ANALYST, OCTOBER 1989, VOL. 114 1215 Determination of Underivatised Efaroxan and ldazoxan in Blood Plasma by Capillary Gas Chromatography With Mass-selective Detection John D. Nichols, Neil A. Hyde and Keith Sugden Pharmaceutical Division, Reckitt & Colman plc, Dansom Lane, Kingston-upon-Hull, Hull HU8 7DS, UK Sensitive and specific methods for the determination of efaroxan and idazoxan in blood plasma have been developed based on solvent extraction, chromatographic separation and quantification by selected-ion monitoring using a quadrupole mass-selective detector. The use of a short non-polar bonded-phase capillary gas chromatography (GC) column facilitated rapid separation of the compounds of interest from internal standards, metabolites and endogenous material. Of equal significance was the ability to chromatograph these basic compounds without prior derivatisation.The application of bonded-phase capillary GC coupled to selected-ion monitoring resulted in robust analytical procedures with sub-ng ml-1 sensitivity and high selectivity. Keywords: Efaroxan; idazoxan; plasma assay; capillary gas chromatography; mass-selective detection Efaroxan. 2-[(2-ethyl-2,3-dihydrobenzofuran-2-yl)]-2-imid- azoline hydrochloride, and idazoxan, 2-[ (1,4-benzodioxan-2- yl)]-2-imidazoline hydrochloride, are potent and highly selec- tive a2-adrenoreceptor antagonists currently under investiga- tion for a range of therapeutic indications.1.2 Initial animal studies and early Phase 1 studies with idazoxan in man utilised a high-performance liquid chromatographic (HPLC) assay to determine plasma concentrations of the drug.3 The minimum levels that could be quantified with this technique were in the range 5-10 ng ml-1 in plasma.Subsequent single-dose studies to evaluate pharmacokinetic parameters required a specific assay procedure with an order of magnitude improvement in assay sensitivity to characterise fully the absorption and terminal elimination phases of the drug profile. Efaroxan was recognised early in its development as being more potent than idazoxan and this necessitated that a highly sensitive assay be developed before Phase 1 human studies could commence. This prompted the investigation of potentially more sensitive assay procedures which in turn led to the development of the methods described here.Experimental Apparatus Chromatographic separation was performed using a Hewlett- Packard 5890A gas chromatograph fitted with a 15m X 0.25 mm i.d., 0.25-pm film thickness DB1 fused-silica capil- lary column (J&W Scientific, Folsom, CA, USA). Sample introduction was via a Hewlett-Packard 7673A autosampler with fast injection into an inert injection port liner containing a glass-wool plug. A Hewlett-Packard 5970 mass-selective detector (MSD) was used for peak detection. Data handling, including integration and calculation of peak-area ratios, was performed on a Hewlett-Packard 59960 MS Chemstation. Reagents Acetate buffer (0.2 M, p H 5.0). Prepared by mixing 0.2 M sodium acetate solution and 0.2 M acetic acid. Carbonate - hydrogen carbonate buffer (0.4 M, p H 10.5).Prepared by mixing 0.4 M sodium carbonate and 0.4 M sodium hydrogen carbonate solutions. Diethyl ether. Glass-distilled grade (Rathburn Chemicals, Walkerburn, UK). Chloroform. Laboratory-reagent grade , special for chroma- tography (BDH, Poole, Dorset, UK). Standard Solutions for Efaroxan Assay Efaroxan solution. Prepared by dissolving the drug in distilled water and diluting the solution to give 0.1 and 0.01 pg ml-1 solutions. Internal standard solution, RX83 1001 { 2-[ (2,3-dihydro-2- propylbenzofuran-2-yl)]-2-irnidazoline hydrochloride} . Pre- pared by dissolving the salt in distilled water and diluting the solution to give a 0.1 pg ml-1 solution. Standard Solutions for Idazoxan Assay ldazoxan solution. Prepared by dissolving the drug in distilled water and diluting the solution to give 1.0, 0.1 and 0.01 pg ml-1 solutions.lnternal standard solution. Prepared by dissolving efaroxan in distilled water and diluting the solution to give a 50 ng ml-1 solution. Plasma Standards for Calibration Control plasma (1 ml) and an appropriate volume of the internal standard solution (SO pl) were placed in a round- bottomed 15-ml test-tube. For the efaroxan assay 25, 50, 100 or 200 pl of 0.01 pg ml-1 or 50, 100 or 250 pl of 0.1 pg ml-1 efaroxan solutions were added (0.25,0.5,1,2,5,10 or 25 ng of efaroxan, respectively). For the idazoxan assay 50, 100 or 250 pl of 0.01 pg ml-1 or 50, 100 or 250 p1 of 0.1 pg ml-1 or 50 pl of 1.0 pg ml-1 idazoxan solutions were added (0.5,1,2.5, 5 , 10, 25 or 50 ng of idazoxan, respectively).Test Samples Sample plasma (1 ml) and internal standard solution (50 pl containing 5 ng of RX831001 or 2.5 ng of efaroxan, respec- tively) were placed in a round-bottomed 15-ml test-tube. Extraction Procedure Acetate buffer (1 ml) was added to each sample or standard test-tube. The tube was stoppered and vortex-mixed for 15 s. Diethyl ether ( 5 ml) was added followed by further mixing for 1 min and centrifugation for 5 min to separate the phases. The tubes were frozen in a cardice - acetone bath and the diethyl ether was decanted off and discarded. The tube contents were thawed and mixed with l m l of carbonate - hydrogen carbonate buffer (15 s) prior to further addition of diethyl1216 cn w .- 5 40000 2 P ._ I) & 30000 6 C m -0 3 c 20000 z ANALYST.OCTOBER 1989, VOL. 114 - - - - 80 -100 :::: 50 100 rnlz 150 187 h -1 200 Fig. 1. line) and RX831001 peak (inverted below zero line) Comparison of mass spectra from efaroxan peak (above zero 12 000 ._ E 10000 c 3 > 5 8000 c ._ e 6000 0 C m 4000 3 a 2000 I I I I I I 5.5 6.0 6.5 7.0 7.5 8.0 Time/min Fig. 2. Chromatogram of the ion at m/z 187 for a 2-p.1 injection of the test sample. Efaroxan elutes at 6.1 min and the internal standard at 6.5 min. The sample was calculated to contain 2.4 ng ml-1 of efaroxan 80 70 a; c m -0 C 3 I) Q 50 100 150 200 mlz Fig. 3. Mass spectrum of idazoxan peak 12 000 v) c .- 5 10000 2 F 2 8000 a; 2 6000 .I- m m -0 S 2 4000 Q 2ooo t 0- 6.0 6.5 7.0 Tim eim i n 50 000 i" L 10 000 0- 6.0 6.5 7.0 Time/m i n Fig. 4. ( a ) Chromatogram of the ion at m/z 174; and ( h ) chromato- gram of the ion at m/z 187, for a 2 4 injection of a test sample found to contain I .1 ng ml-1 o f idazoxan. ldazoxan elutes at 6.5 min and the internal 3tandard at 6.2 min ether ( 5 ml). The tubes were vortexed (1 min), centrifuged and frozen as before, and the diethyl ether was decanted off into a 12-ml conical test-tube. The aqueous layer was re-extracted with a further 5-ml aliquot of diethyl ether and the extracts were combined and allowed to reach ambient temperature. Any aqueous phase carried over was removed by freezing the tip of the tube and decanting the diethyl ether off into a further clean conical tube. The diethyl ether was then evaporated under a stream of nitrogen and the residue reconstituted in chloroform (50 PI), transferred into a low- volume insert in an autosampler bottle, capped and taken for analysis.Chromatography Injections were made while the injection port was switched to the splitless mode, the split valve being opened after 3 min. The gas chromatograph was temperature programmed as follows: inlet pressure, 50 kPa helium; injection port temper- ature, 250 "C (constant) (efaroxan method), 200 "C (constant) (idazoxan method); column oven temperature, 45 "C for 3 min, 45-180 "C at 70 "C min- , 180 "C for 5 min, 180-280 "C at 70 "C min-1, 280 "C to end of run (16 min in total); and transfer line temperature, 275 "C (constant). Data Acquisition The gas chromatography (GC) column was interfaced directly to the MSD which was run in the selected-ion monitoring (SIM) mode.Acquisition parameters for the efaroxan assay were: ion, mlz 187; dwell time, 200 ms; and acquisition cycles per second, 3.7. Acquisition parameters for the idazoxan assay were: ion, mlz 187 and 174; dwell time, 25 and 250 ms, respectively; and acquisition cycles per second, 2.8. The MSD was switched on only during the expected elution of the peaks of interest, i.e. , 5-8 min. For the efaroxan procedure the raw data were processed by the Chemstation and the resulting chromatogram was inte- grated. The efaroxan peak area was divided by the internal standard peak area and the peak-area ratio reported together with the chromatogram and sample details. For the idazoxan procedure the raw data were processed to give two extracted ion chromatograms due to the ions at mlz 187 and 174, respectively.The chromatograms were inte- grated and the peak-height ratio was calculated by dividing the idazoxan peak height in the m/z 174 chromatogram by the internal standard peak height in the mlz 187 chromatogram.ANALYST. OCTOBER 1989, VOL. 114 1217 Table 1. Calibration data for the efaroxan procedurc Efaroxan concentration/ ng ml- 1 0.25 0.50 1 .o 2.0 5.0 10.0 25 .o Peak-area ratio (cfaroxan to internal standard) I 2 3 4 5 0.0745 0.0669 0.0741 0.05 14 0.O6O7 0.0928 0.0959 O.IO96 0.1007 0.1048 0.1856 0 . I745 0.1814 0.1748 0.1716 0.3143 0.3001 0.3400 0.3 142 0.3967 0.9165 0.8408 0.8173 0.9125 0.8876 1.955 1.936 1.794 I .875 2.163 4.745 4.685 4.505 4.263 4.299 Table 2. Calibration data for the idazoxan proccdurc Idazoxan concentration/ Peak-height ratio (idazoxan to internal standard) ng ml-1 0.25 I .o 2.5 5.0 10.0 25 .o 50.0 1 0.0692 0.1266 0.2583 0.5 102 0.9750 2.161 3.095 2 3 0.091 1 0.0765 0.1078 0. I36 1 0.2706 0.2708 0.43 19 0.5289 0.850 1 1.01 1 1.909 2.221 3.980 4.397 The peak-height ratio was reported together with the two chromatograms of interest and the sample details. Results The feasibility of the above methods was initially investigated using 1-ul injections of a 0.1 mg tnl-1 chloroform solution of efaroxan or idazoxan base. The GC conditions were optimised to those described above with additional temperature ramping being included to avoid possible late-running peak contamina- tion from plasma extracts. The MSD was operated in the scan mode, giving full mass spectral information for each eluting peak.The spectra of the peaks of interest were studied in order to determine which was the best ion or ions to use in the analysis. Operation in the SIM mode was necessary to achieve the desired assay sensitivity. In each instance several com- pounds were considered as possible internal standards. Analogues of efaroxan were available and their suitability as an internal standard was systematically tested. RX831001 was found to give good peak shape and was well separated from efaroxan. Comparison of the mass spectra of these two compounds (Fig. 1) showed a common base peak and as a result offered the opportunity to maximise the assay sensitivity by acquiring data for a single ion (mlz 187) only. A typical STM chromatogram is shown in Fig.2. The mass spectrum of idazoxan (Fig. 3) showed a base peak at mlz 174; however, despitc running many analogues of idazoxan, no compound could be found which had a common base peak. Efaroxan was eventually chosen as the internal standard owing to its availability and good chromatographic separation from idazoxan. Fig. 4(u) and (b) shows typical SIM chromatograms for idazoxan and the internal standard, respectively. The efficiency of the extraction procedure was determined for both efaroxan and idazoxan by spiking plasma with a radiolabelled drug of known purity and specific activity and determining the recovery by liquid scintillation counting of the extract. Replicate extracts ( n = 6) gave recoveries of 84.0% [coefficient of variation (CV), 3.4%] and 85.0% (CV, 2.6%) for efaroxan and idazoxan, respectively, at concentrations appropriate to the assay (10 ng ml-1).'The extraction proce- dure was the rate-limiting step in sample analysis; however, up to 30 samples per day could easily be analysed by one technician. 4 0.0715 0.1297 0.2500 0.574 1 0.981 1 1.834 5.092 5 0.0908 0.1347 0.2563 0.5895 1.084 2.489 4.900 Mean k SD 0.0655 k 0.0097 0.1008 k 0.0067 0.1776 * 0.0057 0.3331 rf: 0.0384 0.8749 k 0.0441 1.945 2 0.137 4.499 k 0.2 I8 Mean k SD 0.07% k 0.0105 0. I270 * 0.01 14 0.2612 t 0.0092 0.5269 i 0.0622 0.9802 k 0.0847 2.123 k 0.262 4.293 k 0.799 cv, % 14.8 6.6 3.2 11.5 5.0 7.0 4.8 cv, Yo 13.2 9.0 3.5 11.8 8.6 12.3 18.6 l o I 0 .- +- 2 1 2 ? m Y m a, a 0.1 1 I I 0.1 1 10 Concentratiodng ml-1 Fig. 5. Calibration graph for the determination of efaroxan in blood plasma.Calculated linc of best fit (solid line) and 95% confidence limits (broken lines) 0 .- +- p 1 f +- L G7 a) Y m a .- 0.1 1 I I 0.1 1 10 Concentratiodng ml-1 Fig. 6. Calibration graph for the determination of idazoxan in blood plasma. Calculated linc of best fit (solid linc) and 95% confidence limits (broken lines)1218 ANALYST, OCTOBER 1989, VOL. 114 For both assay procedures multiple calibration points (five) were obtained for each of seven standard concentrations (Tables 1 and 2). The relationship between concentration and peak-area ratio was defined using least-squares regression of the log - log transformed data. Rartlett’s test4 was used to check the variance stability over the calibrated range and the goodness-of-fit was tested using the F-value for each poly- nomial following an analysis of variance.The line of best fit for efaroxan was described by four regression coefficients. This is shown, together with the calculated 95% confidence limits, in Fig. 5. The multiple correlation coefficient of tl e line was 0.998. Assay precision (CV, Yo) was calculated as 11% at 20 ng ml-1 and 13% at 0.5 ng ml-1. The idazoxan assay gave a linear calibration graph; this is shown together with the 95% confidence limits in Fig. 6. The multiple correlation coefficient of the line was 0.996 and the precision (CV, YO) was calculated to be 15.5% over the calibrated range. Test samples were run concomitantly with quality control (QC) standards to ensure the accuracy of the methods.For the test data to be valid, the QC results had to fall within the defined 95% tolerances. Concentrations in the test samples were computed directly from the calibration graphs. Discussion The availability of bonded-phase fused-silica capillary column technology has resulted in a renaissance in GC in many areas where HPLC had become the separation method of choice. It is now possible to chromatograph many compounds by GC without employing time-consuming and laborious derivatisa- tion procedures. Efaroxan and idazoxan are two such com- pounds, their basic nature and chemical structure offering few opportunities for derivatisation. Gas chromatography of the native compounds on packed columns proved very difficult, with degradation problems and poor peak shapes.However, using a short non-polar bonded-phase capillary system, good, consistent peak shapes were obtainable with the additional benefit of improved chromatographic resolution. The most specific detection technique for GC is mass spectrometry. Specificity and sensitivity were major consider- ations in the development of a bioanalytical procedure, and when used in the STM mode mass-selective detection offered the potential for both high sensitivity and specificity. By combining the unique properties of bonded-phase capillary GC with the convenience and power of mass-selective detection in the SIM mode, robust, sensitive and specific analytical procedures for the determination of efaroxan and idazoxan in blood plasma samples were successfully devel- oped. Automation of the chromatographic procedures to facilitate the analysis of large numbers of samples over an extended period was also a key objective achieved in method development. To date, over 3000 samples have been analysed using these procedures, with a single column handling approximately 700 samples over a 4-month period. References 1. 2. Chapleo, C. B., Doxey, J. C., Myers, P. L., and Roach, A. G., Br. J . Pharmacol., 1981, 74. 842. Chapleo, C. B.. Myers. P. L., Butler, R. C. M., Davis, J. A., Doxey, J . C., Higgins, S. D., Myers, M., Roach, A. G., Smith, C. F. C., Stillings, M. R., and Welbourn, A. P.,J. Med. Chem., 1984, 27, 570. Muir, N . C., Lloyd-Jones, J . G., Nichols, J. D., and Clifford. J . M., Eur. J . Clin. Pharmacol., 1986, 29, 743. Snedecor, G. W., and Cochrane. W. G., “Statistical Methods,” Sixth Edition, Iowa State University Press, Amcs, 1967. 3 . 4. Paper 9100048H Received January 4th, 1989 Accepted May 23rd, 1989