- Email: [email protected]
Determination of Nitrosourea Compounds in Brain Tissue by Gas Chromatography and Electron Capture Detection
Determination of Nitrosourea Compounds in Brain Tissue by Gas Chromatography and Electron Capture Detection SAMUEL
J. HASSENBUSCH~~, 0.MICHAELCOLVINS, AND JAMESH. ANDERSON§
ReceivedJanuary 4, 1995, from the Wniversiw of Texas M. D. Anderson Cancer Center, Depatfment of Neurosurgery, C9075, 1515 Holcombe Boulevard Houston, TX 77030. Accepted for publication March 8, 1995@. Present addresses: *Pharmacology Oncology, Oncology 1-121, Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287, and §Radiology Research Center, Traylor 330, Johns Hopkins University School of Medicine, 600 North Wolfe Street, Baltimore, MD 21287. BCNU
Abstract 0 A relatively simple, high-sensitivitygas chromatographic assay is described for nitrosourea compounds, such as BCNU [1,3-bis(2chloroethy1)-1-nitrosourea] and MeCCNU [1-(2-chloroethyl)-3-(trans-C methylcyclohexyl)-1-nitrosoureal, in small biopsy samples of brain and other tissues. After extraction with ethyl acetate, secondary amines in BCNU and MeCCNU are derivatized with trifluoroacetic anhydride. Compounds are separated and quantitated by gas chromatography using a capillary column with temperature programming and an electron capture detector. Standard curves of BCNU indicate a coefficient of variance of 0.066 f. 0.018, a correlation coefficient of 0.929, and an extraction efficiency from whole brain of 68% with a minimum detectable amount of 20 ng in 5-10 mg samples. The assay has been facile and sensitive in over 1000 brain biopsy specimens after intravenous and intraarterial infusions of BCNU.
Introduction
n
Nitrosourea compounds, such as BCNU [1,3-bis(Z-chloroethyl)-1-nitrosourea]and MeCCNU [ 1-(2-~hloroethyl)-3-(trum4-methylcyclohexyl)-l-nitrosourea],remain the backbone of many chemotherapy regimens, especially for brain tumors. With disappointing clinical results, however, there is an increasing use of these agents in higher doses and a need for measurement of actual concentrations of the agents not only in blood but in small biopsy samples of target tissues, such as the brain. Previous assay methods for nitrosourea antitumor agents have used less-sensitive colorimetric more complicated chemical ionization mass spectrometry3 or gas chromatographic-mass spectrometry4-8 methods, or highpressure liquid chromatography (HPLC)g-ll or gas chromatography (GC) rnethodsl2 with low sensitivity for analysis of nitrosourea compounds in small tissue samples. This report presents a relatively simple GC method with high sensitivity suitable for drug determination in small biopsy samples of brain and other tissues.
Methods Tissue Extraction and Derivatization Method-The derivatization method was adapted from a previously reported reaction using denitrosation of the BCNU with subsequent acetylation of the secondary nitrogen atoms with halogen groups.13 Both BCNU and MeCCNU were studied in this manner. Rabbit brain biopsies (10-20 mg) were placed in microcentrifuge tubes and freeze-thawed for three cycles of 30 s each. Ethyl acetate (EtOAc) (500 pL)containing MeCCNU (512 ng) was added to each tube, and the tubes were shaken and centrifuged (5000 rpm, Eppendorf microcentrifuge model 5415C, Brinkmann Instruments Inc., Westbury, NY)in a cold room. EtOAc was decanted from tubes into reaction vials, and the samples were dried (15 min, 15 "C) under a @
Abstract published in Advance ACS Abstracts, June 1 , 1995.
840 / Journal of Pharmaceutical Sciences Vol. 84, No. 7, July 1995
F.C-F FC.F F F
+ SOH
Figure 1-Derivatization reaction scheme for BCNU with trifluoroacetic anhydride. Trifluoroacetylation of secondary amines allows analysis by an electron capture detector during gas chromatography. nitrogen stream. Trifluoroacetic anhydride (TFAA, 40 pL) and acetonitrile (30 p L ) were added to each vial (Figure 1). Each sample was heated a t 85 "C (6 h), dried under a nitrogen stream, reconstituted with EtOAc (100 pL), and injected into the GC (1p L ) . The derivatizing agents (trifluoroacetic anhydride, pentafluoropropionic anhydride, and heptafluorobutyric anhydride), reaction temperature, and reaction time were varied to provide the most complete derivatization, as evidenced by GC peak heights, of BCNU or MeCCNU. For standard curves from BCNU powder, BCNU powder (10-600 ng in 2-fold increments) was mixed with EtOAc (500 p L ) containing the internal standard MeCCNU (512 ng) in reaction vials. The samples were then dried under a nitrogen stream, acetonitrile and TFAA were added, and the samples were derivatized and assayed as described above. For BCNU standard curves from brain samples, untreated rabbit brain biopsies (10-20 mg) were placed in microcentrifuge tubes. BCNU (10-600 ng in 2-fold increments) in 10 pL of 1%ethanol was added to each tube. BCNU was extracted and assayed as described above. GC Configuration-Because of the derivatization method, which utilized halogen group attachment to the BCNU molecule, a n electron capture detector (ECD) with 63Niradioactive foil was used on a gas chromatograph with an autosampler (Model 3500 gas chromatograph, Varian Associates, Sunnyvale, CAI. The GC was configured as follows: SPBl 60 m capillary column, 0.25 mm i.d. fused silica, 0.25 pm stationary phase coating (catalog 2-4030, Supelco Incorporated,
0022-3549/953184-0840$09.00/0
0 1995, American Chemical Society and American Pharmaceutical Association
JBCNU J MeCCNU
.3
.1
L L
L(
Discussion
J 1
0
10
,
/
,
/
30 40 50 Retention Time (minutes) 20
,
1
60
1
height and peak area calculations showed a linear relationship with a correlation coefficient of 0.956. Use of this assay over a 2 year period revealed no statistically significant difference between powder and brain samples in the intraday or interday variation of standard curve. The intraday variation in the standard curve was 1.2%& 0.4%. The interday variation of the standard curve over a 2 week period and approximately 150 samples, was 4.2%f 1.3%with a trend toward lower peak heights and peak areas for each amount of BCNU in the standard curve.
/
70
Figure 2-Representative gas chromatogram with a capillary column. BCNU is assayed with MeCCNU as an internal standard. Bellefonte, PA)with a 1 m capillary precolumn, an ECD with 63Ni radioactive foil, a 1pL injection with a 3:l split, and a capillary column pressure of 30 psi. The column temperatures were 50 "C for 5 min, then from 50 to 100 "C at 10 "C/min and from 100 to 105.5 "C at 0.1 "C/min. The injector temperature was 200 "C. The chromatogram was collected as analog data by a strip chart recorder and as digital data by a computer. BCNU:MeCCNU peak height and peak area ratios were calculated from the analog record and the digitized data on computer record. BCNU:MeCCNU peak height ratios were plotted versus the amount of BCNU in the samples. A similar graph of BCNU:MeCCNU peak area ratios versus the amount of BCNU in the samples was also done. Columns of different polarities (nonpolar, intermediate, polar) and size (conventional bore and capillary column) as well as different oven temperature schemes (isothermal and temperature programming) were studied to maximize sensitivity and ability to separate specimen peaks.
Results BCNU Standard Curve from Powder-BCNU usually had a retention time of 30-31 min and MeCCNU had a retention time of 55-57 min. A typical GC chromatogram for this assay is shown in Figure 2. The derivatized BCNU peak comes before the derivatized MeCCNU peak. For the peak height standard curve, statistical evaluation by the method of least square analysis indicated a correlation coefficient of 0.967 and a coefficient of variance (standard error of meadmean) for each amount level of BCNU of 2.5%f 1.8%. BCNU Standard Curve with Brain Samples-The chromatograms for BCNU extracted from brain samples were not significantly different from the BCNU powder studies in terms of unwanted peaks or baseline noise, The minimum detectable amount of BCNU in each sample was 20 ng. The overall coefficient of variance (determined at each BCNU amount level assayed) was 6.6% f 1.8%. Statistical evaluation by least square analysis of this standard curve showed a slope of 0.00243 with a y-intercept of 0.080 and a coefficient of correlation of 0.929. The extraction efficiency from whole brain samples (68%)was determined by comparison of the least square analysis equations from this curve and that of the curve from BCNU powder above. For both standard curves, computer-derived peak area ratios showed similar coefficients of variance and correlation coefficients by least square analysis. Comparisons of peak
Previously Described Nitrosourea Assays-A simple and inexpensive colorimetric assay for BCNU has been reported1-2based upon the sulfanilamide assay developed by Bratton and Mar~ha1l.l~ The lower limit of sensitivity is 0.5 pg (500 ng) of BCNU in a sample. A major advantage of the colorimetric assay is that only the intact BCNU molecule and no decomposition products are quantitated by this assay. This method, however, has only been reported for use with plasma and not with whole blood or tissue samples. Previous HPLC methods assay the nonderivatized, intact parent molecule and are reported to detect 50 ng of BCNU in a ~ample.~-llThe assay has been reported with 500-2000 pL blood samples but not with 5-10 mg tissue samples with a volume of approximately 6-12 pL. It is possible that, as in this assay, 5-10 mg tissue samples with a volume of approximately 6-12 pL, could be homogenized into a larger volume with subsequent extraction. This might maintain the lower limit of sensitivity with HPLC techniques. A possibly lower extraction efficiency with brain tissue samples as compared to plasma, however, might raise the lower limit of sensitivity with such methods. Considering the complexity of this gas chromatographic-electron detection (GC-EC) method as well as these factors, it might be concluded that the GC-EC method might represent a roughly equivalent alternative to reported HPLC methods, thus allowing the choice of a specific method to be based upon available equipment and expertise. Various other methods have been reported using GC to separate derivatized samples with mass spectrometry to assay the amount of BCNU.4-sJ3 The lower limit of sensitivity is 1 ng of BCNU in a sample but requires the use of a mass spectrometer for each sample. Another GC assay, not using mass spectrometry as a detector, has also been reported.12In this assay, BCNU is extracted from plasma by benzene. The underivatized BCNU residue is injected onto a capillary, 25 m, 0.33 mm i.d., nonpolar column. A thermionic nitrogenphosphorus-specific ( " N P ) detector is used. The lower limit of sensitivity has been reported to be 2-3 ng of total BCNU from 500 pL plasma samples. The assay has not been studied on tissue (e.g., brain) samples. The sensitivity of these previous assays makes it difficult to assay BCNU in small biopsies of brain or other tissues. Gas chromatographic-mass spectrometric assays for BCNU have high sensitivity for the detection of very small amounts of BCNU. Although the use of mass spectrometry is increasing, there is still limited availability especially when analysis of a large number of samples from a variety of animals is necessary. Radiolabeled BCNU also has been used to assay for BCNU tissue concentrations in animal studies. BCNU radiolabels, whether 3H or 14C,usually either involve the carbamoylating part of the molecule or are nonspecific and involve both ends of the molecule. For BCNU, the alkylating part of the molecule, however, is the active moiety and the carbamoylating part is the nonactive moiety. Journal of Pharmaceutical Sciences / 841 Vol. 84, No. 7, July 1995
BCNU Gas Chromatographic GC Assay-This GC assay is practical yet provides sensitivity for BCNU drug concentrations in brain tissue samples with a mass of 5-10 mg. For example, this provides the ability to quantitate BCNU concentrations in biopsies at multiple sites in and around tumorbearing tissue. With the widespread ability to perform stereotactic brain biopsies in humans, this allows study of important parameters that might influence distribution of BCNU and ultimately its potential therapeutic efficacy. As compared to ethyl ether, ethyl acetate minimized extraction of unwanted compounds from brain samples. An electron capture detector was chosen for this assay because of its high sensitivity to halogen groups. TFAA was chosen as a derivatizing agent because heptafluorobutyric anhydride and pentafluoropropionicanhydride appeared to react less completely with BCNU but more extensively with other unwanted compounds in the brain extractions. In a technical sense, this assay derivatizes denitrosated BCNU. In tissue, after denitrosation of BCNU, the resultant molecule decomposes into the alkylating and carbamoylating moieties in a very short time period, probably less than 1 s, and the decomposition products travel only a very small distance, probably only microns. For this reason, this assay, in a practical sense, can be used to quantitate intact BCNU. The separation of BCNLJ, MeCCNU, and other unwanted peaks was accomplished with the use of temperature programming and a nonpolar capillary column of small diameter (0.25 mm) and long length (60 m). Daily Use of the Assay-This GC assay has been applied to real samples for over 2 years and in over 1000 brain biopsy specimens from over 60 rabbits. The method was very useful in the determination of BCNU concentrations in different areas of the brain in this study of intravenous and intraarterial infusions via the common carotid artery. The assay was invaluable in the determination of BCNU in the small punch biopsies from this study. The GC assay was found to be easy and sensitive and has been used to determine BCNU concentrations in 2 mm diameter punch biopsies from 3 mm thick brain slices. Each biopsy had a mass of 5-10 mg. The concentration of BCNU in each sample was calculated from the amount of BCNU in each brain sample divided by the actual weight of each biopsy. The resultant concentrations document increased BCNU concentrations in the infused hemisphere during intraarterial infusion as compared to intravenous infusion (Figure 3). This GC-EC method provides a facile GC method for detection of BCNU in relatively small tissue samples without the need for mass spectrometry. It enables the important
842 /Journal of Pharmaceutical Sciences Vol. 84,No. 7, Ju& 1995
2
v cp
travenous Infusion)
BCNU - 300 mglsquare-meter over 15 minutes
1
nB “0
10 20 30 40 Time after Start of 15 minute infusion (minutes)
50
Figure 3-Example of use of assay to determine BCNU concentrations in small brain biopsies at different time points during and after intraarterial and intravenous infusions.
potential clinical assay of BCNU in actual human brain biopsies obtained by needle biopsy under computerized tomography guidance. Hopefully this might provide information that can be used to adjust clinical doses, either by intravenous or intraarterial infusion routes.
References and Notes 1. Loo, T. L.; Dion, R. L. J Pharm. Sci. 1965, 54, 809-810. 2. Kari, P.; McConnell, W. R.; Finkel, J. M.; Hill, D. L. Cancer Chemother. Pharmacol. 1980, 4, 243-248. 3. Weinkam, R. J.; Wen, J. H.; Furst, D. E.; Levin, V. A. Clin. Chem. 1978,24, 45-49. 4. Weinkam, R. J.; Lin, H. S. Anal. Chem. 1979, 51, 972-975. 5. Smith, R. G.; Cheung, L. K. J . Chromatogr. 1982, 229, 464469. 6. Smith, R. G.; Cheung, L. K.; Feun, L. G.; Loo, T. L. Biomed. Mass Spect. 1983,10,404-407. 7. Caddy, B.; Idowu, 0. R. Analyst 1982,107, 301-306. 8. Weinkam, R. J.; Liu, T. J. J. Pharm. Sci. 1982, 7 1 , 153-157. 9. Hochberg, F. H.; F’ruitt, A. A.; Davis, K.; DeBruyn, G. Neurology 1983, 33 [Suppl 21, 109. 10. Krull, I. S.; Straws, J.; Hochberg, F.; Zervas, N. T. J . Anal. Toxicol. 1981, 5, 42-46. 11. Yeager, R. L.; Oldfield, E. H.; Chatterji, D. C. J . Chromatogr. 1984,305, 496-501. 12. El-Yaziei. A,: Martin. C . R. Pharm. Res. 1988.5. 220-225. 13. Smith, k.’G.: Blackstock, S. C.; Cheung, L. K.;’ Loo, T. L. Anal. Chem. 1981,53, 1205-1208. 14. Bratton, A. C.; Marshall, E. K. J . Biol. Chem. 1939, 128, 537-550. 559500058
J. HASSENBUSCH~~, 0.MICHAELCOLVINS, AND JAMESH. ANDERSON§
ReceivedJanuary 4, 1995, from the Wniversiw of Texas M. D. Anderson Cancer Center, Depatfment of Neurosurgery, C9075, 1515 Holcombe Boulevard Houston, TX 77030. Accepted for publication March 8, 1995@. Present addresses: *Pharmacology Oncology, Oncology 1-121, Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287, and §Radiology Research Center, Traylor 330, Johns Hopkins University School of Medicine, 600 North Wolfe Street, Baltimore, MD 21287. BCNU
Abstract 0 A relatively simple, high-sensitivitygas chromatographic assay is described for nitrosourea compounds, such as BCNU [1,3-bis(2chloroethy1)-1-nitrosourea] and MeCCNU [1-(2-chloroethyl)-3-(trans-C methylcyclohexyl)-1-nitrosoureal, in small biopsy samples of brain and other tissues. After extraction with ethyl acetate, secondary amines in BCNU and MeCCNU are derivatized with trifluoroacetic anhydride. Compounds are separated and quantitated by gas chromatography using a capillary column with temperature programming and an electron capture detector. Standard curves of BCNU indicate a coefficient of variance of 0.066 f. 0.018, a correlation coefficient of 0.929, and an extraction efficiency from whole brain of 68% with a minimum detectable amount of 20 ng in 5-10 mg samples. The assay has been facile and sensitive in over 1000 brain biopsy specimens after intravenous and intraarterial infusions of BCNU.
Introduction
n
Nitrosourea compounds, such as BCNU [1,3-bis(Z-chloroethyl)-1-nitrosourea]and MeCCNU [ 1-(2-~hloroethyl)-3-(trum4-methylcyclohexyl)-l-nitrosourea],remain the backbone of many chemotherapy regimens, especially for brain tumors. With disappointing clinical results, however, there is an increasing use of these agents in higher doses and a need for measurement of actual concentrations of the agents not only in blood but in small biopsy samples of target tissues, such as the brain. Previous assay methods for nitrosourea antitumor agents have used less-sensitive colorimetric more complicated chemical ionization mass spectrometry3 or gas chromatographic-mass spectrometry4-8 methods, or highpressure liquid chromatography (HPLC)g-ll or gas chromatography (GC) rnethodsl2 with low sensitivity for analysis of nitrosourea compounds in small tissue samples. This report presents a relatively simple GC method with high sensitivity suitable for drug determination in small biopsy samples of brain and other tissues.
Methods Tissue Extraction and Derivatization Method-The derivatization method was adapted from a previously reported reaction using denitrosation of the BCNU with subsequent acetylation of the secondary nitrogen atoms with halogen groups.13 Both BCNU and MeCCNU were studied in this manner. Rabbit brain biopsies (10-20 mg) were placed in microcentrifuge tubes and freeze-thawed for three cycles of 30 s each. Ethyl acetate (EtOAc) (500 pL)containing MeCCNU (512 ng) was added to each tube, and the tubes were shaken and centrifuged (5000 rpm, Eppendorf microcentrifuge model 5415C, Brinkmann Instruments Inc., Westbury, NY)in a cold room. EtOAc was decanted from tubes into reaction vials, and the samples were dried (15 min, 15 "C) under a @
Abstract published in Advance ACS Abstracts, June 1 , 1995.
840 / Journal of Pharmaceutical Sciences Vol. 84, No. 7, July 1995
F.C-F FC.F F F
+ SOH
Figure 1-Derivatization reaction scheme for BCNU with trifluoroacetic anhydride. Trifluoroacetylation of secondary amines allows analysis by an electron capture detector during gas chromatography. nitrogen stream. Trifluoroacetic anhydride (TFAA, 40 pL) and acetonitrile (30 p L ) were added to each vial (Figure 1). Each sample was heated a t 85 "C (6 h), dried under a nitrogen stream, reconstituted with EtOAc (100 pL), and injected into the GC (1p L ) . The derivatizing agents (trifluoroacetic anhydride, pentafluoropropionic anhydride, and heptafluorobutyric anhydride), reaction temperature, and reaction time were varied to provide the most complete derivatization, as evidenced by GC peak heights, of BCNU or MeCCNU. For standard curves from BCNU powder, BCNU powder (10-600 ng in 2-fold increments) was mixed with EtOAc (500 p L ) containing the internal standard MeCCNU (512 ng) in reaction vials. The samples were then dried under a nitrogen stream, acetonitrile and TFAA were added, and the samples were derivatized and assayed as described above. For BCNU standard curves from brain samples, untreated rabbit brain biopsies (10-20 mg) were placed in microcentrifuge tubes. BCNU (10-600 ng in 2-fold increments) in 10 pL of 1%ethanol was added to each tube. BCNU was extracted and assayed as described above. GC Configuration-Because of the derivatization method, which utilized halogen group attachment to the BCNU molecule, a n electron capture detector (ECD) with 63Niradioactive foil was used on a gas chromatograph with an autosampler (Model 3500 gas chromatograph, Varian Associates, Sunnyvale, CAI. The GC was configured as follows: SPBl 60 m capillary column, 0.25 mm i.d. fused silica, 0.25 pm stationary phase coating (catalog 2-4030, Supelco Incorporated,
0022-3549/953184-0840$09.00/0
0 1995, American Chemical Society and American Pharmaceutical Association
JBCNU J MeCCNU
.3
.1
L L
L(
Discussion
J 1
0
10
,
/
,
/
30 40 50 Retention Time (minutes) 20
,
1
60
1
height and peak area calculations showed a linear relationship with a correlation coefficient of 0.956. Use of this assay over a 2 year period revealed no statistically significant difference between powder and brain samples in the intraday or interday variation of standard curve. The intraday variation in the standard curve was 1.2%& 0.4%. The interday variation of the standard curve over a 2 week period and approximately 150 samples, was 4.2%f 1.3%with a trend toward lower peak heights and peak areas for each amount of BCNU in the standard curve.
/
70
Figure 2-Representative gas chromatogram with a capillary column. BCNU is assayed with MeCCNU as an internal standard. Bellefonte, PA)with a 1 m capillary precolumn, an ECD with 63Ni radioactive foil, a 1pL injection with a 3:l split, and a capillary column pressure of 30 psi. The column temperatures were 50 "C for 5 min, then from 50 to 100 "C at 10 "C/min and from 100 to 105.5 "C at 0.1 "C/min. The injector temperature was 200 "C. The chromatogram was collected as analog data by a strip chart recorder and as digital data by a computer. BCNU:MeCCNU peak height and peak area ratios were calculated from the analog record and the digitized data on computer record. BCNU:MeCCNU peak height ratios were plotted versus the amount of BCNU in the samples. A similar graph of BCNU:MeCCNU peak area ratios versus the amount of BCNU in the samples was also done. Columns of different polarities (nonpolar, intermediate, polar) and size (conventional bore and capillary column) as well as different oven temperature schemes (isothermal and temperature programming) were studied to maximize sensitivity and ability to separate specimen peaks.
Results BCNU Standard Curve from Powder-BCNU usually had a retention time of 30-31 min and MeCCNU had a retention time of 55-57 min. A typical GC chromatogram for this assay is shown in Figure 2. The derivatized BCNU peak comes before the derivatized MeCCNU peak. For the peak height standard curve, statistical evaluation by the method of least square analysis indicated a correlation coefficient of 0.967 and a coefficient of variance (standard error of meadmean) for each amount level of BCNU of 2.5%f 1.8%. BCNU Standard Curve with Brain Samples-The chromatograms for BCNU extracted from brain samples were not significantly different from the BCNU powder studies in terms of unwanted peaks or baseline noise, The minimum detectable amount of BCNU in each sample was 20 ng. The overall coefficient of variance (determined at each BCNU amount level assayed) was 6.6% f 1.8%. Statistical evaluation by least square analysis of this standard curve showed a slope of 0.00243 with a y-intercept of 0.080 and a coefficient of correlation of 0.929. The extraction efficiency from whole brain samples (68%)was determined by comparison of the least square analysis equations from this curve and that of the curve from BCNU powder above. For both standard curves, computer-derived peak area ratios showed similar coefficients of variance and correlation coefficients by least square analysis. Comparisons of peak
Previously Described Nitrosourea Assays-A simple and inexpensive colorimetric assay for BCNU has been reported1-2based upon the sulfanilamide assay developed by Bratton and Mar~ha1l.l~ The lower limit of sensitivity is 0.5 pg (500 ng) of BCNU in a sample. A major advantage of the colorimetric assay is that only the intact BCNU molecule and no decomposition products are quantitated by this assay. This method, however, has only been reported for use with plasma and not with whole blood or tissue samples. Previous HPLC methods assay the nonderivatized, intact parent molecule and are reported to detect 50 ng of BCNU in a ~ample.~-llThe assay has been reported with 500-2000 pL blood samples but not with 5-10 mg tissue samples with a volume of approximately 6-12 pL. It is possible that, as in this assay, 5-10 mg tissue samples with a volume of approximately 6-12 pL, could be homogenized into a larger volume with subsequent extraction. This might maintain the lower limit of sensitivity with HPLC techniques. A possibly lower extraction efficiency with brain tissue samples as compared to plasma, however, might raise the lower limit of sensitivity with such methods. Considering the complexity of this gas chromatographic-electron detection (GC-EC) method as well as these factors, it might be concluded that the GC-EC method might represent a roughly equivalent alternative to reported HPLC methods, thus allowing the choice of a specific method to be based upon available equipment and expertise. Various other methods have been reported using GC to separate derivatized samples with mass spectrometry to assay the amount of BCNU.4-sJ3 The lower limit of sensitivity is 1 ng of BCNU in a sample but requires the use of a mass spectrometer for each sample. Another GC assay, not using mass spectrometry as a detector, has also been reported.12In this assay, BCNU is extracted from plasma by benzene. The underivatized BCNU residue is injected onto a capillary, 25 m, 0.33 mm i.d., nonpolar column. A thermionic nitrogenphosphorus-specific ( " N P ) detector is used. The lower limit of sensitivity has been reported to be 2-3 ng of total BCNU from 500 pL plasma samples. The assay has not been studied on tissue (e.g., brain) samples. The sensitivity of these previous assays makes it difficult to assay BCNU in small biopsies of brain or other tissues. Gas chromatographic-mass spectrometric assays for BCNU have high sensitivity for the detection of very small amounts of BCNU. Although the use of mass spectrometry is increasing, there is still limited availability especially when analysis of a large number of samples from a variety of animals is necessary. Radiolabeled BCNU also has been used to assay for BCNU tissue concentrations in animal studies. BCNU radiolabels, whether 3H or 14C,usually either involve the carbamoylating part of the molecule or are nonspecific and involve both ends of the molecule. For BCNU, the alkylating part of the molecule, however, is the active moiety and the carbamoylating part is the nonactive moiety. Journal of Pharmaceutical Sciences / 841 Vol. 84, No. 7, July 1995
BCNU Gas Chromatographic GC Assay-This GC assay is practical yet provides sensitivity for BCNU drug concentrations in brain tissue samples with a mass of 5-10 mg. For example, this provides the ability to quantitate BCNU concentrations in biopsies at multiple sites in and around tumorbearing tissue. With the widespread ability to perform stereotactic brain biopsies in humans, this allows study of important parameters that might influence distribution of BCNU and ultimately its potential therapeutic efficacy. As compared to ethyl ether, ethyl acetate minimized extraction of unwanted compounds from brain samples. An electron capture detector was chosen for this assay because of its high sensitivity to halogen groups. TFAA was chosen as a derivatizing agent because heptafluorobutyric anhydride and pentafluoropropionicanhydride appeared to react less completely with BCNU but more extensively with other unwanted compounds in the brain extractions. In a technical sense, this assay derivatizes denitrosated BCNU. In tissue, after denitrosation of BCNU, the resultant molecule decomposes into the alkylating and carbamoylating moieties in a very short time period, probably less than 1 s, and the decomposition products travel only a very small distance, probably only microns. For this reason, this assay, in a practical sense, can be used to quantitate intact BCNU. The separation of BCNLJ, MeCCNU, and other unwanted peaks was accomplished with the use of temperature programming and a nonpolar capillary column of small diameter (0.25 mm) and long length (60 m). Daily Use of the Assay-This GC assay has been applied to real samples for over 2 years and in over 1000 brain biopsy specimens from over 60 rabbits. The method was very useful in the determination of BCNU concentrations in different areas of the brain in this study of intravenous and intraarterial infusions via the common carotid artery. The assay was invaluable in the determination of BCNU in the small punch biopsies from this study. The GC assay was found to be easy and sensitive and has been used to determine BCNU concentrations in 2 mm diameter punch biopsies from 3 mm thick brain slices. Each biopsy had a mass of 5-10 mg. The concentration of BCNU in each sample was calculated from the amount of BCNU in each brain sample divided by the actual weight of each biopsy. The resultant concentrations document increased BCNU concentrations in the infused hemisphere during intraarterial infusion as compared to intravenous infusion (Figure 3). This GC-EC method provides a facile GC method for detection of BCNU in relatively small tissue samples without the need for mass spectrometry. It enables the important
842 /Journal of Pharmaceutical Sciences Vol. 84,No. 7, Ju& 1995
2
v cp
travenous Infusion)
BCNU - 300 mglsquare-meter over 15 minutes
1
nB “0
10 20 30 40 Time after Start of 15 minute infusion (minutes)
50
Figure 3-Example of use of assay to determine BCNU concentrations in small brain biopsies at different time points during and after intraarterial and intravenous infusions.
potential clinical assay of BCNU in actual human brain biopsies obtained by needle biopsy under computerized tomography guidance. Hopefully this might provide information that can be used to adjust clinical doses, either by intravenous or intraarterial infusion routes.
References and Notes 1. Loo, T. L.; Dion, R. L. J Pharm. Sci. 1965, 54, 809-810. 2. Kari, P.; McConnell, W. R.; Finkel, J. M.; Hill, D. L. Cancer Chemother. Pharmacol. 1980, 4, 243-248. 3. Weinkam, R. J.; Wen, J. H.; Furst, D. E.; Levin, V. A. Clin. Chem. 1978,24, 45-49. 4. Weinkam, R. J.; Lin, H. S. Anal. Chem. 1979, 51, 972-975. 5. Smith, R. G.; Cheung, L. K. J . Chromatogr. 1982, 229, 464469. 6. Smith, R. G.; Cheung, L. K.; Feun, L. G.; Loo, T. L. Biomed. Mass Spect. 1983,10,404-407. 7. Caddy, B.; Idowu, 0. R. Analyst 1982,107, 301-306. 8. Weinkam, R. J.; Liu, T. J. J. Pharm. Sci. 1982, 7 1 , 153-157. 9. Hochberg, F. H.; F’ruitt, A. A.; Davis, K.; DeBruyn, G. Neurology 1983, 33 [Suppl 21, 109. 10. Krull, I. S.; Straws, J.; Hochberg, F.; Zervas, N. T. J . Anal. Toxicol. 1981, 5, 42-46. 11. Yeager, R. L.; Oldfield, E. H.; Chatterji, D. C. J . Chromatogr. 1984,305, 496-501. 12. El-Yaziei. A,: Martin. C . R. Pharm. Res. 1988.5. 220-225. 13. Smith, k.’G.: Blackstock, S. C.; Cheung, L. K.;’ Loo, T. L. Anal. Chem. 1981,53, 1205-1208. 14. Bratton, A. C.; Marshall, E. K. J . Biol. Chem. 1939, 128, 537-550. 559500058