Anahst, February, 1979, Vol. 104, pp. 117-123 117 Fluorimetric Determination of Acetohexamide in Plasma and Tablet Formulations Using 1-Methylnicotinamide Pamela Girgis-Takla and loannis Chroneos Welsh School of Pharmacy, University of Wales Imtitztte of Science and Technology, King Edward V I I Avenue, Cardifl, CF1 3NU A sensitive method is described for the fluorimetric determination of aceto- hexamide in plasma or in tablets by means of its reaction with l-methyl- nicotinamide, which is shown t o be a useful reagent for the determination of ketonic compounds. The limit of detection is approximately 0.2 pg ml-l and the relative standard deviation is 3.1% for 2 pg ml-l in plasma. Acetoacetic acid usually does not interfere, but can be separated, if necessary, from acetohexamide by means of a washing technique.No interference is caused by the presence of insulin, other (non-ketonic) oral hypoglycaemic drugs, acetone or pyruvic or a-ketoglutaric acid, Keywords : A cetolaexamide determination ; plasma ; tablets ; 1- fiaetlzylnicotin- amide reagent ; spectrofluorimetry The finding1 that the oral hypoglycaemic compound acetohexamide can exist in at least two polymorphic forms made it of interest to develop a sensitive and relatively specific method for the routine determination of plasma levels of the unchanged drug, especially in view of a recent report2 that marked variations in bioavailability can exist between different batches of tablets containing the related sulphonylurea tolbutamide. Serum concentrations of acetohexamide can be measured colorimetrically by a modification3 of Spingler's pro- ~ e d u r e . ~ Both the drug and its metabolite, hydroxyhexamide, can be determined in serum or plasma by isotope dilution analysis,5 or by a two-component spectrophotometric pro- cedure6 utilising measurements at 247 and 228 nm, a disadvantage of the last method being that blank absorbance values are high.A more sensitive gas - liquid chromatographic method has been proposed by Kleber et aL7 for which the calibration concentrations of the drug and its metabolite range from 5 to 40 pg ml-l in plasma. The method described here is suitable for plasma samples containing 0.5-2.5 pg ml-l of acetohexamide. It introduces a new use for 1-methylnicotinamide as a reagent for the fluorimetric determination of ketones, with which it is known8 to react readily, after treatment with an alkali in order to convert it into the highly reactive a-carbinol.The reaction has previously been applied only to determining 1-methylnicotinamide by condensation in the presence of an alkali with a variety of different methyl ketones. Huff and Perlzweig8 and Carpenter and Kodiceks have described similar procedures for 1-methylnicotinamide, which involve condensation in aqueous alkali, followed by acidification to about pH 0.5 with hydrochloric acid, a short period of heating and finally the addition of potassium dihydrogen orthophosphate in order to buffer the mixture at about pH 2. In another modification, Clark et a1.l0 converted the 1-methylnicotinamide into a fluorescent derivative by treatment with acetophenone in alcoholic potassium hydroxide, and then acidified the solution with 99% formic acid.The acidification in each procedure reversibly changed the fluorescence from greenish blue to blue and enhanced it. The method developed for acetohexamide in plasma can also be applied to the determina- tion of the drug in tablets, and is simpler and less time consuming than the high-performance liquid chromatographic method of Beyer,ll or the Salim and Hilty spectrophotometric assay,12 which is the official method of the United States Pharma~opeia.1~ Other techniques that have been proposed for tablet preparations include non-aqueous titration,l4>l5 colorimetry using either cobalt acetate16 or 2 ,4-dinitrophenylhydrazine17 and polarography.18118 GIRGIS-TAKLA AND CHRONEOS : FLUORIMETRIC DETERMINATION OF Apparatus Fluorimetric measurements were carried out on a Perkin-Elmer, Model 1000, fluorescence spectrophotometer in 1 x 1 cm silica cells with a 371-nm filter at an angle of 20" for excita- tion and an emission wavelength setting of 437 nm, slit width N and scale expansion x 1 or x 2.The infrared spectra were recorded from potassium bromide discs using a Perkin-Elmer, Model 357, grating spectrometer. Spectrophotometric measurements were made using a Unicam SP500 Series 2 spectro- photometer and 1-cm silica cells. Ultrafiltration of plasma solutions of acetohexamide was carried out using Amicon Centri- flo Membrane Cones CF 50A with conical supports and 50-ml centrifuge tubes. Analyst, Vol.104 Experimental Reagents of fluorescent contaminants. All reagents were of analytical-reagent grade, and were checked before use for the presence Acetohexamide. 1-Methylnicotinamide iodide reagent, 3% m/V in lo-* M hydrochloric acid. This solution should be freshly prepared. Dissolve 9 g of nicotinamide in 50 ml of dimethylformamide and add 10ml of methyl iodide. Allow the mixture to stand, preferably overnight, then separate the product by filtration, wash it with about 50 ml of dimethylformamide and dry it in air (melting-point 205-208" C). Formic acid solution, 50% V/V. Hydrochloric acid, 2 M and 10% m/m. Sodium hydroxide solutiozzs, 0.01, 0.1 and 5 M. Chloroform. De-ionised water was used in the preparation of all solutions. Kindly supplied by Eli Lilly and Company Ltd.Prepared from 98-100% m/m formic acid. Preparation of hydroxyhexamide A solution of 1 g of acetohexamide in 60 ml of 2yo m/V sodium hydroxide solution was prepared and 150 mg of sodium tetrahydroborate( 111) were added and dissolved by shaking. The solution was left to stand for 1 h at room temperature and then acidified with 40 ml of 10% m/m hydrochloric acid. The product was separated immediately by filtration, washed with approximately 150 ml of water and recrystallised from dilute ethanol] to give 0.7 g of h ydroxyhexamide ( N - [ (cyclohexylamino) carbonyll-4- (1 -hydroxyethyl) benzenesulphon- amide), melting at 146-149 "C. The compound was identified by thin-layer chromatography on pre-coated silica gel 60 F,,, plates with chloroform - formic acid (97 + 3) as the solvent system6 and by nuclear magnetic resonance and infrared absorption spectrometry.Thin- layer chromatography showed a trace amount of acetohexamide in the product that was difficult to remove by further recrystallisation. The infrared spectrum (Fig. 1) showed the presence of an OH band at 3520 cm-l. The slight shoulder a t 1690 cm-1 provided evidence that a carbonyl group had been reduced when the spectrum was compared with that of Wave len g t h/p m 2.5 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10 12 14 16 20 30 40 100 -7- 80 a- & 60 E 40 4-J + .- c p 20 0 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200 W avenumber/cm-' Fig. 1. Infrared spectrum of hydroxyhexamide.February, 1979 ACETOHEXAMIDE IN PLASMA AND TABLETS USING 1-METHYLNICOTINAMIDE 119 acetohexamide polymorph B,l as this shows a clear doublet with peaks a t 1690 and 1665 cm-1.Attempts to prepare hydroxyhexamide on a small scale by catalytic hydro- genation, which has been used by previous workers,lSJO were not successful as the reduction proceeded too far, and the product was N- [ (cyclohexylamino)carbonyl]-4-ethylbenzene- sulphonamide. Procedures Assay of acetohexamide in plasma Mix 2.0 ml of plasma with 1.5 ml of 2 M hydrochloric acid in a glass- stoppered tube. Add 20.0ml of chloroform, shake thoroughly and allow to stand for at least 30 min in order to allow the phases to separate. Transfer the bulk of the chloroform layer into a stoppered centrifuge tube and centrifuge. Measure 5.0ml of the chloroform extract into a 10-ml calibrated flask and evaporate to dryness in a stream of nitrogen at room temperature.(Up to three 5-ml portions of the chloroform extract can be evaporated to dryness, and subjected separately to the fluorimetric procedure to improve the confidence limit of the assay.) Dissolve the residue in the 10-ml flask in 0.10 ml of 0.01 M sodium hydroxide solution, add 0.10 ml of 1-methylnicotinamide reagent, followed immediately by 0.10 ml of 5 M sodium hydroxide solution and mix thoroughly. Exactly 2 min after addition of the 5 M sodium hydroxide solution make up to volume with 50% V/V formic acid. Measure the fluorescence intensity (maximum value) after 1.5-2 h. Correct the observed fluorescence by subtracting the fluorescence intensity measured using the same procedure on a drug-free plasma sample taken from the same subject prior to drug administration.Determine the concentration of acetohexamide in the plasma by reference to a calibration graph obtained by carrying out the assay procedure using 2.0-ml aliquots of standard solutions of acetohexamide in 0.01 M sodium hydroxide solution or in plasma containing 0, 0.5, 1.0, 1.5, 2.0 and 2.5 pg ml-l of drug. [Fluorescence intensity values after subtraction of the appropriate blank (zero drug concentration) are the same, irrespective of whether the drug is dissolved in plasma or in aqueous alkali.] If necessary, dilute the test fluorescent solution with 50% V/V formic acid, in order to bring the fluorescence intensity within the range of the calibration graph. Determine the approximate concentra- tion of acetohexamide in the plasma by reference to the calibration graph.Repeat the assay using a dilution of the plasma in 0.01 M sodium hydroxide solution accurately prepared to contain about 2.0 pg ml-l of acetohexamide. Carry out a plasma blank determination using an equivalent dilution in 0.01 M sodium hydroxide solution of the drug-free plasma sample. Plasma extraction. Fluorimetric procedure. Calculation. Assay of acetohexamide in tablets Accurately weigh an amount of the powder equivalent to 500 mg of acetohexamide into a 100-ml calibrated flask, add 60 ml of 0.1 M sodium hydroxide solution and shake for 30 min. Make up to volume with 0.1 M sodium hydroxide solution, mix and filter, and discard the first 20 ml of filtrate. Dilute 5.0 ml of filtrate to 50.0 ml with water.Transfer 1.0 ml of this solution into a 100-ml calibrated flask, add 0.01 M sodium hydroxide solution to volume and mix. To 1.0 ml of the final dilute solution in a 100-ml calibrated flask, add 1.0 ml of 1-methylnicotinamide reagent followed immediately by 1.0 ml of 5 M sodium hydroxide solution and mix thoroughly. Exactly 2 min after addition of the 5 M sodium hydroxide solution, make up to volume with 50% V/V formic acid. Measure the fluorescence intensity (maximum value) after 1.5-2 h. Determine the concentration of acetohexamide in the final dilute solution by reference to a calibration graph, obtained by carrying out the fluorimetric procedure on 1 .O-ml aliquots of standard solutions of acetohexamide in 0.01 M sodium hydroxide solution containing 0, 2.0, 4.0, 6.0, 8.0 and 10.0 pg ml-l of drug.Calculate the amount of aceto- hexamide in milligrams per tablet using the expression Sample preparation. Weigh and powder 20 tablets. Fluorimetric psocedure. Calculatioiz. myytl x c, x 100 Mass per tabletlmg = m2120 GIRGIS-TAKLA AND CHRONEOS : FLUORIMETRIC DETERMINATION OF Analyst, VoZ. 104 where C, pg ml-1 is the concentration of acetohexamide in the final sample solution and m, and m2 g are the average mass of the tablets and mass of sample taken, respectively. Results and Discussion Factors Affecting the Fluorimetric Procedure The significance of the reagent concentrations and reaction times selected for the recom- mended method was shown by carrying out the fluorimetric procedure described for the assay of acetohexamide in tablets, using a 10 pg ml-l solution of acetohexamide, and varying the concentration of the different reagents individually. Changing the concentration of 1-methylnicotinamide iodide in the reagent solution over a range from 0.5 to 10% m/V showed that a concentration of at least 2-3% was necessary for the highest fluorescence intensity to be obtained [Fig.2(a)], and that when reagent concentrations were reduced to 1 or o.5y0, fluorescence readings decreased to 78 or 52% of the maximum, respectively. When using a 3% reagent concentration, maximum fluorescence develops within 1.5 h and remains stable, decreasing by not more than about 2% of its value over 24 h. When the reagent concentration is increased to 5 or loyo, maximum fluorescence is measured after 1 h, after which time readings decrease gradually because of quenching caused by a yellow colour that slowly develops in the solution after acidification.This quenching effect can be avoided by extracting the aqueous solution with dichloromethane immediately after acidifi- cation and measuring the fluorescence in the organic layer at the usual wavelength settings. I I I I I 2 4 6 8 1 0 : I g 200 E 100 - LL 0 1 2 3 4 5 Concentration of 'I-methylnicotinamide, % m/V Time/min 600 =1 500 2 E cn *.' .- C + 400 m >: z 300 + 0, C + .- . 8 200 2 C g 100 - 0 2 4 6 8 1 0 Sodium hydroxide concentration/N Fig. 2. Influence on fluorescence intensity developed from 10 pg of acetohexamide of : ( a ) , 1-methylnicotinamide iodide reagent concentration ; ( b ) , duration of reaction with 2, 3 or 5 N sodium hydroxide solution; and (c), normality of sodium hydroxide soh tion.February, 1979 ACETOHEXAMIDE IN PLASMA AND TABLETS USING 1-METHYLNICOTINAMIDE 121 By following this procedure it was shown that there is no significant change in fluorescence intensity when the l-methylnicotinamide concentration is increased from 3 to 5 or 10%.Substituting l-methylnicotinamide chloride for the iodide does not result in any enhancement of fluorescence, and the iodide was therefore used throughout the work as it is easier to prepare. The fluorescence intensity is also influenced by the duration of the reaction in alkaline solution, as well as by the concentration of alkali used. Fig. 2(b) shows that the fluorescence intensity reaches a maximum when the reaction in alkali is allowed to proceed for 2 min, but that it is reduced if a longer reaction time is used.This observation is the same irrespec- tive of whether the fluorescence is measured in aqueous solution, or after extraction into dichloromethane. The highest fluorescence readings are obtained when the sodium hydroxide reagent concentration is 5 M or above [Fig. 2(c)]. With the higher alkali concentrations (8 or 10 M), however, the fluorescence is slightly less stable because of the quenching effect already mentioned. The fluorescence intensity is about ten times higher in acidic than in alkaline solution , and the final fluorimetric measurement is therefore made after acidification with formic acid.This acid produced a higher fluorescence than hydrochloric acid, possibly because with the latter a heating step is necessary in order to develop the fluorescence. The optimum concentration of formic acid in the final solution is approximately 50% VlV. With concentrations higher than this a yellow colour again develops and the fluorescence is less stable; with lower concentrations fluorescence develops to the same extent but more slowly. Precision and Accuracy The calibration graph for the assay in plasma was linear, and fluorescence intensity measurements (arbitrary units), after subtraction of the blank value, ranged from 130 for 0.5 pg ml-l to 653 for 2.5 pg ml-l of acetohexamide in human plasma or in 0.01 M sodium hydroxide solution. In order to test the precision of the procedure, eight replicate deter- minations were made on each of two plasma solutions containing 1.0 and 2.0 pg ml-l of acetohexamide.The relative standard deviations of the assay were calculated to be 3.7 and 3.1y0, respectively. The accuracy of the procedure was tested by assaying a sample of plasma to which had been added 40.0 pg ml-l of acetohexamide and comparing the results obtained by using the fluorimetric method and the spectrophotometric procedure of Smith et aL6 The average recovery (see Table I) was 99.2% by the fluorimetric method, while by the spectrophotometric procedure it was 93.0% calculated by means of the Smith equation, or 90.0% calculated directly using an value of 538.9 at 247 nm, which had been deter- mined for the acetohexamide sample used.The lower recovery by the spectrophotometric procedure, which does not apply a correction for the losses in the extraction steps involved, is in accordance with recoveries of 91 and 94% reported by Smith et al. The spectrophoto- metric procedure is not sufficiently sensitive to allow a comparison of the two methods of assay at lower levels of drug concentration in plasma. Defining the detection limit as the amount giving twice the background (blank) fluorescence, the limit calculated for aceto- hexamide in plasma is 0.2 pg ml-l, The calibration graph was also linear for the assay of tablets, and fluorescence intensity measurements (arbitrary units), after subtraction of the blank value, ranged from 113 for 2.0 pg ml-l to 564 for 10.0 pg ml-l of acetohexamide in TABLE I COMPARISON OF THE FLUORIMETRIC AND SPECTROPHOTOMETRIC METHODS FOR THE ASSAY OF PLASMA CONTAINING 40 pg ml-1 OF ADDED ACETOHEXAMIDE Fluorimetric method Spectrophotometric methods Acetohexamide Recovery, Acetohexamide Recovery, found/pg ml-1 % found/pg ml-l Yo 39.6 99.0 37.3 93.3 39.7 99.3 37.6 94.0 39.7 99.3 36.8 92.0 36.1 90.0 37.6 94.0 Mean: 99.2 Mean: 93.0122 GIRGIS-TAKLA AND CHRONEOS : FLUORIMETRIC DETERMINATION OF Analyst, VoZ.104 0.01 M sodium hydroxide solution. In order to test the precision and accuracy of the fluorimetric procedure, results were compared with those obtained by the spectrophotometric method of the United States Pharmacopeia.l3 Five portions, each equivalent to about 500 mg of acetohexamide, were taken from the same powdered sample of tablets, accurately weighed, and made up to 100 ml in 0.1 M sodium hydroxide solution in the usual way.Duplicate aliquots of the filtered solution were assayed by each method. The results are shown in Table 11. TABLE I1 COMPARISON OF THE FLUORIMETRIC AND USP METHODS FOR THE ASSAY OF A COMMERCIAL SAMPLE OF ACETOHEXAMIDE TABLETS Average mass per tablet, 0.653 g; nominal content of acetohexamide, 500 mg. Fluonmetric method USP method r \ r Acetohexamide found Percentage of Acetohexamide found Percentage of Sample mass/g per tabletlmg stated amount per tabletlmg stated amount A A > 0.669 0 500 508 0.650 9 493 495 0.6444 509 503 0.648 1 505 611 0.666 5 614 506 Mean .. .. 504 Relative standard deviation .. . . 1.3% 100.0 101.6 98.6 99.0 101.8 100.6 101.0 102.2 102.8 101.2 500 509 503 512 605 504 504 503 501 502 100.0 101.8 100.6 102.4 101.0 100.8 100.8 100.6 100.2 100.4 100.9 f 1.1 504 100.9 f 0.5 0.7% Specificity The procedure can be used for the determination of acetohexamide in the presence of insulin and other commercially available oral hypoglycaemic drugs, as none of these contains a ketone grouping.The fluorimetric reaction is also not shown by lactic or fl-hydroxy- butyric acids, or by hydroxyhexamide, which is the main route of metabolism for aceto- hexamide in man.21 Acetone and acetoacetic, pyruvic and a-ketoglutaric acids produce fluorescent derivatives in the reaction with 1-methylnicotinamide. When, however, solutions of these compounds, in the concentrations shown in Table 111, were tested by the assay procedure for aceto- hexamide in plasma, only acetoacetic acid was extracted with chloroform and remained behind after evaporation of the solvent in an amount sufficient to make it liable to interfere TABLE I11 FLUORESCENCE MEASUREMENTS MADE BY SUBJECTING SOLUTIONS OF POSSiBLE INTERFERING COMPOUNDS TO THE ASSAY PROCEDURE FOR ACETOHEXAMIDE I N PLASMA Compound Acetoacetic acid .... Acetone . . .. .. .. Hydroxyhexamide . . . . Lactic acid , . .. .. fl-Hydroxybutyric acid .. a-Ketoglutaric acid . . .. Pyruvic acid . . .. .. Concentration in solution/ mg ml-l 5.2 8 6 0.1 0.1 0.36 0.01 Fluorescence intensity, * arbitrary units 474 0 0 0 0 0 10 * Blank fluorescence subtracted.Febt“ZdUYy, 1979 ACETOHEXAMIDE I N PLASMA AND TABLETS USING 1-METHYLNICOTINAMIDE 123 with the assay.The acetoacetic acid can be separated from the acetohexamide by washing an aliquot of the chloroform extract twice, each time with an equal volume of distilled water. Repeated assays using this washing technique showed that, on average, 93% of the chloroform-extracted acetohexamide will remain in the chloroform layer, the remainder passing into the aqueous washings. Experience has shown that interference, if any, from acetoacetic acid in the assay procedure for plasma is likely to be slight. Plasma samples were taken from ten diabetic patients who were not receiving any acetohexamide, but were being controlled either by diet alone, or by treatment with insulin, chlorpropamide or metformin. The average plasma blank reading (arbitrary units) was 27 (range 23-30) by the normal assay procedure, 23 (range 20-25) when the chloroform was washed twice with water and 18 (range 16-19) when aliquots of the chloroform extracts were washed once with an equal volume of 0.5 M sodium carbonate solution before evaporation.The mean blank fluorescence (reagent blank) when the procedure was carried out using 0.01 M sodium hydroxide solution in place of plasma was also 18 (range 17-19). Blanks obtained using plasma samples from non-diabetic patients were in the same ranges. Binding of Acetohexamide to Plasma Attempts to separate acetohexamide from plasma proteins by ultrafiltration showed that the drug is largely bound by these proteins. Solutions of the drug in human plasma con- taining 10 and 100 pg ml-l of acetohexamide were assayed by the fluorimetric procedure.Samples (about 5ml) from each solution were also subjected to ultrafiltration by centri- fuging for about 30 min in Centriflo membrane cones. The ultrafiltrate collected from each solution was then subjected to fluorimetric assay. It was found that the binding of aceto- hexamide was 96.6y0 for the 10 pg ml-l and 95.7% for the 100 pg ml-l solution. These figures support the measurements of binding of the same drug to various human proteins made by JudisZ2 by means of equilibrium dialysis, which showed that acetohexamide can be bound S8yo to albumin, and to a lesser extent to fibrinogen I and a-globulin IV-4. We are grateful to Dr. T. M. Hayes of the Welsh National School of Medicine, Cardiff, and to his colleagues who kindly helped us to obtain plasma samples from diabetic subjects.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 2 2 . References Girgis-Takla, P., and Chroneos, I., J. Phavm. Pharmac., 1977, 29, 640. Simmons, D. L., Legore, ,4. A., Picotte, P., Chbnier, &I., and Jasmin, K., Can. J . Pharm. Sci., 1977, Maha, G. E., Kirtley, W. R., Root, M. A., and Anderson, R. C., Diabetes, 1962, 11, 83. Spingler, H., Klin. Wschr., 1957, 35, 533. Galloway, J. A., McMahon, R. E., Culp, H. W., Marshall, F. J., and Young, E. C., Diabetes, 1967, Smith, D. L., Vecchio, T. J., and Forist, A. A., Metabolism, 1965, 14, 229. Kleber, J. W., Galloway, J . A., and Rodda, B. E., J . Pharm. Sci., 1977, 66, 635. Huff, J . W., and Perlzweig, W. A., J , Biol. Chew., 1947, 167, 157. Carpenter, K. J., and Kodicek, E., Biochem. J . , 1950, 46, 421. Clark, B. R., Halpern, R. M., and Smith, R. A., Analyt. Biochem., 1975, 68, 54. Beyer, W. F., Analyt. Chem., 1972, 44, 1312. Salim, E. F., and Hilty, W. W., J . Pharm. Sci., 1967, 56, 385. “United States Pharmacopeia,” Nineteenth Revision, United States Pharmacopeial Convention Baltazar, J., and Ferreira Braga, M. M., Revta Port. Farm., 1966, 16, 169. Agarwal, S. P., and Walash, M. I., Indian J . Pharm., 1972, 34, 109. Meier, G. N., Kuhn, 0. S., Pierart, F. O., and Cortes, S. J. S., Revta R. Acad. Cienc. Exact. Fis. Nut. .4mer, M. M., and Walash, M. I., Bull. Fac. PharPn. Cairo Univ., 1973, 12, 399. Tammilehto, S., Favnz. Aikak., 1973, 82, 39. Eli Lilly and Co., BY. Pat. 912789, 1962. Marshall, F. J., Sigal, M. V., Jr., Sullivan, H. R., Cesnik, C . , and Root, 34. A., J . Med. Chem., 1963, Welles, J . S., Root, M. A., and Anderson, R. C., Proc. SOC. Exp. Bid. Med., 1961, 107, 583. Judis, J . , J. Pharm. Sci., 1972, 61, 89. 12, 85. 16, 118. Inc., Rockville, Md., 1975, p. 15. Madr., 1971, 65, 653. 6, 60. Received July loth, 1978 Accepted August 15th, 1978