a short (12 m) OV-101 fused silica capillary column. The described procedure is specific and provides between-assay variability of 4.8% CV at 5 micrograms 1(-1) plasma concentration. The method enables quantification down to 1 microgram 1(-1) and hence demonstrates sufficient sensitivity to permit pharmacokinetic or drug monitoring studies.

This is a new method for isolating and concentrating urinary organic acids before gas chromatography--mass spectrometry. Sulfate and phosphate anions are removed by precipitation with Ba(OH)2, and the urine pH is adjusted to 8-8.5. The sample is then applied onto small, disposable, strong-anion-exchange columns. Neutral and basic compounds are washed out with water, which is then removed by centrifugation and by rinsing the columns with methanol and diethyl ether. The organic acids are eluted with a 4/1 (by vol) mixture of an organic solvent--either n-butanol or ethyl acetate--and formic acid containing HSO4- (0.1 mol/L) as a highly selective counter-ion, and finally with methanol alone. Sulfate ions are retained, and the eluate is evaporated before trimethylsilyl derivatization. Analytical recoveries (about 100%) of organic acids compare favorably with those obtained with solvent extraction. This convenient procedure is selective and reproducible and is a suitable alternative to the more cumbersome diethylaminoethyl-Sephadex extraction method.

A specific and sensitive GLC analysis of bromhexine in human plasma is described. After addition of the internal standard and an aqueous triethanolamine solution, bromhexine is extracted at alkaline pH into n-hexane, transferred to an acidic aqueous solution, and back-extracted into n-hexane after alkalinization. Both compounds are derivatized with trifluoroacetic anhydride and quantified by GLC using , 63Ni-electron-capture detector. The method has a sensitivity of approximately 1.0 ng/ml of plasma. Linearity of plasma working curves was good. The extraction recovery from spiked plasma was 90.1 +/- 5.68% (SD). The within-run and within-day precisions (CV) were 6.0% (4.3 ng/ml, n = 8) and 8.6% (11.0 ng/ml, n = 13), respectively. The procedure was applied successfully to measurement of the plasma concentration-time profile in a human volunteer after oral drug administration.

A specific gas-liquid chromatographic method with electron-capture detection is described for the determination of bromhexine at the nanogram level. Structurally analogous internal standards were synthesized and their suitability was investigated, based on their chromatographic properties on different stationary liquid phases. The electron-capture activity of the compounds is increased by trifluoroacetylation. Reaction conditions for this derivatization were studied. Calibration graphs in the range 0-57 ng (in a total of 100 microliters of reaction mixture) showed good linearity.

The presence of p-chlorophenoxyacetamide was detected in the urine of man receiving iproclozide [1-(p-chlorophenoxyacetyl)-2-isopropylhydrazine]. This new metabolite was identified by combined gas-liquid chromatography-mass spectrometry and by comparison with the synthetic compound. An appropriate procedure for the extraction and quantitation of p-chlorophenoxyacetamide by gas-liquid chromatography on a 5% OV-225-packed column with the 2-butyl analog of iproclozide as an internal standard, has been developed. After 20- and 50-mg single dose oral administrations of iproclozide, 5.2 and 8.3% were slowly excreted in urine as p-chlorophenoxyacetamide within 30.5 and 36 hr. respectively. After 20-mg oral intakes by psychiatric patients, the 24-hr urinary excretion of p-chlorophenoxyacetamide amounted to about 2-4% of the iproclozide dose administered.

Fluorouracil in human plasma of a concentration as low as 2 ng ml-1 can be assayed with the use of gas-liquid chromatography and a multiple ion detector. Fluorouracil is isolated from plasma by anion exchange extraction. The extract is butylated and the derivative obtained extracted into a mixture of cyclohexane + dichloromethane (95:5). An aliquot is submitted to combined gas-liquid chromatography mass spectrometry. The use of chlorouracil or oxygen-18 labelled fluorouracil as internal standard is discussed. Linearity is proven from 100 microgram to 5 ng of fluorouracil per ml plasma. The coefficients of variation are 7.2 and 8.4% for within-run and between-run precision, respectively, at the 50 ng ml-1 level. The biological applicability of our procedure is demonstrated by analysing plasma samples obtained at different time intervals from three patients given 1 g of fluorouracil once intravenously and once orally.

A multi-stage screening and identification scheme for the diagnosis of inborn errors in amino acid, fatty acid, carbohydrate and intermediary metabolism is described. The method is based on a computerized analysis of data obtained with dual-column gas chromatography/FID (GC/FID) and gas chromatography/mass spectrometry (GC/MS) profiling of urinary organic acids. It involves: (1) isolation of the compounds by solvent or solid-phase extraction; (2) conversion into trimethylsilyl (TMS) or TMS-oxime derivatives; (3) GC/FID analysis on SE-52 and OV-1701 capillary columns; (4) tentative identification by comparing the methylene unit (MU) values on both columns with a user-built library of reference compounds; (5) quantitative evaluation of the excretion profile; and (6) analysis by GC/MS of samples with an abnormal profile using an automated peak identification programme.

Owing to the lack of sensitivity and/or selectivity of the existing chemical assays, vitamin K deficiency has always been diagnosed indirectly by measuring its effect on blood coagulation. We used our recently developed multidimensional liquid chromatographic assay for what is, to the best of our knowledge, the first systematic investigation of physiological vitamin K levels in human blood. It allowed the unequivocal demonstration of trans-phylloquinone (vitamin K1) and its quantification down to a level of 0.5 nanogram per milliliter of serum (ng/ml. In healthy adults, a mean serum concentration of 2.6 ng/ml was found, with a normal range of 0.9 to 7.8 ng/ml. These values apparently are distributed in a log-normal way.

We report a fast and simple high-speed liquid-chromatographic assay for serum retinol. Only 100 mul of serum is required. The lower detection limit is 50 mug/liter; linearity was demonstrated up to 1.50 mg/liter. On analyzing a serum pool eight times, a CV of 2.5% was obtained. Values by this method are compared with results obtained by a flurometric method [Clin. Chem. 16, 766(1970)].

We describe how riboflavin can be precisely and accurately measured in serum and urine. A structural analog, isoriboflavin, is used as an internal standard. Urine samples are prepared by adding the internal standard and trichloroacetic acid. For serum, proteins are denatured with trichloroacetic acid. A simple Sep-pak treatment of the supernate removes contaminating interferents. Within- and between-run precision is reflected by respective CVs of 2.2 and 4.9% for urine at 180 micrograms/L and 4.4 and 7.3% for serum at 10 micrograms/L. The standard curve is linear far beyond the concentrations encountered in serum and urine. The detection limit is estimated to be 10 micrograms/L and 1 microgram/L for urine and serum, respectively. The normal reference interval, as determined from 50 results for each matrix, is 36 to 349 micrograms/g of creatinine (mean 112 micrograms/g) for urine and 5.5 to 14.4 micrograms/L (mean 8.8 micrograms/L) for serum. Both distributions are skewed. The method is suited for routine use.

Laboratory results submitted to External Quality Assessment Schemes (EQAS) are evaluated against the arithmetic mean (m) and standard deviation (SD) of the result from other participants assuming a normal distribution of individual results. The sample statistics m and SD can only be reliable estimates of the theoretical mean (mu) and standard deviation (sigma), if they are not distorted by outlying values. Outliers are commonly detected by defining a confidence interval m +/- k SD, typically with k-values fixed at 3.0 or 3.2. The objective of our study was to prove the deficiency of the use of these fixed limits in large-scale EQAS, because it results in an unacceptable increase of outlier rate and hence in a serious distortion of the estimate of sigma. We have proposed a modified approach using a variable k-value which is dependent upon sample size and keeping the significance level constant at 5%. The influence of different decision limits on the number of outliers and on the sample statistics m and SD was compared using data from the Belgian National EQAS obtained in 6 consecutive surveys.

In order to study the biological fate of all-trans- and cis-canthaxanthin in the brine shrimp Artemia, a quantitative method was developed for the determination of both carotenoids and their metabolic precursors in encysted embryos (cysts), nauplii, whole animals, organs, and subcellular fractions. This method is based on nonaqueous reversed-phase chromatography, two new exhaustive extraction procedures, and the determination of proteins in the extracted residue. Hydration of Artemia cysts caused a reversible conversion of part of the all-trans- to cis-canthaxanthin. During further pre-emergence embryonic development, there was little change in the levels of both isomers. After hatching of cysts, cis-canthaxanthin was progressively isomerized to the all-trans form, while the total (all-trans + cis) canthaxanthin to protein ratio tended to remain constant. Cis-canthaxanthin rapidly became undetectable in animals fed on algae and reappeared in females at an advanced stage of the reproductive cycle. All-trans-canthaxanthin remained present during the whole Artemia life cycle in addition to its metabolic precursors echinenone and beta-carotene. The carotenoid distribution in organs and subcellular fractions indicated high affinity of cis-canthaxanthin for the female reproductive system, oocytes in general, and yolk in particular. A role for cis-canthaxanthin is suggested at an early developmental stage, i.e., in cysts, before hatching.

A liquid-chromatographic assay for alpha, beta, and gamma-isomeric tocopherols in human serum is reported. The tocopherols and the internal standard (tocol) are absorbed into a silica gel column and are eluted in less than 10 min with n-hexane/isopropanol (99.4/0.6, by vol) at a flow rate of 1 mL/min. The complete analysis requires no longer than 30 min. Within-day precision (CV) was 1.4% (mean = 13.18 mg/L, n = 24), 7.4% (mean = 0.214 mg/L, n = 14), and 1.3% (mean = 1.01 mg/L, n = 14) for alpha-, beta-, and gamma-tocopherol, respectively. Day-to-day precision (CV) was 4.4% (mean = 9.85 mg/L, n = 10) for alpha-tocopherol, 9.1% (mean = 0.222 mg/L, n = 20) for beta-tocopherol, and 3.8% (mean = 1.00 mg/L, n = 20) for gamma-tocopherol. Extraction recoveries for alpha-, beta-, and gamma-tocopherol averaged 92.4 +/- 2.9% (n = 5), 91.4 4/- 7.6% (n = 5), an 92.0 +/- 4.1% (n = 4), respectively. The smallest injected amount detectable is estimated to be 0.03 micrograms for alpha-, and 0.04 micrograms for beta- and gamma-tocopherol.

A procedure is described for the separation of Vitamin D and related compounds by gas-liquid chromatography using flame ionization detection. The method involves a two-step conversion: isomerization to the corresponding (all trans) isotachysterol(s) followed by methyl ether derivatization of the alcohol group(s). The procedure provides a means for identification as well as a possible basis for quantitation of Vitamin D and analogs.